ffdshow tryouts project: Discussion & Development

[clsid]

@mark0077
Just drag&drop a file into it. The graph gets build automatically.

@ikarad
http://gitorious.org/users/astrange

[albain]

There is a problem since version 2787

Revision 2786 is fine
Revision 2790 is not. It corresponds to an update of libmplayer

Problem : ffdshow crashes in Media Center
I have checked, this problem is different from the one I saw earlier.

The crash is located here :

Twscale.cpp

void Tswscale::done(void)
{
if (swsc) libmplayer->sws_freeContext(swsc);swsc=NULL;
}

and the message of VS is "This may be due to a corruption of the heap".

I have no idea of how to analyze it

Thank you for any help

[clsid]

In which revision things go wrong? 2788, 2789, or 2790?

[albain]

I am not sure, I will test each of those and will get back to you

[haruhiko_yamagata]

I don't mean using the DScaler deinterlacing filters in ffdshow.
I mean using ffdshow as a filter in DScaler, to postprocess the picture in DScaler.

Was able to do it in some version of ffdshow way back.
Works for me. Please update to the latest.
I'm writing about preset settings (http://ffdshow-tryout.sourceforge.net/wiki/preset_settings). Does it help? You need a preset settings named ffdscaler.

[albain]

In which revision things go wrong? 2788, 2789, or 2790?

Unfortunately, I have compilation errors on builds 2788 and 2789, so I have to stick to 2790

[Amour]

By the way, in a dialog 'About' page, if I click 'Version details", I see
"ff_theora.dll not found".
It's much appreciated if you implement a missing function 'getVersion'.

And if you have more time, please update ff_theora. The original project
is great, however, our ff_theora is quite outdated.


Please forgive me, but as mentioned earlier this week it's my first time using a compiler, so I will probably not “implement” anything related to theora as I do not know where/how to create this missing “getVersion”.

However, I will soon post a big translation patch.

[clsid]

Unfortunately, I have compilation errors on builds 2788 and 2789, so I have to stick to 2790
Exactly when does ffdshow crash in MC? During decoding or only when using as a post-processor? Are specific settings required? Then I'll see if I can reproduce it on Vista x64 MC. Have you been able to reproduce the crash outside of MC?

In 2788 I have cleaned up the config.h file, removing malloc and free defines. Same functions are defined in mem.c/h and those are now used instead. Only difference is their names: av_free, av_malloc. Maybe a wrong name is still being used somewhere in the code. See if adding the two defines back to config.h solves the crash.

edit: I have spotted a few uses of malloc in yuv2rgb.c. I will replace them with av_malloc. That might fix it.

[haruhiko_yamagata]

Please forgive me, but as mentioned earlier this week it's my first time using a compiler, so I will probably not “implement” anything related to theora as I do not know where/how to create this missing “getVersion”.

However, I will soon post a big translation patch.
OK, never mind.
However, it's a small function. There are many similar "getVersion" in other dlls. If you are interested in C/C++, it's a good place to start. :)

And good news, translations are welcome.

[albain]

Exactly when does ffdshow crash in MC? During decoding or only when using as a post-processor? Are specific settings required? Then I'll see if I can reproduce it on Vista x64 MC. Have you been able to reproduce the crash outside of MC?

In 2788 I have cleaned up the config.h file, removing malloc and free defines. Same functions are defined in mem.c/h and those are now used instead. Only difference is their names: av_free, av_malloc. Maybe a wrong name is still being used somewhere in the code. See if adding the two defines back to config.h solves the crash.

edit: I have spotted a few uses of malloc in yuv2rgb.c. I will replace them with av_malloc. That might fix it.

Crashes occur when browsing into the video folders : the decoders are loaded to display the thumbnails for each video.

Otherwise if I play the video with wmp it works fine.

I will test your fixes when available

[clsid]

I have committed the changes. Hopefully this will fix it.

[albain]

I have committed the changes. Hopefully this will fix it.

That did it !

Thanks

[Amour]

Here is my translation update:
https://sourceforge.net/tracker/?func=detail&aid=2719172&group_id=173941&atid=867362

It also fixes some English phrases, so I had to edit other translations as well. This explains the number of modified files in this patch.

[ikarad]

I found a problem with ffdshow and MPC HOMEcinema

If I want to see DVD movie with separated subtitles (which doesn't on the dvd), subtitle aren't display if I use open dvd fuinction of MPC Homecinema and subtitle function of FFDSHOW.
I use mpc homecinema 1018 and ffdshow 2812

[clsid]

How did you name the subtitle files? If ffdshow doesn't find them, it can't display them. You can also load a subtitle file manually in ffdshow settings if needed.

[ikarad]

How did you name the subtitle files? If ffdshow doesn't find them, it can't display them. You can also load a subtitle file manually in ffdshow settings if needed.

No there isn't problem with subtitle file (.srt) because ffdshow can display them if I don't use opendvd function of MPc homecinema.
Subtitle can't be displayed by ffdshow only if I use opendvd function in mpc-homecinema (function to see dvd movie).
Even manually it doens't work if I use open dvd function of mpc homecinema.
To see subtitle I must not use opendvd function and load manually .vob file.

subtitle file
http://www.zshare.net/info.html?57838523-994f8482b31c7f381c5f87aba3c1024a

dvd sample
http://www.zshare.net/info.html?57839553-f44f5747969ad8972a800c653aafc20f

subtitle file doesn't belong to this dvd but It allow to show the problem.
If I use opendvd function to open dvd and load subtitle function of Mpc homecinema subtitle are displayed. But If i use subtile function of ffdshow instead of load subtitle function of mpc homecinema subtitles are not displayed.

[clsid]

I don't think this has anything to do with how MPC operates. It is more likely related to the "DVD Navigator" source filter. For example the DirectVobSub subtitle filter only loads when a compatible source filter is in the graph. I suspect something similar applies to ffdshow. Autoloading the correct subtitle file is certainly not going to work when playing a DVD, because when ffdshow asks the source filter for the filename it will just get an empty string (unlike when you play a separate .vob file). So you would have to manually specify the sub file in ffdshow configuration. Although that does not seem to work with DVDs either.

I suggest you use the internal subtitle renderer of MPC instead for manually loading the external subs with DVDs. Or wait for Haruhiko to comment on this.

Do you also get an "Integer division by zero" error in MPC when seeking in your sample DVD?

[haruhiko_yamagata]

I found a problem with ffdshow and MPC HOMEcinema

If I want to see DVD movie with separated subtitles (which doesn't on the dvd), subtitle aren't display if I use open dvd fuinction of MPC Homecinema and subtitle function of FFDSHOW.
I use mpc homecinema 1018 and ffdshow 2812
I have re-enabled external subtitles file for DVD. I disabled for a reason, DVDs may have DVD subtitles and closed caption. If three kind of subtitles are displayed at the same time, it's just a mess. And this really happens (strictly speaking, external subtitles are not displayed at the same time as DVD subtitles, but if DVD subtitles disappear, external subtitles appear immediately...), ffdshow just does not crash.

[haruhiko_yamagata]

Here is my translation update:
https://sourceforge.net/tracker/?func=detail&aid=2719172&group_id=173941&atid=867362

It also fixes some English phrases, so I had to edit other translations as well. This explains the number of modified files in this patch.
I couldn't apply the patch for encoding reason. Please pack modified files in a zip, rather than a patch.

[morsa]

Does it open DcvproHD MXFs ?

Is it planned for the future?

[clsid]

ffdshow is not a source filter, it therefore does not handle file formats. ffdshow can not be used to play your file.

[Amour]

I couldn't apply the patch for encoding reason. Please pack modified files in a zip, rather than a patch.
Done.
I separated the language file and the code patch.
Please apply both.

[haruhiko_yamagata]

Done.
I separated the language file and the code patch.
Please apply both.
I have committed as rev 2836.
Please update your working copy of svn and see if I didn't break encoding.

[leeperry]

I use ICL for compiling a couple of files (libmad, libfaad, libdts, etc)
is that also the case for the xvidvideo.ru builds?
and why do the latest builds from there check "reset all ffdshow settings" by default ?!
they also check "decode h264"....that wasn't the case w/ last week packages :o

and sometimes when I want to open IE6, I have to close MPC or it won't open...apparently the winamp2 plugin in ffdshow audio is blocking it from popping up :confused:

[Amour]

I have committed as rev 2836.
Please update your working copy of svn and see if I didn't break encoding.

You broke custom_messages.iss.
Maybe your patching tool isn't encoding-safe. I'll provide the file separatly.

[edit]
The file was added to the patch tracker entry.
https://sourceforge.net/tracker/?func=detail&aid=2719172&group_id=173941&atid=867362

[haruhiko_yamagata]

You broke custom_messages.iss.
Maybe your patching tool isn't encoding-safe. I'll provide the file separatly.

[edit]
The file was added to the patch tracker entry.
https://sourceforge.net/tracker/?func=detail&aid=2719172&group_id=173941&atid=867362
I have committed. Please check again.
To be safest, please pack the file in a zip. Sending process may alter the encoding.

[haruhiko_yamagata]

Revision 2815 - Directory Listing
Modified Wed Mar 25 16:42:12 2009 UTC (4 days, 21 hours ago) by clsid2

Exclude ffmpegmt.dll from x64 installer. I don't think anyone has managed to build a working version yet.
I've managed to compile ffmpeg-mt. I think it's working.
Build (http://ffdshow-tryout.sourceforge.net/doom9/ffdshow_rev2839_20090330_Q_x64.exe)

[clsid]

is that also the case for the xvidvideo.ru builds?
and why do the latest builds from there check "reset all ffdshow settings" by default ?!
they also check "decode h264"....that wasn't the case w/ last week packages :o
You would have to ask the maker of the xvidvideo.ru builds.
Reset settings is only done on the initial installation, not during upgrades. During the initial installation, the installer does not care about any existing settings, it will use the defaults that we have chosen. No settings are changed during upgrades.

[haruhiko_yamagata]

I had to update Inno Setup for ffdshow_rev2839_20090330_Q_x64.exe.

[Eragon4ever]

I have committed. Please check again.
To be safest, please pack the file in a zip. Sending process may alter the encoding.@Amour: If it is still broken, tell me. I've never had any encoding breakages (that I'm aware of) and especially French should be save here.

[Amour]

I have committed. Please check again.

Revision 2843 is fine. Thanks.

[Amour]

Eragon4ever, thanks for offering your help, but now it's fine. Here is a recap of the story:
1) On the tracker item, I provided a .patch file (28th March). haruhiko_yamagata told me it wasn't possible to apply it, so I deleted it.
2) On the tracker item, I provided a different .patch file enclosed in a .zip file (30th March). haruhiko_yamagata applyed it on rev 2836. Unfortunatly, it broke the encoding of one file.
3) I provided the broken file separatly, and it was fixed on rev 2843.

[clsid]

I've managed to compile ffmpeg-mt. I think it's working.
Build (http://ffdshow-tryout.sourceforge.net/doom9/ffdshow_rev2839_20090330_Q_x64.exe)
Does not work either. ffdshow will simply fail to load when ffmpeg-mt is selected for H.264. So I suspect a problem with the pthreads dependency.

[Eragon4ever]

Eragon4ever, thanks for offering your help, but now it's fine. Here is a recap of the story:
1) On the tracker item, I provided a .patch file (28th March). haruhiko_yamagata told me it wasn't possible to apply it, so I deleted it.
2) On the tracker item, I provided a different .patch file enclosed in a .zip file (30th March). haruhiko_yamagata applyed it on rev 2836. Unfortunatly, it broke the encoding of one file.
3) I provided the broken file separatly, and it was fixed on rev 2843.
I tried to commit even before haruhiko_yamagata and noticed the probelm with the patch but I was in a hurry and not sure if it was a problem on my side, so I didn't post about it. By the time I was back, he already made his post...
I was following the stroy ;)

[leeperry]

You would have to ask the maker of the xvidvideo.ru builds.
Reset settings is only done on the initial installation, not during upgrades. During the initial installation, the installer does not care about any existing settings, it will use the defaults that we have chosen. No settings are changed during upgrades.
got it, I will PM alexins.
even when updating, the newest build still check "reset" by default...it was better the way it was the week before :o

[STaRGaZeR]

I've managed to compile ffmpeg-mt. I think it's working.
Build (http://ffdshow-tryout.sourceforge.net/doom9/ffdshow_rev2839_20090330_Q_x64.exe)

Same problem as clsid, ffdshow just fails to connect when -mt is selected as decoder.

[clsid]

even when updating, the newest build still check "reset" by default...it was better the way it was the week before :oWhen updating you should not even see the checkbox.

[Egh]

An interesting bug detected a week or so ago.
Tried this on several latest revisions of ffdshow.

30mb m2ts sample:
http://www.mediafire.com/?mwfm4jj2zmm

Works superb. Plays well by both beliyaals and SVN MPCHC versions.

Unless I do the following:

Splitter -- Haali splitter (latest beta).
Player -- MPC-HC.
Video decoder -- MPC MPEG2 internal. (doesn't really matter here)
AAC Decoder -- ffdshow.

Each time this sample is played ffdshow looses audio after several seconds. If I use internal MPC AAC decoder then everything seems working as usual.

Replicated this problem with both x86 and x64 versions (i.e. both Haali splitter and ffdshow).

Then I even tried building a graph (to exclude potential interference from MPCHC) and was able to reproduce the bug as well.

So here there's a bug for sure, not quite certain yet whenever it is Haali's fault or ffdshow ;)

[alexins]

Does not work either. ffdshow will simply fail to load when ffmpeg-mt is selected for H.264. So I suspect a problem with the pthreads dependency.

In ffdshow 2844 x64 (http://www.xvidvideo.ru/content/view/693/1/), ffmpeg-mt works great! Neither any problems with the building I did not have.

haruhiko_yamagata - :thanks::thanks:

[breez]

my lossless 24/96 DVD-A/SACD transcodes were saturating quite badly sometimes, still need to figure out what to input for LS/RS...if anyone knows please :)

The absolute values of the coeffs for one output channel need to sum to 1, otherwise there may be clipping. Check the normalize matrix, it does just that even if the coeffs do not change in the window.

[georgevalkov]

Two weeks ago I posted about a problem:
http://forum.doom9.org/showthread.php?p=1261369#post1261369

It apears to be fixed somewhere between revisions [rev2786 - rev2832].

broken:
ffdshow_rev2633_20090125_xxl_mt.exe
ffdshow_rev2815_20090325_clsid_sse_icl10.exe
ffdshow_rev2786_20090326_dbt.exe

fixed [works properly]:
ffdshow_beta6_rev2527_20081219.exe
ffdshow_rev2832_20090328_clsid.exe
ffdshow_rev2832_20090329_dbt.exe
ffdshow_rev2844_20090330_xxl.exe

:thanks: Thank You very much for fixing it! :)

[avivahl]

Using the 32bits build of rev2844 from XvidVideo.ru, I see that the "Line Spacing" percentage (under Subtitles->Text) no longer has effect on the output. Regression?

[STaRGaZeR]

Guys, what is exacly needed for building libavcodec x64? I've been playing with MinGW64 with no luck whatsoever.

[alexins]

Guys, what is exacly needed for building libavcodec x64? I've been playing with MinGW64 with no luck whatsoever.

You can install the mingw64 + yasm, and more…
cd /src/ffmpeg
make 64BIT=yes

cd ../mplayer
make 64BIT=yes

cd ../codecs/libmpeg2
make 64BIT=yes

[leeperry]

The absolute values of the coeffs for one output channel need to sum to 1, otherwise there may be clipping. Check the normalize matrix, it does just that even if the coeffs do not change in the window.
indeed, I've added a VST vu-meter and on my loudest 24/96 5.1 tracks, I have to lower the volume to 45% in ffdshow prior to the matrix to avoid clipping > -0.1dB

I'd like to take it in account directly in the Logic 7 matrix, but w/ the inverted phases and stuff I've got no idea how to do so :o

it's not very clear how to compute coeffs : http://ac3filter.net/guides/mixing_matrix

[Leak]

Two weeks ago I posted about a problem:
http://forum.doom9.org/showthread.php?p=1261369#post1261369

It apears to be fixed somewhere between revisions [rev2786 - rev2832].
*cough*Revision 2823*cough* (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout?view=rev&revision=2823) :D

[dev84]

Hi, is ffdshow capable of adding filter like one on photo, or are there any other filters or P.S. 2 shaders capable of this?

http://img128.imageshack.us/img128/8004/snapshot20090401140735.jpg

[leeperry]

this movie was shot w/ 3 cameras to get projected on a dome screen, doing the same on regular movie content is gonna be tough :D

[haruhiko_yamagata]

Using the 32bits build of rev2844 from XvidVideo.ru, I see that the "Line Spacing" percentage (under Subtitles->Text) no longer has effect on the output. Regression?
I have fixed at rev 2849.

[dev84]

this movie was shot w/ 3 cameras to get projected on a dome screen, doing the same on regular movie content is gonna be tough :D

That`s a bummer :(

[avivahl]

I have fixed at rev 2849.Thank you (will test once alexins compile the new revision).

I also noticed that the behavior of the "Vertical Position" has changed a bit. I now have to use ~95% (instead of 90%) to get it where I want it. Isn't 90% too low as a default?

[EpsilonX]

About the subtitles in FFDShow...
Can it be used to position subtitles from a DVD...
Let's say, below the picture as in the black bar...
So it won't be on top of the video...
If it's possible, is it possible with ZoomPlayer or MPC-HC..?
IIRC both player use different method in displaying subtitle...
CMIIW...
Hoping to be able to use the smoothing option available...
Unless it's actually active all this time... :D

BTW...
IIRC before build 28xx, I didn't have to enable FFDShow's subtitle to display subtitles in ZP...
But now, it will cause the most bottom part of the monitor to be "highlighted" in white...
Like 2-5 pixel in height across the whole screen width...
But if I enable FFDShow's subtitle, it's all fine...
Tried it with a clean profile, same thing...
It only happens with the Resize filter active though...
No biggie, just curious... :D

[breez]

indeed, I've added a VST vu-meter and on my loudest 24/96 5.1 tracks, I have to lower the volume to 45% in ffdshow prior to the matrix to avoid clipping > -0.1dB

I'd like to take it in account directly in the Logic 7 matrix, but w/ the inverted phases and stuff I've got no idea how to do so :o

it's not very clear how to compute coeffs : http://ac3filter.net/guides/mixing_matrix

Sum the absolute values of the coeffs for one channel and then divide every coeff by that number. That guarantees no clipping.

Example (the first matrix you posted): 0.853 + 0.5 + 0.14 + 0.5 + 0.5 + 0.5 = 2.993

Coeffs become 0.285, 0.167, -0.047 and so on.

[leeperry]

Sum the absolute values of the coeffs for one channel and then divide every coeff by that number. That guarantees no clipping.

Example (the first matrix you posted): 0.853 + 0.5 + 0.14 + 0.5 + 0.5 + 0.5 = 2.993

Coeffs become 0.285, 0.167, -0.047 and so on.
well OK, but lowering the volume to 45% in ffdshow prior to the matrix just "works"...I never clip, even on my most dynamic 24/96 lossless stuff and these psychoacoustic coeffs do wonders :eek:

any idea how to reach 45% if 100% is 1+0.707+0.9-0.38+0.707 please? :o

http://www.image-load.eu/out.php/t156027_logic7.png (http://www.image-load.eu/out.php/i156027_logic7.png) http://www.image-load.eu/out.php/t156028_uuuu.png (http://www.image-load.eu/out.php/i156028_uuuu.png)

[leeperry]

BTW, is there any working source filter for FLAC? so I could use ffdshow's decoder?

I'm not too lucky w/ madflac on 24/96 files these days....as soon as I add ffdshow on top of it I get garbled audio on some files :confused:

[STaRGaZeR]

You can install the mingw64 + yasm, and more…
cd /src/ffmpeg
make 64BIT=yes

cd ../mplayer
make 64BIT=yes

cd ../codecs/libmpeg2
make 64BIT=yes

I've already installed them, even make64 but this doesn't matter. The problem is that as soon as it starts the compilation I get this error:

http://thumbnails12.imagebam.com/3146/898bdc31457128.gif (http://www.imagebam.com/image/898bdc31457128)

I'm using MinGW64 from here: http://sourceforge.net/project/showfiles.php?group_id=202880&package_id=311650, mingw-w64-bin_i686-mingw_20090326.zip to be exact.

[Brazil2]

http://www.image-load.eu/out.php/i155923_mix00.png
What's the point to have a negative value for the FR channel in the Left output and a negative value for the FL channel in the Right output ? I don't get it... :confused:

[leeperry]

What's the point to have a negative value for the FR channel in the Left output and a negative value for the FL channel in the Right output ? I don't get it... :confused:
phase inversion my friend, these values are based on psychoacoustic principles...try it, it's great :)

ok MPC HC has a standalone FLAC source filter, hopefully ffdshow will play all my 24/96 FLAC files :)

[Brazil2]

phase inversion
AFAIK that should be OK for DTS only.
But anyway since there is no output from the FR to the Left output I still don't get it. As we are talking about FRONT channels and not the Back and Surround ones.

[leeperry]

it's the Logic 7 matrix from Lexicon, it uses psychoacoustic principles to widen the soundstage...mostly it sends a very low copy of the opposite front channel to create a sorta "out of head" sound, but not as hollow as Dolby Headphone/Virtual Speaker, much more subtle...it's really awesome on headphones

it seems that the MPC HC flac source filter is tied to its own audio decoder, it doesn't let ffdshow decode on my XP box.....and I sure hope that it doesn't downsample 24/96 to 16/48, but at least it doesn't give garbled audio like madflac :)

HMS takes over for FLAC embedded in MKV, so ffdshow does the decoding :)

PS: I got 2 profiles for music, one for 6 channels audio, one for stereo...it works w/ madflac/WavPack/.wav files but not w/ the internal MPC HC decoder...ffdshow always loads the stereo profile even though it shows "6 channels PCM uncompressed" as input :confused:

and problem remains if I make one profile for stereo and one for the rest...

here's a 5.1 24/96 FLAC sample :
http://www.megaupload.com/?d=LYMN7DI4

here's what MPC HC says, and yet ffdshow will pick my "2 channels" profile even though it says "6 channels uncompressed" as input ?!

same problem occurs w/ AC3Filter...apparently the number of channels condition is buggy :(

CLSID: {1930D8FF-4739-4E42-9199-3B2EDEAA3BF2}
Filter: C:\96000.flac
Pin: Output

- Connection media type:

Audio: FLAC 96000Hz 6ch 13824Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {1541C5C0-CDDF-477D-BC0A-86F8AE7F8354}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 18

WAVEFORMATEX:
wFormatTag: 0xf1ac
nChannels: 6
nSamplesPerSec: 96000
nAvgBytesPerSec: 1728000
nBlockAlign: 1
wBitsPerSample: 24
cbSize: 18 (extra bytes)

[haruhiko_yamagata]

I've already installed them, even make64 but this doesn't matter. The problem is that as soon as it starts the compilation I get this error:

http://thumbnails12.imagebam.com/3146/898bdc31457128.gif (http://www.imagebam.com/image/898bdc31457128)

I'm using MinGW64 from here: http://sourceforge.net/project/showfiles.php?group_id=202880&package_id=311650, mingw-w64-bin_i686-mingw_20090326.zip to be exact.
Try 'make clean' and rebuild.

[STaRGaZeR]

Try 'make clean' and rebuild.

That did the trick :)

[puffpio]

software used:

media player classic home cinema 64 bit
ffdshow tryouts rev 2832 64 bit
windows 7 64 bit

I MPC-HC I uncheck the internal filters for H264
But ffdshow will only decode H264 if it is set to use libavcodec. If I set it to use ffmpeg-mt, then I get an error that says the following pin(s) could not find a connectable filter. is this a known issue? or how can I get ffmpeg-mt to work?

D:\test\vid.mp4::GPAC ISO Video Handler

Media Type 0:
--------------------------
Video: MPEG4 Video (H264) 320x240 29.97fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {31637661-0000-0010-8000-00AA00389B71}
formattype: FORMAT_MPEG2_VIDEO {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 161

VIDEOINFOHEADER:
rcSource: (72081392,35106736)-(34981368,1078664988)
rcTarget: (72081424,48)-(34981368,1078664988)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 333673

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 320
dwPictAspectRatioY: 240
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

MPEG2VIDEOINFO:
dwStartTimeCode: 0
cbSequenceHeader: 29
dwProfile: 0x00000042
dwLevel: 0x0000000d
dwFlags: 0x00000004

BITMAPINFOHEADER:
biSize: 40
biWidth: 320
biHeight: 240
biPlanes: 1
biBitCount: 24
biCompression: avc1
biSizeImage: 0
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 69 17 05 00 00 00 00 00 ........i.......
0030: 00 00 00 00 00 00 00 00 40 01 00 00 f0 00 00 00 ........@...ð...
0040: 00 00 00 00 00 00 00 00 28 00 00 00 40 01 00 00 ........(...@...
0050: f0 00 00 00 01 00 18 00 61 76 63 31 00 00 00 00 ð.......avc1....
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0070: 00 00 00 00 1d 00 00 00 42 00 00 00 0d 00 00 00 ........B.......
0080: 04 00 00 00|00 15 67 42 c0 0d 9a 74 0a 0f d0 80 ......gBÀ.št..Ѐ
0090: 00 00 32 00 00 0b b5 47 8a 15 50 00 04 68 ce 32 ..2...µGŠ.P..hÎ2
00a0: c8 È


Media Type 1:
--------------------------
Video: MPEG4 Video (H264) 320x240 29.97fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {31435641-0000-0010-8000-00AA00389B71}
formattype: FORMAT_MPEG2_VIDEO {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 161

VIDEOINFOHEADER:
rcSource: (72081392,35106736)-(34981656,1078664988)
rcTarget: (72081424,48)-(34981656,1078664988)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 333673

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 320
dwPictAspectRatioY: 240
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

MPEG2VIDEOINFO:
dwStartTimeCode: 0
cbSequenceHeader: 29
dwProfile: 0x00000042
dwLevel: 0x0000000d
dwFlags: 0x00000004

BITMAPINFOHEADER:
biSize: 40
biWidth: 320
biHeight: 240
biPlanes: 1
biBitCount: 24
biCompression: AVC1
biSizeImage: 0
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 69 17 05 00 00 00 00 00 ........i.......
0030: 00 00 00 00 00 00 00 00 40 01 00 00 f0 00 00 00 ........@...ð...
0040: 00 00 00 00 00 00 00 00 28 00 00 00 40 01 00 00 ........(...@...
0050: f0 00 00 00 01 00 18 00 41 56 43 31 00 00 00 00 ð.......AVC1....
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0070: 00 00 00 00 1d 00 00 00 42 00 00 00 0d 00 00 00 ........B.......
0080: 04 00 00 00|00 15 67 42 c0 0d 9a 74 0a 0f d0 80 ......gBÀ.št..Ѐ
0090: 00 00 32 00 00 0b b5 47 8a 15 50 00 04 68 ce 32 ..2...µGŠ.P..hÎ2
00a0: c8 È

[Leak]

ffdshow tryouts rev 2832 64 bit

*cough*Revision 2838*cough* (http://forum.doom9.org/showthread.php?p=1267600#post1267600)

[haruhiko_yamagata]

About the subtitles in FFDShow...
Can it be used to position subtitles from a DVD...
Let's say, below the picture as in the black bar...
So it won't be on top of the video...
If it's possible, is it possible with ZoomPlayer or MPC-HC..?
IIRC both player use different method in displaying subtitle...
CMIIW...
Hoping to be able to use the smoothing option available...
Unless it's actually active all this time... :D

It's impossible or very hard. DVD application requires 4:3 or 16:9 display aspect ratio. If ffdshow add borders, DVD menu won't work.

BTW...
IIRC before build 28xx, I didn't have to enable FFDShow's subtitle to display subtitles in ZP...
But now, it will cause the most bottom part of the monitor to be "highlighted" in white...
Like 2-5 pixel in height across the whole screen width...
But if I enable FFDShow's subtitle, it's all fine...
Tried it with a clean profile, same thing...
It only happens with the Resize filter active though...
No biggie, just curious... :D
Could you give me more details?
What is the source you are playing? Is it a file or DVD?
What is the resolution of video?
Could you send me your settings (go to 'Profiles/ Preset settings' and press 'Save to file')?

[haruhiko_yamagata]

Same problem as clsid, ffdshow just fails to connect when -mt is selected as decoder.
Please try this.

launch \Program Files\Microsoft SDKs\Windows\v6.0\Bin\Depends.exe
Drag and drop ffmpegmt.dll there.

What does it say?

[haruhiko_yamagata]

Thank you (will test once alexins compile the new revision).

I also noticed that the behavior of the "Vertical Position" has changed a bit. I now have to use ~95% (instead of 90%) to get it where I want it. Isn't 90% too low as a default?
Yes, the behavior has been changed.
I have changed default to 94%.

[Leak]

It's impossible or very hard. DVD application requires 4:3 or 16:9 display aspect ratio. If ffdshow add borders, DVD menu won't work.
I think you should say "using DVD menus with the mouse won't work". Using a remote control or the keyboard (or just taking into account that the mouseover spots are actually above/below the image) will continue to work with it...

[mark0077]

haruhiko, did you make any changes to ffdshow regarding enforcing 4:3 or 16:9 or are you still not sure what to do...

[clsid]

Please try this.

launch \Program Files (x86)\Microsoft SDKs\Windows\v6.0\Bin\Depends.exe
Drag and drop ffmpegmt.dll there.

What does it say?
Here is a screenshot:
http://i40.tinypic.com/zjch8p.png

[STaRGaZeR]

Please try this.

launch \Program Files (x86)\Microsoft SDKs\Windows\v6.0\Bin\Depends.exe
Drag and drop ffmpegmt.dll there.

What does it say?

It was in \Program Files\Microsoft SDKs\Windows\v6.0\Bin\Depends.exe

http://thumbnails12.imagebam.com/3152/7ae09c31517167.gif (http://www.imagebam.com/image/7ae09c31517167)

If you need anything else let me know. ffmpegmt.dll is from xvidvideo.ru's rev2849 x64.

[EpsilonX]

Could you give me more details?
What is the source you are playing? Is it a file or DVD?
What is the resolution of video?
Could you send me your settings (go to 'Profiles/ Preset settings' and press 'Save to file')?

Source is DVD...
Tried using clean install rev2849 from XvidVideo.Ru...
The only setting changed are...

1.Enable libmpeg2 + DVD Decoding option in Codecs
2.Set Resize to 1360x768, select "No Aspect Ratio Correction"

Notice the white line across the screen when the Subtitle option is disabled...

[STaRGaZeR]

Source is DVD...
Tried using clean install rev2849 from XvidVideo.Ru...
The only setting changed are...

1.Enable libmpeg2 + DVD Decoding option in Codecs
2.Set Resize to 1360x768, select "No Aspect Ratio Correction"

Notice the white line across the screen when the Subtitle option is disabled...

My experience:

1.- ffdshow displays DVD subtitles even when the subtitle filter is disabled and MPC's internal subtitle renderer is enabled. It does not autoload with normal files.
2.- The white line appears only if resize is enabled and ffdshow is displaying autoloaded subtitles and the subtitle filter is disabled. Changing any of these will make it disappear.

[haruhiko_yamagata]

Here is a screenshot:

Which ffmpegmt.dll did you use?
It looks like a problem in libpthreadGC2.a.

[haruhiko_yamagata]

It was in \Program Files\Microsoft SDKs\Windows\v6.0\Bin\Depends.exe

http://thumbnails12.imagebam.com/3152/7ae09c31517167.gif (http://www.imagebam.com/image/7ae09c31517167)

If you need anything else let me know. ffmpegmt.dll is from xvidvideo.ru's rev2849 x64.
It looks perfect...
I don't know why it does not load.

[clsid]

I used the one I build myself.

Yes, I also suspect a problem with libpthreadGC2.a. Which one are you currently using?
The one from MinGW64 project page (http://sourceforge.net/project/showfiles.php?group_id=202880&package_id=293298) gives me linking errors. The one build by xxl links ok, but gives broken dll.

The linking errors are:
libavcodec/pthread.o:pthread.c:(.text+0x61): undefined reference to `_imp__pthread_mutex_lock'
libavcodec/pthread.o:pthread.c:(.text+0x68): undefined reference to `_imp__pthread_mutex_unlock'
...
the library only contains definitions without the "_imp__" prefix.

[STaRGaZeR]

It looks perfect...
I don't know why it does not load.

I also tried with graphstudio64, ffdshow just does not accept the connection.

[haruhiko_yamagata]

The one from MinGW64 project page (http://sourceforge.net/project/showfiles.php?group_id=202880&package_id=293298) gives me linking errors. The one build by xxl links ok, but gives broken dll.

The linking errors are:

the library only contains definitions without the "_imp__" prefix.If you mean the one in the self extracting exe file, it's for x86. You have to apply pthreads-w64.patch to the source code and compile or download .a and .dll.

[clsid]

No, I am using the standalone libpthreadGC2.a file.

[puffpio]

*cough*Revision 2838*cough* (http://forum.doom9.org/showthread.php?p=1267600#post1267600)
nice! thanks. I'll grab myself a more recent build

[puffpio]

nice! thanks. I'll grab myself a more recent build
I installed revision 2850 of ffdshow 64 bit build (from here http://www.xvidvideo.ru/content/view/699/5/)
and still, when I set H264 to be decoded by ffmpeg-mt, media player classic home cinema still can't find a connectable filter..switching to libavcodec works fine still...

any ideas?

[haruhiko_yamagata]

@puffpio
Please read a few pages from the latest of this thread.

[morsa]

Will anybody add support for DVCproHD in FFDSHOW?

[fastplayer]

Will anybody add support for DVCproHD in FFDSHOW?
Your question has already been answered:
http://forum.doom9.org/showthread.php?p=1267163#post1267163

[ikarad]

I have re-enabled external subtitles file for DVD. I disabled for a reason, DVDs may have DVD subtitles and closed caption. If three kind of subtitles are displayed at the same time, it's just a mess. And this really happens (strictly speaking, external subtitles are not displayed at the same time as DVD subtitles, but if DVD subtitles disappear, external subtitles appear immediately...), ffdshow just does not crash.

Thanks. But There is a problem.
Now, external subtitles are displayed but they are desynchronized with video.
At the start of the dvd there is nearly 10 seconds of ad or logos displays and then, the movie start (normally subtitles start also with movie and not logos or ad).

But here, subtitles start with logos and ad. Consequently when the movie start, some subtitles are displayed before the start of movie and there is a gap with voice in the movie and subtitles.

the gap depends on the lenght of ad or logos displayed before the movie.

here an example:
http://www.zshare.net/info.html?58171633-15b887cac10caf87421ce669c23063c5

subtitle file:
http://www.zshare.net/info.html?58171999-660b36e99b3ecc4c87a68229c75747ad

ffdshow 2850,

fdshow tracker
https://sourceforge.net/tracker/?func=detail&aid=2730487&group_id=173941&atid=867360

video captured by fraps to show the problem
http://www.zshare.net/info.html?58172420-7bf85ad1e00b79286035a52df9223be1

[haruhiko_yamagata]

Thanks. But There is a problem.
Now, external subtitles are displayed but they are desynchronized with video.
At the start of the dvd there is nearly 10 seconds of ad or logos displays and then, the movie start (normally subtitles start also with movie and not logos or ad).


I knew it, please add delay (in the dialog 'Delay:' plus 'Keys & remote' numpad + and -) until I have time to fix it.

[haruhiko_yamagata]

haruhiko, did you make any changes to ffdshow regarding enforcing 4:3 or 16:9 or are you still not sure what to do...
Maybe the AR must match either with 4:3, 16:9 or container AR.
BTW, I'm working on DVD's ff and rew.

[Joniii]

Is there any way to get ffdshow working again on Windows 7 (build 7068) with Xvid/H.264/VC-1/MPEG-2?

[Leak]

Is there any way to get ffdshow working again on Windows 7 (build 7068) with Xvid/H.264/VC-1/MPEG-2?
You might want to read this thread (http://forum.doom9.org/showthread.php?t=145906); somewhere further down there's some discussion what you need to fiddle with in the registry to make MS's codecs less "special"...

(NB: I haven't used Windows 7 yet, so I can't guarantee it'll work...)

np: Autechre - Draun Quarter (Envane)

[ikarad]

I knew it, please add delay (in the dialog 'Delay:' plus 'Keys & remote' numpad + and -) until I have time to fix it.

thanks but I hope you will fix as soon as possible

[yesgrey]

anyone know what 45% of -0.38 is :confused: Not -0.844 apparently...

-0.38*0.45 = -0.171

[leeperry]

-0.38*0.45 = -0.171
indeed, you nailed it! I thought because it was out of phase it had to be divided by 0.45 :rolleyes:

ah well, hopefully the random glitches on my 24/96 5.1 content will vanish...it never occured before I started playing around w/ the ffdshow volume filter :confused:

:thanks:

[yesgrey]

iI thought because it was out of phase it had to be divided by 0.45 :rolleyes:
No, the out-of-phase is handled by the minus sign.;)

[mark0077]

Seeing as you guys are talking about audio, why does the ffdshow mixer feature to expand stereo to rear channels only give them 50% volume. Shouldn't it just send an exact copy of the front channels to the rear channels. This is what I am hoping to achieve.

[leeperry]

it's all explained on that site from David Griesinger(main DSP engineer for Lexicon/Harman Kardon) : http://www.davidgriesinger.com/

especially this PPT :
http://www.davidgriesinger.com/surround_from_stereo2.ppt

and especially this slide :
http://www.image-load.eu/out.php/t156390_mixx.png (http://www.image-load.eu/out.php/i156390_mixx.png)

you need to lower the volume of the rear channels otherwise the loudness would be too high.

if you look hard enough, he publishes all his matrixes for 5.1>stereo, stereo>5.1 and stereo>7.1....this would be a great asset to have them in ffdshow, for uncompromised SQ :)

his Logic7 matrixes are the best on the market, as I understand it!

but Haruhiko is apparently the only active ffdshow coder, and he mostly focuses on ffdshow video...so fix the matrixes yourself, the most valuable infos I found are on this engineer's site....he's got it all figured out :devil:

[albain]

@clsid, Haruhiko, _xxl

I have prepared a set of 2 C# projects

Project 1 FFDShowAPI .Net library : this library (used by my plugin Media Control) can be integrated into third party players to communicate with and control FFDShow.
Project 2 sample application in C# too that uses some features of the library (find a FFDShow instance, control playback, list and switch subtitle/audio streams...)


This code is free to use and may be promoted to enhance the integration of FFDShow inside other projects.
The only requirement is to have a .Net compiler (visual studio)


I would like to know how to integrate it into subversion.

Here is the zip package of the sources :
FFDShowAPI .Net sources (http://damienbt.free.fr/FFDShowAPI.zip)


Features of the library :

Enable and set postprocessing intensity
Enable and set picture properties (brightness, hue...)
Enable and set horizontal or vertical crop
Enable and set resize
Enable sharpen
Enable noise reduction
Enable deinterlacer
Set subtitles delay
Set audio delay
List and select an external subtitle file
List and select subtitle/audio streams
List and select chapters (whenever supported)
Fast forward/rewind functionality
Set active video preset
Set default audio or video preset
Set an OSD message to be displayed at a given position


FYI, this library uses the FFDShowRemoteAPI integrated into FFDShow video decoder.

[mark0077]

it's all explained on that site from David Griesinger(main DSP engineer for Lexicon/Harman Kardon) : http://www.davidgriesinger.com/

especially this PPT :
http://www.davidgriesinger.com/surround_from_stereo2.ppt

and especially this slide :
http://www.image-load.eu/out.php/t156390_mixx.png (http://www.image-load.eu/out.php/i156390_mixx.png)

you need to lower the volume of the rear channels otherwise the loudness would be too high.

if you look hard enough, he publishes all his matrixes for 5.1>stereo, stereo>5.1 and stereo>7.1....this would be a great asset to have them in ffdshow, for uncompromised SQ :)

his Logic7 matrixes are the best on the market, as I understand it!

but Haruhiko is apparently the only active ffdshow coder, and he mostly focuses on ffdshow video...so fix the matrixes yourself, the most valuable infos I found are on this engineer's site....he's got it all figured out :devil:

Very interesting set of slides but he seems to talk about complicated mixing with some sort of goal different to mine (just hear my stereo content from front and rear). Even talking about panning, etc makes me think this is not what I am after AT ALL.

I just want the original stereo sound untouched sent to rear channels, nothing more. This has always sounded gorgeous to me compared to any of these advanced expanders, including the dolby ones. They never have sounded good to me for some reason.

I can easily create a custom matrix in ffdshow.... I only want that matrix used obv when stereo content comes in, not on 5.1 content. Can't seem to achieve this through ffdshow presets, tried a few months back and changing preset on input channel number didn't work for some reason. I shall try again.

[leeperry]

he seems to talk about complicated mixing with some sort of goal different to mine (just hear my stereo content from front and rear)
[...]
Can't seem to achieve this through ffdshow presets, tried a few months back and changing preset on input channel number didn't work
well the ffdshow audio channels detection for automatic profiles doesn't work with some decoders(MPC HC/AC3filter), that might be your problem.

I've filed a ticket on the ffdshow bug tracker, hopefully someone will be able to look into it http://forum-images.hardware.fr/images/perso/sniperr.gif

well these algorithms make very gorgeous multichannel conversions from stereo, nothing to do w/ DPL2 and stuff...this is far more refined ;)

if I had known about ffdshow when I was a kid, I would have continued coding in K-Seka assembler on Amiga :rolleyes:

[mark0077]

well the ffdshow audio channels detection for automatic profiles doesn't work with some decoders(MPC HC/AC3filter), that might be your problem.

I've filed a ticket on the ffdshow bug tracker, hopefully someone will be able to look into it http://forum-images.hardware.fr/images/perso/sniperr.gif

well these algorithms make very gorgeous multichannel conversions from stereo, nothing to do w/ DPL2 and stuff...this is far more refined ;)

if I had known about ffdshow when I was a kid, I would have continued coding in K-Seka assembler on Amiga :rolleyes:

Yeah would be great to get that fixed. My soundcard drivers only has some of the dolby expanders, a dts neo one I think, and some D3D GX which I don't have a clue what it is. Would be fantastic to be able to force ffdshow to be loaded in EVERY app just to perform this stereo -> multichannel conversion, even games etc. Would love to plug it into winamp especially, just to do the expansion.

[leeperry]

well I've ditched all the other audio players because 1)ffdshow is unmatched for audio processing(too bad the winamp plugin support doesn't work too well, but Haruhiko & Christian Budde are looking into it) and 2)MPC opens really fast, 1/10" of a second...I like that..foobar is just as slow as KMP, they need like 1 or 2 seconds to open :rolleyes:

you can force ffdshow audio in WinDVD9 and many other apps from my experience if you check "uncompressed : all" in its settings.

PS: having the "number of channels" condition working w/ the MPC HC decoder would enable me to stop using madflac...which is giving static in WaveOut on XP :(

[fastplayer]

@clsid, Haruhiko, _xxl

I have prepared a set of 2 C# projects

Project 1 FFDShowAPI .Net library : this library (used by my plugin Media Control) can be integrated into third party players to communicate with and control FFDShow.
Project 2 sample application in C# too that uses some features of the library (find a FFDShow instance, control playback, list and switch subtitle/audio streams...)


This code is free to use and may be promoted to enhance the integration of FFDShow inside other projects.
The only requirement is to have a .Net compiler (visual studio)
Impressive work! The code is very "readable" :)
To get this project better promoted like you said, you might want to write a small summary about both managed and unmanaged remote APIs in ffdshow.
This would fit perfectly into our "new" wiki:
http://ffdshow-tryout.sourceforge.net/wiki/

[_xxl]

I have prepared a set of 2 C# projects

* Project 1 FFDShowAPI .Net library : this library (used by my plugin Media Control) can be integrated into third party players to communicate with and control FFDShow.
* Project 2 sample application in C# too that uses some features of the library (find a FFDShow instance, control playback, list and switch subtitle/audio streams...)

You should commit the source to ffdshow-tryouts svn.

[clsid]

@_xxl
64bit builds fails after your recent changes:
TimgFilterPostproc.obj : error LNK2019: unresolved external symbol simple_idct_mmx_P referenced in function "private: void __cdecl TimgFilterPostprocSpp::filter(unsigned char *,unsigned char const *,int,int,unsigned int,unsigned int,signed char const *,int,bool)" (?filter@TimgFilterPostprocSpp@@AEAAXPEAEPEBEHHIIPEBCH_N@Z)

Edit: fixed

[Jeremy Duncan]

I'm not sure what Leak did to ffdshow avisynth.

I think he said it doesn't use buffers anymore, but I see a buffer tab in the configuration options?
If I choose no buffers I see no picture but hear sound.

The performance increase is not there imho.
I have tested it with the frame interpolator by fizick.
If anything it goes from low to high cpu useage spazmatically. :D

:angry:

[leeperry]

I'm not sure what Leak did to ffdshow avisynth.
I also didn't see much change, prolly he meant that these changes only mattered for IVTC and stuff..

[Leak]

I'm not sure what Leak did to ffdshow avisynth.
I replaced the internal buffers that ffdshow was using for each buffered frame with storing the buffered frames directly in AviSynth's frame cache, saving one full-frame copy operation per buffer.

It probably has the biggest impact with HD-sized video...

And where did I write that I removed buffering?

If I choose no buffers I see no picture but hear sound.
Must be something on your end, as I get a picture no matter what.

np: Gui Boratto - Ballroom (Take My Breath Away)

[albain]

Impressive work! The code is very "readable" :)
To get this project better promoted like you said, you might want to write a small summary about both managed and unmanaged remote APIs in ffdshow.
This would fit perfectly into our "new" wiki:
http://ffdshow-tryout.sourceforge.net/wiki/

Thanks, I have added a whole section to the wiki.

I have also cleaned the code and added the missing XML documentation.
Now there is a CHM help file that describes all the APIs.

[fastplayer]

Yes, I've noticed. It's very thorough :)
If you don't mind I'd like to fix a few typos, layout issues, and the screenshots. By the way, if you want to take screenshots without cutting and cropping manually, try Screenshot Captor (http://www.donationcoder.com/Software/Mouser/screenshotcaptor/index.html) (it's free, sign-up needed though)

[leoliver]

Hi Everyone,

I have ffdshow tryouts revision 2527 installed on my Win XP Pro computer. I wanted to configure FFDShow to play a video, in the Matroska format with WMP 11. But I didn't see any .MKV Codecs in FFDShow's Codecs selection window.
Unless it's listed by another name other than Matroska .
So my question is can FFDShow be configured so I could watch a Matroska video in WMP 11?
I'm attaching a Media Info text file on the Matroska video.
Feedback will be appreciated.

[fastplayer]

MKV is a container, not a video format. Install this and your MKVs should play fine with WMP:
http://haali.cs.msu.ru/mkv/MatroskaSplitter.exe

[tsal]

For matroska *.mkv support in winxp or vista you have to install haali media splitter.

http://haali.cs.msu.ru/mkv/

Then you look for the wrong thing in ffdshow, matroska is just the name of the container, its the video and audio codecs inside this container that matter to ffdshow.

For example a lot of user made *.mkv contain h.264 codec video,
then all you need to do is enable h.264 video decoding, which by default is already on.

Hope this helps.

[Jeremy Duncan]

I replaced the internal buffers that ffdshow was using for each buffered frame with storing the buffered frames directly in AviSynth's frame cache, saving one full-frame copy operation per buffer.

It probably has the biggest impact with HD-sized video...

And where did I write that I removed buffering?

I will supply the blu ray sample and somebody else who knows how can test Leak's new avisynth tweak.

Here's the sample of the mpeg2 mt2s. link to sample from m2ts (http://www.mediafire.com/?etcix9wyz5y)

I read wrong that you removed buffering.

[e-t172]

I will supply the blu ray sample and somebody else who knows how can test Leak's new avisynth tweak.

Here's the sample of the mpeg2 mt2s. link to sample from m2ts (http://www.mediafire.com/?etcix9wyz5y).

I just tested your sample.

With avisynth disabled: no problem
With avisynth enabled: no problem
With avisynth enabled, buffering 8/8: no problem

Using ffdshow r2867 (public release).

[Leak]

With avisynth disabled: no problem
With avisynth enabled: no problem
With avisynth enabled, buffering 8/8: no problem

Using ffdshow r2867 (public release).
Didn't have time yesterday to get that sample, but I get the same with a DVD image I always use for testing.

EDIT: Just tested that sample here at work - it works just fine with and without buffering.

(With plain vanilla non-MT AviSynth 2.5.8 from AviSynth.org, FWIW...)

[albain]

Yes, I've noticed. It's very thorough :)
If you don't mind I'd like to fix a few typos, layout issues, and the screenshots

No problem

[fastplayer]

No problem
Thanks!
Are you planning to write about the unmanaged remote API that you've developed long ago?

[e-t172]

I've been playing with ffdshow's structures and I have a question.

If I understand YV12 correctly, in YV12 pictures Y values are 6-bit and UV values are 3-bit. This means that Y values are in the range [0 .. 63] (2^6) and UV values are in the range [0 .. 7] (2^3).

However, in ffdshow's FF_CSP_420P picture data, Y plane (TffPict::data[0]), I have values spread all over the [0 .. 255] range. Is this normal? Are the values normalized to [0 .. 255] by the decoder? Or am I doing something wrong? Is this the same for UV planes?

[squid_80]

If I understand YV12 correctly, in YV12 pictures Y values are 6-bit and UV values are 3-bit. This means that Y values are in the range [0 .. 63] (2^6) and UV values are in the range [0 .. 7] (2^3).
I don't think you do understand it correctly. All values are 8-bit.

[albain]

Thanks!
Are you planning to write about the unmanaged remote API that you've developed long ago?

Yes, when I will have some time, it is not on my top priority list for now, but I will do it.

[fastplayer]

Yes, when I will have some time, it is not on my top priority list for now, but I will do it.
There's no need to hurry :)

By the way, I noticed from your screenshots that the UI controls in Vista are not "themed". Can somebody look into that?

[e-t172]

I don't think you do understand it correctly. All values are 8-bit.

Indeed. I got it now: the 12 in YV12 means "12 bits per pixel", which means 8 bit for each Y value and 8 bit for each UV value because of the 4:2:0 subsampling (thus the 8-bit U/V values are spread over 2x2 pixel blocks, which means 2-bit U/V information per pixel, 8 + 2 + 2 = 12). Thanks for your time anyway.

[Jeremy Duncan]

I was thinking the sample would be used to test fps to see how the new avisynth tweak Leak and e-t172 made compares to using how Leak had it before.

[leoliver]

MKV is a container, not a video format. Install this and your MKVs should play fine with WMP:
http://haali.cs.msu.ru/mkv/MatroskaSplitter.exe

Thanks for the download link.

[Thunderbolt8]

guys, is there any news on multithreading support for the other 2 codecs, VC-1 and MPEG-2? as I once said before playback of high bitrate AVC stuff is no problem any more, but even on a c2q I get CPU problems with VC-1, especially when trying to add deinterlacing and/or sharpening (although playback alone of HighDef stuff without any other effect is already a little too much load to get it fluid for those codecs). would be great if that could be done as well.

[leoliver]

For matroska *.mkv support in winxp or vista you have to install haali media splitter.

http://haali.cs.msu.ru/mkv/

Then you look for the wrong thing in ffdshow, matroska is just the name of the container, its the video and audio codecs inside this container that matter to ffdshow.

For example a lot of user made *.mkv contain h.264 codec video,
then all you need to do is enable h.264 video decoding, which by default is already on.

Hope this helps.

The video codec is different than h.264 . So I'll try the haali media splitter .

[iron2000]

Is MPC internal DXVA H.264 decoder better then ffmpeg-mt?

Played a Full HD video with animated subtitles.
MPC internal codec handled it better than ffmpeg-mt.

[avih]

Hi,

I see an incompatible behavior between ffdshow tryouts 20080927 (Rev. 2156) and beta6 (or latest clsid build 20090409):

I use ffdshow to capture H.264 using vdub (1.9). My source is interlaced (TV SD - PAL) and I use the input image processing option of ffdshow to deinterlace and resize (in that order on the filters list on the left). On the 2156 build it works well, but on beta6 and latest build it doesn't (tried few deinterlacers such as ffmpeg's, yadif and others). Other filters on the input image processing seem to work (at least resize and image properties which I've tried). Other than the different build, all other options are identical as far as I can tell, and were properly automatically transferred to the new builds.

Has anyone tried it and can reproduce the issue? Is it a known bug?

[Leak]

My source is interlaced (TV SD - PAL) and I use the input image processing option of ffdshow to deinterlace and resize (in that order on the filters list on the left). On the 2156 build it works well, but on beta6 and latest build it doesn't (tried few deinterlacers such as ffmpeg's, yadif and others).
Have you checked "Process frames flagged as progressive"? I wouldn't be surprised if your video wasn't flagged as interlaced and thus not deinterlaced...

np: Amorphous Androgynous - Yantra (The Peppermint Tree & The Seeds Of Superconciousness)

[ACrowley]

TrueHD/libavcodec is broken with 2871 ?!

The Output is distored/wrong when i play a M2TS with TrueHD .

Works perfect with 2867 (maybe higher).
Its not caused by MPC MpegSplitter.

[avih]

Have you checked "Process frames flagged as progressive"? I wouldn't be surprised if your video wasn't flagged as interlaced and thus not deinterlaced...
...

Retried with this option checked, worked. Thanks.

In that case, and due to the fact that the user already checked the "deinterlace" filter and now expects the video to get deinterlaced, probably regardless of frame flags, I suggest a change: either somehow alert the user to the fact that a video might not be flagged as interlaced (capture using vdub wouldn't be too uncommon IMHO, and that produces the unwanted behavior), at which case deinterlace wouldn't work, or check it by default (not so nice) or, my preferred solution, rename the option (and reverse the functionality) to "Disable for frames flagged as progressive" and leave it unchecked by default.

Regardless, thanks for the help.

[clsid]

There still seems to be a problem with ffdshow and wmp/media center in the current builds. A few people have reported crashes when the player is generating thumbnails. Playback itself is ok.

[Jeremy Duncan]

ffdshow (http://sourceforge.net/project/downloading.php?group_id=173941&filename=ffdshow_rev2871_20090409_clsid_sse_icl10.exe&a=97416276)
media player classic. updated december 31, 2008 (http://downloads.sourceforge.net/guliverkli2/mplayerc_20081210.zip?modtime=1229004652&big_mirror=0)
Avisynth. march 10, 09 (http://downloads.sourceforge.net/avisynth2/Avisynth_258.exe?use_mirror=voxel)
Reclock. Updated february 16, 2009 (http://files.filefront.com/SetupReClock1834zip/;13306098;/fileinfo.html)
mvtools (http://www.megaupload.com/?d=5RQK6DR2)
2.5.7 MT. march 15, 09 (http://www.megaupload.com/?d=6YJAUXG7)
avisynth 258 mt dll (http://www.mediafire.com/?mbbymniejem)

________ advanced ffdshow: framerate doubler. (ntsc 60hz monitor compatible) ________

Codecs tab
Set Mpeg2 to Libmpeg2, and check "DVD decoding".
Set Avisynth to Avisynth, Raw video to All supported

Subtitles tab, Unchecked
Uncheck "Decode closed captions"
Uncheck "Accept embedded subs"
Uncheck "Accept SSA, ASS, ASS2 Subtitle (experimental)
Vobsub subpage, uncheck Vobsub Enable.

Avisynth tab checked
YV12 checked,
Add FFdshow Video source checked,
3:2 Pulldown box: Ignore Pulldown checked,
check Buffer back/Ahead: 0, 2

MT("""
super = MSuper(pel=2,hpad=8,vpad=8)
backward_vec = MAnalyse(super,chroma=false, blksize=16, overlap=4, isb = true, search=5, searchparam=1)
forward_vec = MAnalyse(super, chroma=false, blksize=16, overlap=4, isb = false, search=5, searchparam=1)
MFlowFps(super, backward_vec, forward_vec, num=2997, den=50, mask=0, ml=49)
""",2)

Resize & aspect tab checked
Multiply by:
1.844 (for 1280x720 16:9 aspect ratio)
2.729 (for 1920x1080 16:9 aspect ratio)
Process Pixel aspect ratio internally checked
No aspect ratio correction checked
spline
Luma Sharpen: 0.00
Accurate rounding checked

Queue & Output tab
Queue output samples unchecked
Very unstable. May cause player to crash.
No box is checked.

Output tab
rgb32 checked
high quality yv12 to rgb conversion checked

rgb conversion tab
ycbcr: itu-bt 601
contrast: standard
Output levels: Tv/Projector

mpc settings
output: vmr9 renderless, reclock audio renderer.
external filters: ffdshow audio and video decoders.

reclock settings: everything default but the video settings tab.
In the video settings tab I unchecked Determine framerate of dvd's, and set both frame rate boxes to 30.

The video has a funny artifacts where there is a trail of frames instead of a steady frame.

In this ffdshow version there is no trail of video artifact: FFDshow. Updated march 10, 2009 (http://downloads.sourceforge.net/ffdshow-tryout/ffdshow_rev2754_20090310_clsid.exe?use_mirror=osdn)

video sample (http://www.mediafire.com/?gnyhfry0xhd)
Blade, ntsc, sd dvd widescreen

The sample was ripped with dvdfab and the 50 MB sample was taken with dgsplit.

I only play from the disk and this artifact is present when played from the disk as well.

[mark0077]

TrueHD/libavcodec is broken with 2871 ?!

The Output is distored/wrong when i play a M2TS with TrueHD .

Works perfect with 2867 (maybe higher).
Its not caused by MPC MpegSplitter.

I confirm True HD is broken in 2871. Using mpc-hc for decoding True HD now. I may have found another little bug, but maybe someone can clear it up.

I have a m2ts file with True HD and some other Dolby 5.1 tracks on it. The following are the results I get with the True HD track selected.

AC3 - libavcodec
MLP - libavcodec
Result: Stuttering sound (as of new 2871 version)
Expect Result: True HD coming from my speakers, decoded properly by ffdshow.

AC3 - libavcodec
MLP - disabled
Result: Sounds OK, but filter list shows ffdshow is still doing the decoding.
Expected Result: I expected ffdshow to show its getting uncompressed audio, with mpc-hc doing the decoding of the mlp / True HD track, but ffdshow is still decoding something. I just hope its (mpc-hc?) not selecting a lesser quality 5.1 track instead of the True HD one (but True HD track still shows as being the one in use in mpc-hc menu).

AC3 - disabled
MLP - disabled
Result: Sounds OK, mpc-hc doing the decoding, but mpc-hc pin says "Audio: Dolby AC3 48000Hz 6ch 640Kbps"
Expected Result: Shouldn't it say "Dolby True HD" in pin info?

Getting lots of unexpected results.

[albain]

There still seems to be a problem with ffdshow and wmp/media center in the current builds. A few people have reported crashes when the player is generating thumbnails. Playback itself is ok.

I don't have any crash with latest builds (2871). Do you have more details (revision where is occurs, eventually FFDShow video configuration : FFDShow WMP11 checkbox ticked and alternate decoder used for some formats) so that I look more into this ?

TrueHD/libavcodec is broken with 2871 ?!

The Output is distored/wrong when i play a M2TS with TrueHD .

Works perfect with 2867 (maybe higher).
Its not caused by MPC MpegSplitter.

Weird, I haven't made any test yet but the mlp decoder has been updated in revision 2867 but not after. Can you confirm me this to make me spare some analysis time ? Thanks

*EDIT* : I confirm that the problems comes from revision 2867, probably from the update of mlpdec.c

[clsid]

One user has rev 2867 on Vista x64. Settings at defaults, raw video is disabled.

Edit: perhaps the problem is related to this bug report:
http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=1125


You can (partially) revert the MLP changes if it fixes the regression. Either the new code from FFmpeg is broken, or it is somehow incompatible with ffdshow parsing code, since it deals with access units and major sync units?
The corresponding FFmpeg revisions are: 18335 to 18338.

[princenigma]

Today I tried to load ffdshow raw video filter (ffdshow-rev2871_20090409_sse.zip download from free-codecs) in windows media encoder 9.0.2980, WME report 0x80040207 error.
I googled and found this http://forums.logitech.com/logitech/board/message?board.id=quickcam_software&thread.id=59231
seems the gus have the same error.
After some unsuccessful test, I download ffdshow-20041012-sse2.exe from sourceforge,and IT'S OK !
And I just don't know why.
Is there any chance to make the newest ffdshow builds as compatible with WME as ffdshow-20041012-sse2.exe does ?

Best Regards

[Leak]

Is there any chance to make the newest ffdshow builds as compatible with WME as ffdshow-20041012-sse2.exe does ?
0x80040207 means "VFW_E_NO_ACCEPTABLE_TYPES (There is no common media type between these pins.)" - did you enable "Raw video" on ffdshow's codecs page, i.e. set it to "all supported"?

np: Ricardo Villalobos - Dexter (Alcachofa)

[Thunderbolt8]

got a bit of image corruption with the slumdog millionaire BD (US) at 33:20, which is not present when using coreavc to play.

http://thumbnails4.imagebam.com/3253/f98fd232529454.gif (http://www.imagebam.com/image/f98fd232529454)

[fastplayer]

got a bit of image corruption with the slumdog millionaire BD (US) at 33:20, which is not present when using coreavc to play.
Which ffdshow build? libavcodec or ffmpeg-mt? What renderer? etc.
Next time you post a screenshot, use an image hoster that doesn't advertise porn... :rolleyes:

[Thunderbolt8]

ffdshow 2852, ffmpeg-mt. the problem doesnt occur with libavcodec. renderer, tried haali and avr, but didnt make a difference

[clsid]

Well, ffmpeg-mt is still experimental and is known for having such bugs. You'll have to wait until MrVacBob fixes it.

[princenigma]

0x80040207 means "VFW_E_NO_ACCEPTABLE_TYPES (There is no common media type between these pins.)" - did you enable "Raw video" on ffdshow's codecs page, i.e. set it to "all supported"?

np: Ricardo Villalobos - Dexter (Alcachofa)

yes of course :)

[ImAhNoBoDy]

I have 2 mkv files that have 2 idx/sub each (embedded), but the behavior of each file is not the same when ffdshow is being used to display the subs.

1st mkv file will only play the very first sub throughout the whole movie. Both idx/sub files will do the same thing. The 2nd mkv file will display correctly, but the one of the subs are missing when ffdshow lists them. If I extracts the idx/sub, ffdshow displays them just fine. It's only when the subs are embedded into the mkv container. I did remux them, but still the same issue. Vobsub is able to display the subs correctly. Is this a known issue? If you need a sample, I can provide it later tonight.

[albain]

One user has rev 2867 on Vista x64. Settings at defaults, raw video is disabled.

Edit: perhaps the problem is related to this bug report:
http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=1125


You can (partially) revert the MLP changes if it fixes the regression. Either the new code from FFmpeg is broken, or it is somehow incompatible with ffdshow parsing code, since it deals with access units and major sync units?
The corresponding FFmpeg revisions are: 18335 to 18338.

I remove the partial portion of code that caused a problem. It was a test on the codecId. I didn't digg it further.

[Thunderbolt8]

Well, ffmpeg-mt is still experimental and is known for having such bugs. You'll have to wait until MrVacBob fixes it.
np, heres a sample: http://www.sendspace.com/file/qb7fjc

[Jeremy Duncan]

I hope Leak is fixing the bug I reported in post 7135. What's happening leak?

[bwlonsdale]

These things are probably already known, but whatever:

High quality YV12 to RGB conversion is broken in x64 builds. It doesn't crash the player, but the video is really discolored.

DeBand is broken in x64 builds. It doesn't crash the player, but the video turns into static.

I'm sure there are other things, but those I use often.

Thank you for your reports, but I can't reproduce.

Sorry to go over (very) old ground, but I've started using x64 builds recently and have run into same issue as this very old post.

Using latest x64 build from clsid and noticed that deband turns the video into a slideshow while the audio continues as normal. When turning off deband, the video resumes where it left off, so the audio is ahead.

Is there a fix for this? I keep reading about turning on the OSD to fix it, but that doesn't seem to work.

[haruhiko_yamagata]

Sorry to go over (very) old ground, but I've started using x64 builds recently and have run into same issue as this very old post.

Using latest x64 build from clsid and noticed that deband turns the video into a slideshow while the audio continues as normal. When turning off deband, the video resumes where it left off, so the audio is ahead.

Is there a fix for this? I keep reading about turning on the OSD to fix it, but that doesn't seem to work.
MSVC2008 doesn't optimize MMX intrinsics, it's just very very slow. Please use x86 build or give up deband for now.

[haruhiko_yamagata]

I have 2 mkv files that have 2 idx/sub each (embedded), but the behavior of each file is not the same when ffdshow is being used to display the subs.

1st mkv file will only play the very first sub throughout the whole movie. Both idx/sub files will do the same thing. The 2nd mkv file will display correctly, but the one of the subs are missing when ffdshow lists them. If I extracts the idx/sub, ffdshow displays them just fine. It's only when the subs are embedded into the mkv container. I did remux them, but still the same issue. Vobsub is able to display the subs correctly. Is this a known issue? If you need a sample, I can provide it later tonight.
I think it's a known issue. Vobsub support in ffdshow is very incomplete. But could you send us the samples? I would like to add them to my collection for the occasion I get time to fix.
(It's long term todo, and my "long term" is years.)

[fastplayer]

(It's long term todo, and my "long term" is years.)
:D Dunno why but this line cracked me up! :D

[mark0077]

haruhiko thanks for the work on ff / rewind on dvd's. Just a question to ask if you have any plans to update the ffdshow colorspace / color conversion / resize code to work internally at higher bit depths. It seems madshi and his madVR renderer has pointed out not so great colorspace conversion especially when resizing in ffdshow. madshi is using 16-bit internally everywhere in his new decoder, and dithering down to the screens bit-depth at the very last stage.

Could ffdshow be brought upto scratch?

[ImAhNoBoDy]

I think it's a known issue. Vobsub support in ffdshow is very incomplete. But could you send us the samples? I would like to add them to my collection for the occasion I get time to fix.
(It's long term todo, and my "long term" is years.)

LOL, we know that. I think that's how you typed it up on the first page long ago. Anyways, here it is:

http://www.megaupload.com/?d=1NRTBD51

I put readme files in them just in case you might forget. We'll all be waiting.

[haruhiko_yamagata]

LOL, we know that. I think that's how you typed it up on the first page long ago. Anyways, here it is:

http://www.megaupload.com/?d=1NRTBD51

I put readme files in them just in case you might forget. We'll all be waiting.
Thank you for your patience.

[haruhiko_yamagata]

haruhiko thanks for the work on ff / rewind on dvd's.
It took longer than expected, 2 weeks.
Just a question to ask if you have any plans to update the ffdshow colorspace / color conversion / resize code to work internally at higher bit depths. It seems madshi and his madVR renderer has pointed out not so great colorspace conversion especially when resizing in ffdshow. madshi is using 16-bit internally everywhere in his new decoder, and dithering down to the screens bit-depth at the very last stage.

Could ffdshow be brought upto scratch?
I have a plan to implement dithering in RGB conversion.
As for resize, it needs a new internal color space YUV 8:8:8. I think it's a bit too much work. The gain of quality wouldn't be much enough to match the effort in real images.

[mark0077]

It took longer than expected, 2 weeks.

I have a plan to implement dithering in RGB conversion.
As for resize, it needs a new internal color space YUV 8:8:8. I think it's a bit too much work. The gain of quality wouldn't be much enough to match the effort in real images.

Thanks for all the work, greatly appreciated.

[STaRGaZeR]

I have a plan to implement dithering in RGB conversion.
As for resize, it needs a new internal color space YUV 8:8:8. I think it's a bit too much work. The gain of quality wouldn't be much enough to match the effort in real images.

This sounds amazing. It could be used with every renderer that accepts RGB32 input instead of relying on madVR. Do you know what's the current bit depth used for calculations, steps, etc. ffdshow is doing when high quality YUV to RGB is selected?

[haruhiko_yamagata]

This sounds amazing. It could be used with every renderer that accepts RGB32 input instead of relying on madVR. Do you know what's the current bit depth used for calculations, steps, etc. ffdshow is doing when high quality YUV to RGB is selected?
ffdshow calculates in higher than 10bit. Dithering is the one that makes his renderer special.
But please do not expect too much, I just have plan to do, as is usual my todo is... However it has higher priority than long term :D.

[STaRGaZeR]

Haha :D Well you know, your contributions are greatly appreciated, we don't care if you take any amount of time to do it.

[therealjoeblow]

haruhiko thanks for the work on ff / rewind on dvd's.

Sorry if I come across as dense here, but I read back several dozen pages, well back into March and can't find any discussion about ff/rw for DVD's, but am also having issues with it and wondering if it's something on the table for improvement. And if it's already been improved as would be implied by Mark0077's comment, can someone point me in the direction of the correct version please?

I tried the latest available build ffdshow_rev2874_20090413_dbt.exe and with Libmpeg2 it crashes in Zoom Player when loading a DVD - the screen goes black for a second on load, and then ZP crashes with the MS error report popup. Doesn't mater what renderer, this on XP-32 SP3. Libavcodec loads and plays DVD's fine in this version, but has the issues with FF/RW where I get screen corruption when in use and they just don't work smooth and fast as they should at all, just frozen frames and jumps in between, and not really responding to key presses, etc.

I also tried the next available build ffdshow_rev2871_20090409_clsid_sse_icl10.exe and in that one Libmpeg2 loads and plays DVD's fine, but FF/RW don't work very well at all, they do the same thing as all other previous builds.

Thanks for any help here,
The REAL Joe

[clsid]

Get build 2880 from sourceforge.

[therealjoeblow]

Get build 2880 from sourceforge.

Thanks clsid, that one loads fine with Libmpeg2, and FF/RW work nice and smooth!

One last issue with DVD playback now, when I enable "Set interlace flag in output media type" (Auto and Auto for method and field order), If I either manually change positions in the video uwing the time-bar, or simply chapter skip, the picture gets overall resized to about 1/2 normal, and on top of that, squashed in the vertical direction too (looks like 15-20%ish). If I toggle resize on and back off, then it resets the picture to normal size and aspect ration, but skipping ahead or back reinstates the distortion again.

If the interlace flag option is not enabled, then this behavior does not occur.

Is this a known issue, or an anomaly that can be corrected?

Many thanks

[haruhiko_yamagata]

Thanks clsid, that one loads fine with Libmpeg2, and FF/RW work nice and smooth!

One last issue with DVD playback now, when I enable "Set interlace flag in output media type" (Auto and Auto for method and field order), If I either manually change positions in the video uwing the time-bar, or simply chapter skip, the picture gets overall resized to about 1/2 normal, and on top of that, squashed in the vertical direction too (looks like 15-20%ish). If I toggle resize on and back off, then it resets the picture to normal size and aspect ration, but skipping ahead or back reinstates the distortion again.

If the interlace flag option is not enabled, then this behavior does not occur.

Is this a known issue, or an anomaly that can be corrected?

Many thanks
Which video card, video renderer, output color space? Please update your video driver.

[therealjoeblow]

Which video card, video renderer, output color space? Please update your video driver.

Sorry haruhiko, here are the details:

WinXP (32bit) SP3
ATI Radeon HD3200 (IGP on Gigabyte ma78gm-s2hp_rev.2.0 motherboard w/780G chipset)
Using the latest Catalyst v-9.4 driver
Overlay Renderer
YUY2 output colorspace (when I enable this feature YV12 is greyed out and not available; without this feature, YV12 normally gets used)

It doesn't do this with VMR9 renderers, but with VMR9, I get a black screen most times when stopping and restarting video play, and have to chapter skip to get the picture to show again. In any case, I can't use any of the VRM modes without serious tearing problems, and FS Exclusive mode disables OSD and other ZP menus, so it's not wife and kid friendly anymore :-( so I need overlay for user friendly functionality.

It also does this exactly with my Dell laptop that has an Intel Mobile 965 IGP graphics adapter, so I don't think it's specific to the video card or driver.

ANOTHER ISSUE,

I just noticed that DVD menu navigation seems to be broken, at least with the current rev 2880 clsid build. I can get to the menu, and use the cursor keys to move from item to item, but when I hit enter or mouse click to select an item, it freezes and I need to use taskman to kill ZP. It does not do this with the theatertek (nvidia) decoder or pdvd decoders, only ffdshow.

[mark0077]

ffdshow calculates in higher than 10bit. Dithering is the one that makes his renderer special.
But please do not expect too much, I just have plan to do, as is usual my todo is... However it has higher priority than long term :D.

Here is a link which seems to nicely show the improvements possible with madVR vs ffdshow for conversion. I am not sure if this increased quality is due to the dithering alone, or dithering, 16-bit internal repsentation, or all of those plus the higher color depth used during resize. A nice comparison though.

http://forum.doom9.org/showthread.php?p=1274551#post1274551

[EpsilonX]

I just noticed that DVD menu navigation seems to be broken, at least with the current rev 2880 clsid build. I can get to the menu, and use the cursor keys to move from item to item, but when I hit enter or mouse click to select an item, it freezes and I need to use taskman to kill ZP. It does not do this with the theatertek (nvidia) decoder or pdvd decoders, only ffdshow.

I can confirm this...
rev2874 is fine...
rev2877 and rev2880 from XvidVideo.ru crashed ZP...
Vista+EVR...

[leeperry]

ah...the latest ICL10 build from xvidvideo.ru still asks me what video decoders I want to enable, and whether I'd want to reset ffdshow...it worked in the earlier builds, but now the problem is back again :o

[therealjoeblow]

The crop tools don't appear to be able to autocrop a 2.35:1 movie down to 1.778:1 (ie, pan-and-scan a super widescreen down to regular widescreen in lay-man's terms) unless I'm missing something?

Would it be possible to add a feature request to do so, ie, use the same algorithm that autocrop uses to find the actual picture dimensions when removing black bars, but discard the horizontal dimension and crop to 1.778x the vertical dimension?

Or if I missed it, could someone kindly point me in the right direction?

Many thanks!

[mark0077]

haruhiko, not sure if its due to the last few builds or not, but using build 2880, I am having terrible trouble getting reliable movement through dvd menus again.

I am using the Queen Greatest Hits Menu that I uploaded to you before. Sometimes I get a freeze at the DTS Entertainment advertisement. Sometimes I get as far as the DVD Menu, can click on "Options", then highlight and click on "DTS 5.1" find and then the player just doesn't move, ends up freezing (vista saying its not responding etc).

Maybe you could check it out, something isn't right with dvd menu's still.

[therealjoeblow]

haruhiko, not sure if its due to the last few builds or not, but using build 2880, I am having terrible trouble getting reliable movement through dvd menus again.

I am using the Queen Greatest Hits Menu that I uploaded to you before. Sometimes I get a freeze at the DTS Entertainment advertisement. Sometimes I get as far as the DVD Menu, can click on "Options", then highlight and click on "DTS 5.1" find and then the player just doesn't move, ends up freezing (vista saying its not responding etc).

Maybe you could check it out, something isn't right with dvd menu's still.

EpsilonX and I reported same just a few posts above... Which player are you using for interest sake?

[mark0077]

mpc-hc beliyaal's build, version 25. I had some menu navigation slowness / pausing in recent builds but no complete freezes like I get now. Still not as good as my hardware dvd player in the living room but almost there. :D

[haruhiko_yamagata]

Sorry haruhiko, here are the details:

WinXP (32bit) SP3
ATI Radeon HD3200 (IGP on Gigabyte ma78gm-s2hp_rev.2.0 motherboard w/780G chipset)
Using the latest Catalyst v-9.4 driver
Overlay Renderer
YUY2 output colorspace (when I enable this feature YV12 is greyed out and not available; without this feature, YV12 normally gets used)

It doesn't do this with VMR9 renderers, but with VMR9, I get a black screen most times when stopping and restarting video play, and have to chapter skip to get the picture to show again. In any case, I can't use any of the VRM modes without serious tearing problems, and FS Exclusive mode disables OSD and other ZP menus, so it's not wife and kid friendly anymore :-( so I need overlay for user friendly functionality.

It also does this exactly with my Dell laptop that has an Intel Mobile 965 IGP graphics adapter, so I don't think it's specific to the video card or driver.

I can't reproduce. Please make sure no other filter is inserted between ffdshow and video renderer.

ANOTHER ISSUE,

I just noticed that DVD menu navigation seems to be broken, at least with the current rev 2880 clsid build. I can get to the menu, and use the cursor keys to move from item to item, but when I hit enter or mouse click to select an item, it freezes and I need to use taskman to kill ZP. It does not do this with the theatertek (nvidia) decoder or pdvd decoders, only ffdshow.
I have fixed this.

[mark0077]

I can't reproduce. Please make sure no other filter is inserted between ffdshow and video renderer.

I have fixed this.
Thanks :D

EDIT: DVD Menu's very responsive and look perfect for me in this latest version. SWEET

[therealjoeblow]

I can't reproduce. Please make sure no other filter is inserted between ffdshow and video renderer.

I went thru the entire graph, there's only the renderer, ffdshow audio and video decoders, and the dvd navigator (m$). I checked that all of ffdshow's settings are reset to defaults. There is no other software of filters inserted in the graph.

It does this (the weird resize and aspect ratio scrunch) *only* with the OVERLAY mixer, it does not do this with VMR7, VMR9 or Haali renderer, but it is assuredly doing it with overlay on both of my (very different) machines, and only when the "set interlace flag" is turned on.

I tried both ZP6 and MPCHC_x86_svn1044_beliyaal_25 and it does the same thing on both, so I don't think it's specific to the player.

EDIT: Also, it does not change the size or aspect ratio (in overlay mode) if I simply drag the VTS_01_1.VOB file from the DVD into the player, then it plays the file fine even if I jump around back and forth in the file, but when I open anything in DVD mode with the interlace flag on, it does the weirdness. And I don't get any of these issues with nvidia or pdvd decoders in hardware mode, only with ffdshow.

??? not sure what else the problem could be...

[Mark_A_W]

I was just wondering if there is any hope for DTS-MA decoding?

It depends on FFmpeg right? And I guess that depends on getting the right info to write a decoder...

[fastplayer]

I was just wondering if there is any hope for DTS-MA decoding?
http://www.phoronix.com/scan.php?page=article&item=ffmpeg_05_interview&num=2

[leeperry]

DVD Menu's very responsive and look perfect for me in this latest version. SWEET
even in HR?

[mark0077]

I don't use HR, it won't let me play DVDs... macrovision error lol
evr-cp is what I use

[tetsuo55]

With all these changes,

What is now the recommended DVD/Bluray Mpeg2 settings(including audio) settings?

[leeperry]

I don't use HR, it won't let me play DVDs... macrovision error lol
evr-cp is what I use
they play fine in KMP on XP SP3 w/ HR, too bad the damn menus buttons don't work...ah well

[therealjoeblow]

First of all, a BIG THANK YOU to all of the developers!! I don't know how you guys find the time to deal with all of these issues, in most cases on a day by day basis, but, really, it is *very* much appreciated.

@haruhiko_yamagata,

Thanks for the DVD fixes, menus are working great now, and so is the fast-forward and rewind. Again, much appreciated!

SUSPECTED AUDIO BUG REPORT (DTS):
--------------------------------------

Can someone please take a look at DTS passthru over SPDIF? I was going nuts trying out every video setting and driver update to try and get over a judder problem, when I read on another forum that someone thinks it's ffdshow's handling of DTS passthru, and that AC3Filter doesn't do this. I installed AC3Filter and can now confirm that there is something wrong with the way FFDShow does passthru. I tried on 2 very different machines and it does exactly the same thing on both. I never noticed it before because I only just started using SPDIF a few months ago, before that I always had DTS simply decoded and played on the soundcard.

Almost exactly every 6 seconds, there's a slight frame jump/judder. It's most apparent when the camera is panning, and in scrolling credits, etc.

It doesn't do this with AC3 at all, nor when decoding DTS, only when passing thru DTS to SPDIF.

Many thanks once again!

[haruhiko_yamagata]

I went thru the entire graph, there's only the renderer, ffdshow audio and video decoders, and the dvd navigator (m$). I checked that all of ffdshow's settings are reset to defaults. There is no other software of filters inserted in the graph.

It does this (the weird resize and aspect ratio scrunch) *only* with the OVERLAY mixer, it does not do this with VMR7, VMR9 or Haali renderer, but it is assuredly doing it with overlay on both of my (very different) machines, and only when the "set interlace flag" is turned on.

I tried both ZP6 and MPCHC_x86_svn1044_beliyaal_25 and it does the same thing on both, so I don't think it's specific to the player.

EDIT: Also, it does not change the size or aspect ratio (in overlay mode) if I simply drag the VTS_01_1.VOB file from the DVD into the player, then it plays the file fine even if I jump around back and forth in the file, but when I open anything in DVD mode with the interlace flag on, it does the weirdness. And I don't get any of these issues with nvidia or pdvd decoders in hardware mode, only with ffdshow.

??? not sure what else the problem could be...
I have tested this with Intel 82865G on Xp and it works for me.
Thank you for your detailed info, but it doesn't mean it's a bug of ffdshow. I still think it's a bug of overlay mixer + your video drivers. Other decoders just do not trigger the bugs for some reason.

[nijiko]

Is there any chance to improve COOK audio (RA8 32K Stereo) ?

[arestarh]

Hello!
I updated Russian translation and uploaded to the tracker:
http://sourceforge.net/tracker/?group_id=173941&atid=867362
Sorry for my bad English!

[travolter]

hi guys!! Fantastic proyect!!

I miss VC-1 multithread support.
I have a 9550 cuad core cpu and its not enough to play 1080 VC-1 files + ffdshow filters (sharp.. etc).

Multicore support for VC-1 would be nice to improve HD videos with ffdshow filters

[clsid]

We do not make the VC-1 decoder. We use the code from the FFmpeg project. They have a branch (ffmpeg-mt) for multi-threaded decoding, but it is still in early development. You'll have to be very patient and wait.

[STaRGaZeR]

hi guys!! Fantastic proyect!!

I miss VC-1 multithread support.
I have a 9550 cuad core cpu and its not enough to play 1080 VC-1 files + ffdshow filters (sharp.. etc).

Multicore support for VC-1 would be nice to improve HD videos with ffdshow filters

Why not use wmv9 instead? It's multithreaded.

[leeperry]

is there a way to change b and c for bicubic resize? so the negative parameter is c? and b is automatically calculated so b+(2*c)=1?

[Eragon4ever]

Hello!
I updated Russian translation and uploaded to the tracker:
http://sourceforge.net/tracker/?group_id=173941&atid=867362
Sorry for my bad English!

Thank you very much. Commited by clsid at r2886.

[Gleb Egorych]

Bug in LPCM decoder: cracking sound just after rewind. Tried Panasonic Blu-Ray demo disk, there are tracks with 2ch and 8ch LPCM, same bug for both. There are also tracks with DTS audio, no problem for them.

ffdshow r2886 from "ffdshow_rev2886_20090420_clsid.exe", MPC-HC MPEG splitter 1.2.908, ZP6.

In MPC-HC 1.2.1056 with internal TS splitter (v1.2.1056 I guess) the situation is better, but cracking noise still present.

[flanger216]

Hi!

I'm having a problem with bass management in the ffdshow audio decoder.

When enabling "LFE Crossover", all audio that is already present in the discreet LFE channel of a DD5.1 ac3 signal is muted. Hence the ".1 channel" is completely lost!
Correct would be to maintain the LFE signal and simply add the bass from the other 5 channels to it.

I'm also getting this problem in the latest builds. The 'LFE crossover' function works great on sources with no .1 channel. However, for sources with an .1 channel, it sounds like the crossover is actually replacing the LFE channel. Obviously, this kills the bass. Preferably, we should be able to specify whether to add crossover to the .1 channel, or whether the crossover is simply to be ignored when a .1 channel is present.

There wasn't anything in the bug tracker about this, so I went ahead and added it.

[clsid]

What is the last known working revision? That info is always useful for us to pinpoint where the bug was introduced.

[arestarh]

Hello!
I updated Russian translation and you committed it at r2886. Thank you!
But, during the translation some strings are missing (don't load because of
buggy code):
1)http://s39.radikal.ru/i086/0904/65/f623270f8bc7.png
2)http://s61.radikal.ru/i171/0904/b6/deaa7cf85fac.png
3)http://s47.radikal.ru/i117/0904/83/c822fa89dcf5.png

Some strings I can not translate at all(in translate mode):
1)http://s46.radikal.ru/i111/0904/26/a33cda9cbb37.png
2)http://i073.radikal.ru/0904/b6/fe2d65695de2.png
3)http://i045.radikal.ru/0904/9a/1fc0aa58e1f4.png
4)http://s56.radikal.ru/i151/0904/ff/faf501eefb49.png
5)http://i028.radikal.ru/0904/48/3bc78c9cd172.png
6)http://i031.radikal.ru/0904/4a/386c8275ffb5.png

Is it possible to correct/fix this?

[STaRGaZeR]

fix DVD timestamps calculations (libavcodec wrapper)

Thanks Haruhiko, I was going to report it. libavcodec played DVDs at ~20fps :p

[travolter]

Why not use wmv9 instead? It's multithreaded.

cause I want to use the ffdshow filter collection with my videos ;)

Thanks Clsid and STaRGaZeR for the replies!!!

Ill wait to ffmpeg-mt :)

[Mark_A_W]

cause I want to use the ffdshow filter collection with my videos ;)

Thanks Clsid and STaRGaZeR for the replies!!!

Ill wait to ffmpeg-mt :)

Pick wmv9 as the decoder for VC-1, from the fddshow codecs section, instead of libavcodec (from memory).


Then you get both.


(Or you could load ffdshow after the microsoft decoder in the filter chain, but that is not as elegant - I do this for CoreAVC.)

[travolter]

Pick wmv9 as the decoder for VC-1, from the fddshow codecs section, instead of libavcodec (from memory).


Then you get both.


(Or you could load ffdshow after the microsoft decoder in the filter chain, but that is not as elegant - I do this for CoreAVC.)

hey pal!!! big thanks for the tip! I Installed VC1 codec and its working better! ;)

[clsid]

I just discovered another weird issue with DVD playback.

"Movie FPS" on the Info&CPU page shows a wrong value. If I play a 25fps DVD, then it starts at around 19 or 20fps and the value continually increases, converging to 25fps. At the beginning it increments fast, gradually slowing down, like a parabolic function. MPC osd stats show the correct fps.

[LSudlow]

That sounds like it's doing a cumulative average FPS rather than a rolling average...

[her34]

wmv crash with ffdshow

2895, 2886, 2880 crash with wmv files

2857 works

sample file:
http://www.gametrailers.com/player/48326.html
pick HD version, click the download button to choose format

[therealjoeblow]

It's likely that the Subtitle filter in ffdshow was never intended to allow external subs to be used with DVD's, and I'm not sure if it's possible to do so technically, but I'll ask anyway.

I am able to point the subtitle filter to an external .srt file while a DVD is playing, and if I play it from the beginning without jumping around in the file, then the timecodes in the .srt match up with the proper time the subs are expected in the video, but I guess the subtitle filter isn't able to read the actual timestamp from either the video stream or the player like it can for a normal media files, so when I chapter slip or even just jump ahead a minute or two, the subs get out of sync with the video.

Is it possible to modify the behavior so that external .srt files could be loaded and read the proper timecode and display at the correct time when playing DVD's, the same way they do for other media files? This would save a lot of hassle when watching foreign films that don't have english subs on the disc itself.

Many thanks!

[max-pain]

I have problem with the audio in this video (with the last couple of SVN builds):

http://www.dcresource.com/reviews/canon/powershot_sx200-review/MVI_0093.MOV

Source: http://www.dcresource.com/reviews/canon/powershot_sx200-review/

[clsid]

That file contains uncompressed PCM audio. It plays OK without ffdshow. Enabling processing of uncompressed audio in ffdshow indeed results in bad sound.

[max-pain]

I don't have audio without ffdshow in MPC-HC:( But that's obviously not a ffdshow problem.

[STaRGaZeR]

I have problem with the audio in this video (with the last couple of SVN builds):

http://www.dcresource.com/reviews/canon/powershot_sx200-review/MVI_0093.MOV

Source: http://www.dcresource.com/reviews/canon/powershot_sx200-review/

I hear that sample without issues, with and without ffdshow.

[ikarad]

I knew it, please add delay (in the dialog 'Delay:' plus 'Keys & remote' numpad + and -) until I have time to fix it.

excuse me but i would like to known if you expect to fix this problem soon or later (some months for example)?

[Snowknight26]

Since I'm not sure how to contact the devs of ffmpeg-mt, I thought I'd post this in the next best place. Would be nice if someone could pass it on to the devs.

Some corruption is noticable in this clip with ffmpeg-mt but not with libavcodec: http://stfcc.org/misc/ffmpeg-mt.mkv

[clsid]

This is the dev you are looking for:
http://forum.doom9.org/member.php?u=72044

[Shakey_Jake33]

Apologies if this is a known issue, it's quite a mission to trawl back through this thread to find out. I use ffdshow to deinterlace content decoded by other codecs (by enabling 'Raw Video->All Supported' in the codec options), because YADIF does an excellent job. So I have CoreAVC decoding my H.264 videos, and ffdshow deinterlaces.

This doesn't work for me in Windows 7 however, even though ffdshow is set up correctly, it is unable to place itself after CoreAVC in the filter chain. This means I either have to deinterlace using CoreAVC (giving inferior results) or decode using ffdshow. This is probably a Windows 7 problem, but I figure it's worth mentioning.

Using Windows 7 Release Candidate (7100) 64-bit, MPC-HD Beliyaal build 32-bit (SVN 1044, Beliyaal fixes ver. 25), ffdshow 32-bit 2896, EVR Custom renderer.

[clsid]

That's a Windows 7 problem. Setting ffdshow as a preferred external filter in MPC might help.

[Shakey_Jake33]

^Setting 'ffdshow raw video filter' as preferred in MPC did solve the problem, thanks!

[ikarad]

edit error

[Monarc]

one question: how can i use the crf Mode of the x264 encoder inside ffdshow? is it supported?

[haruhiko_yamagata]

It's likely that the Subtitle filter in ffdshow was never intended to allow external subs to be used with DVD's, and I'm not sure if it's possible to do so technically, but I'll ask anyway.

I am able to point the subtitle filter to an external .srt file while a DVD is playing, and if I play it from the beginning without jumping around in the file, then the timecodes in the .srt match up with the proper time the subs are expected in the video, but I guess the subtitle filter isn't able to read the actual timestamp from either the video stream or the player like it can for a normal media files, so when I chapter slip or even just jump ahead a minute or two, the subs get out of sync with the video.

Is it possible to modify the behavior so that external .srt files could be loaded and read the proper timecode and display at the correct time when playing DVD's, the same way they do for other media files? This would save a lot of hassle when watching foreign films that don't have english subs on the disc itself.

Many thanks!
You are right, it's not supported. As far as my skill goes, it's technically impossible.

[STaRGaZeR]

one question: how can i use the crf Mode of the x264 encoder inside ffdshow? is it supported?

Also, it's possible to add x264 to the VfW part of ffdshow64 now that it compiles for 64-bit Windows?

[STaRGaZeR]

Bug with the lastest clsid x64 build: when an interlaced DVD is played in MPC and yadif is used for deinterlacing, jumping letters can be seen in the menus for example. This does not happen with rev2895.

DVD menu sample: here (http://www.megaupload.com/?d=R3JHNSQH) (22MB). You only have to open it as DVD in MPC and use yadif to deinterlace the menu.

[haruhiko_yamagata]

Bug with the lastest clsid x64 build: when an interlaced DVD is played in MPC and yadif is used for deinterlacing, jumping letters can be seen in the menus for example. This does not happen with rev2895.

DVD menu sample: here (http://www.megaupload.com/?d=R3JHNSQH) (22MB). You only have to open it as DVD in MPC and use yadif to deinterlace the menu.
Thanks, I have fixed at rev 2914.

[clsid]

Are all known DVD related problems fixed now?

Recently I have seen a lot of reports from people that get a crash in Vista Media Center when they browse the video library. Can anyone here confirm if this problem is still present in current ffdshow builds?

[tetsuo55]

Are all known DVD related problems fixed now?

Recently I have seen a lot of reports from people that get a crash in Vista Media Center when they browse the video library. Can anyone here confirm if this problem is still present in current ffdshow builds?

Not everything iirc, and definitely not with all combination's of filters.

[movmasty]

Thanks, I have fixed at rev 2914.
In 2895 huffyuv encoding is broken, in the output vid appears a black bar of 16 pixels on the right


-
downloaded here http://www.free-codecs.com/download/FFDshow.htm
on source forge there is another version, 2527 ,where to dl the right then?
-
same prob also with 2913, dont find the 2914
-
back to the 2527, it is ok

[mark0077]

Are all known DVD related problems fixed now?

Recently I have seen a lot of reports from people that get a crash in Vista Media Center when they browse the video library. Can anyone here confirm if this problem is still present in current ffdshow builds?

I will do a full test this evening with some DVD menus that often crash / freeze / act extremely slowly with some recent ffdshow builds.

I also don't remember haruhiko adding the code he talked about here, to enforce 16:9 or 4:3 aspect ratios during resizing of dvd's (like some renderers seem to do) as this often leads to bad aspect ratios. This isn't an ffdshow problem but a workaround that I feel needs to be in there to make resizing of dvd's useful. Will test this later also unless haruhiko has input on it.

[clsid]

I have just tested the latest revision on Vista x64 and ffdshow still crashes Media Center. It happens while browsing the video library when MC is generating thumbnails for multiple files. It seems to be some kind of race condition problem.

@Albain & Haruhiko, any idea on a solution for this nasty problem?

@all, anyone here with 32-bit Vista that can confirm if ffdshow also in 32-bit Media Center?

[Mercury_22]

I CAN'T play (some) files with LPCM sound stream(s) although MPC-HC's internal LPCM filter it's playing fine the same file(s). (FFDShow svn 2913 default settings)

Is this a known problem (bug) ?

Can this be fixed ?

Sample (http://rapidshare.com/files/226414424/LPCM_TEST.m2ts.html) or Mirror (http://www.megaupload.com/?d=AO7343UP)

[albain]

I have just tested the latest revision on Vista x64 and ffdshow still crashes Media Center. It happens while browsing the video library when MC is generating thumbnails for multiple files. It seems to be some kind of race condition problem.

@Albain & Haruhiko, any idea on a solution for this nasty problem?

@all, anyone here with 32-bit Vista that can confirm if ffdshow also in 32-bit Media Center?

I have reproduced this problem in recent revisions but randomly and only in release builds.
However I remember that the boost library was in cause something like a missing lock in a thread (and I think that the problem matches with a bug report on sourceforge tracker).
Is this possible to set in boost the number of threads to 1 for testings ?

[clsid]

I did some testing with a couple old builds and I think the problem starts at revision 2876.

[therealjoeblow]

The "set interlace flag in output media type" (formerly called HW deinterlacing) doesn't appear to be working for me. I tried it in several recent builds from 2880 to the current 2918, and it doesn't seem to do anything (XP32-SP3, VMR7 or VMR9, using ffdshow libmpeg2 to decode DVD's). When enabled, the video is still showing combing just the same as when it's disabled, it doesn't seem to trigger the hardware deinterlacing routines. PDVD and TheaterTek do, when their "use hardware accelerator" tickbox is enabled DVD's get deinterlaced by the video card.

I tried on 2 different machines, one has a gigabyte mb with an AMD 780G chipset (Radeon HD3200) and the other one has an asus mb with an AMD 690 chipset (Radeon X1250). I tried on several different versions of Catalyst, including the latest v-9.4 and it doesn't seem to make any difference, hardware deinterlacing never actually gets invoked with this ffdshow setting.

[Leak]

The "set interlace flag in output media type" (formerly called HW deinterlacing) doesn't appear to be working for me.
Make sure you force ffdshow to output NV12 video by unchecking all other colorspaces.

[clsid]

A tip useful for reproducing the MC thumbnails issue:
1) Enable view of hidden system files in the Explorer folder options
2) Delete the files "ehthumbs_vista.db" to clear the thumbnail cache. This file is created in each folder of the video library of MC.

[therealjoeblow]

Make sure you force ffdshow to output NV12 video by unchecking all other colorspaces.

I tried that, and using the OSD to verify, it shows it's outputting "NV21,VU" (not NV12), even though that's the only colorspace enabled; and it doesn't make any difference any way, the hardware deinterlacing never kicks in.

[alter4]

In all my mpeg1 25 fps 320x240 video files seeking does not work (video freeze but sound carries on) when decoding with libMpeg2 (libavcodec is ok)!
Tested on WINXp sp3 x86, MPC and WMP11, ffdshow beta6 and latest build rev2913
CPU core 2 duo 6550, nvidia 630i/7100 motherboard
Can anybody confirm the bug?

[albain]

A tip useful for reproducing the MC thumbnails issue:
1) Enable view of hidden system files in the Explorer folder options
2) Delete the files "ehthumbs_vista.db" to clear the thumbnail cache. This file is created in each folder of the video library of MC.

I have reproduced the problem but I need some help :

FFDShow crashes in TimgFilterSubtitles.cpp:572 :
boost::unique_lock<boost::mutex> lock(mutex_prefs);

An assertion is raised there :
basic_timed_mutex.hpp:64
void lock()
{ BOOST_VERIFY(timed_lock(::boost::detail::get_system_time_sentinel()));
}


http://damienbt.free.fr/Capture.PNG

EDIT : apparently there is a bug that has been fixed in boost 1.38 concerning this line.
Which version of boost are we using ?

[haruhiko_yamagata]

I tried that, and using the OSD to verify, it shows it's outputting "NV21,VU" (not NV12), even though that's the only colorspace enabled; and it doesn't make any difference any way, the hardware deinterlacing never kicks in.
I think the DVD is flagged progressive. Please try "Force bob" in "Set interlace flag in output media type"-"Method:".

[haruhiko_yamagata]

I have reproduced the problem but I need some help :

Is this issue limited to x64 version?

[albain]

Is this issue limited to x64 version?

No, I am using 32 bits version (Vista, also Windows 7 has the issue).

EDIT :

Also there is a bug with this function which has been fixed in boost 1.38 (released in february 2009) and as far as I know our version is older so this may be it


Also I have seen a similar issue in the following thread :
http://www.nabble.com/-thread--Assertion-failed-get_system_time_sentinel()-td22779251.html

Something like :
"This is undefined behaviour --- your threads are outliving the variables
they are accessing. You need to join with the thread before you destroy
the object. Since you start the thread in the constructor, it would make
sense to call join in the destructor. "

[therealjoeblow]

I think the DVD is flagged progressive. Please try "Force bob" in "Set interlace flag in output media type"-"Method:".

I tried auto, force bob and force weave on both of the different machines with exactly the same result, it still shows an interlaced picture with combing with every setting, the picture is never changed.

It's the R1 NTSC WS version of Good Night and Good Luck (2005) [original factory dvd, not a backup], but I also see the same behavior with all of the SDTV DVD's that I record on my Pioneer DVR-640 standalone dvd recorder, they are interlaced regular definition television content, and ffdshow fails to invoke the hardware deinterlacing on all of these too, while the commercial decoders somehow are be able to.

[albain]

Meanwhile, it seems that defining the following preprocessor macro solves the problem : BOOST_DISABLE_ASSERTS

This is not a definitive solution but it will help before figuring out the real problem.

[Amour]

I installed ffdshow on Windows 7 (Release Candidate), but it is not working!
Instead of that, my videos are being played using the default Windows Media Foundation system.
As I have issues with this, is there a way for me to use ffdshow filters on Windows 7 for any video?

Is ffdshow going to use Windows Media Foundation in the near future in order to override some of Windows 7 default settings?

[clsid]

The Microsoft players will use their own decoders, regardless of what framework ffdshow would use. If you want to use ffdshow then either use a different player or rename the Windows Media Foundation files (Google for more info).

[alter4]

Hey guys! Looks like nobody have read my post (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
Please, help. Could you confirm this?

[therealjoeblow]

Hey guys! Looks like nobody have read my post (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
Please, help. Could you confirm this?

You only posted the message 2 hours ago! What are you expecting - personal on demand service?

Most of the dev's for ffdshow are scattered around the world. I'd hazard a guess that some of them are sleeping right now, and may or may not respond sometime later... Often, replies come days later after someone has time to do some testing.

[_xxl]

Can Boost be updated to 1_38_0?

[rack04]

Hey guys! Looks like nobody have read my post (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
Please, help. Could you confirm this?

It would also be nice if you could post a sample.

[clsid]

ffdshow compiles fine with Boost revision 51018 (= v1.38) and 52652 (latest), but the crashing problem is still present. So some changes to the ffdshow code might also be needed as Albain pointed out. Nevertheless, updating Boost will be a good idea.

I am now building with Albain's temporary workaround.
Edit: the workaround didn't work for me.

[tetsuo55]

It might be a good idea to create a troubleshooting/bug reporting guideline similar to the one i made for MPC-HC.
Often problems reported here are vague and dev's waste a lot of time trying to reproduce

see example here
http://forum.doom9.org/showthread.php?p=1278636#post1278636

(I am willing to write it)

[Amour]

I'm not using any Microsoft players. I'm using MPC.

[clsid]

Set ffdshow as a preferred external filter in MPC. If that doesn't work, go complain to Microsoft.

[albain]

ffdshow compiles fine with Boost revision 51018 (= v1.38) and 52652 (latest), but the crashing problem is still present. So some changes to the ffdshow code might also be needed as Albain pointed out. Nevertheless, updating Boost will be a good idea.

I am now building with Albain's temporary workaround.
Edit: the workaround didn't work for me.


Okay, I will digg this around. I think that this is a design problem as pointed out in the post:
http://www.nabble.com/-thread--Assertion-failed-get_system_time_sentinel()-td22779251.html

Haruhiko solved some similar bugs maybe he has some leads.

EDIT : @clsid, have you made a "touch" (file modification) on the file where the macro is defined (to make it recompile) ?

[haruhiko_yamagata]

Okay, I will digg this around. I think that this is a design problem as pointed out in the post:
http://www.nabble.com/-thread--Assertion-failed-get_system_time_sentinel()-td22779251.html

Haruhiko solved some similar bugs maybe he has some leads.
The thread is joined (terminated) before it is destructed.

I think it will crash if the thread that locks the mutex and is being waited for by another thread crashes, the mutex may be unlocked illeaglly and the thread that is waiting for the mutex would crash with that message.

Do the settings matter? Subtitle filter, external subtitle file?

[movmasty]

You only posted the message 2 hours ago! What are you expecting - personal on demand service?

mine http://forum.doom9.org/showthread.php?p=1278711#post1278711
30 hurs ago :helpful:

[lych_necross]

Hey guys! Looks like nobody have read my post (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
Please, help. Could you confirm this?
Patience alter4, patience.

[albain]

The thread is joined (terminated) before it is destructed.

I think it will crash if the thread that locks the mutex and is being waited for by another thread crashes, the mutex may be unlocked illeaglly and the thread that is waiting for the mutex would crash with that message.

Do the settings matter? Subtitle filter, external subtitle file?

Why would the thread crash ?
The lock is necessary if the variable may be modifed while another thread is reading it. Is it the case here ?

This problem seems to occur when there are several instances of FFDShow running at the same time to extract a frame and build the thumbnails.
I think that the crash occurs because the threads termination is close to the thread creation (extracting one frame is very fast) and maybe before.
I am trying to output the error code that is generated by the assertion, it will explain the exact cause

[haruhiko_yamagata]

In all my mpeg1 25 fps 320x240 video files seeking does not work (freeze) when decoding with libMpeg2 (libavcodec is ok)!
Tested on WINXp sp3 x86, MPC and WMP11, ffdshow beta6 and latest build rev2913
CPU core 2 duo 6550, nvidia 630i/7100 motherboard
Can anybody confirm the bug?
I can't reproduce. Could you send us a sample?

[haruhiko_yamagata]

In 2895 huffyuv encoding is broken, in the output vid appears a black bar of 16 pixels on the right


-
downloaded here http://www.free-codecs.com/download/FFDshow.htm
on source forge there is another version, 2527 ,where to dl the right then?
-
same prob also with 2913, dont find the 2914
-
back to the 2527, it is ok
I don't have time to look into this for now. Sorry.

[haruhiko_yamagata]

Why would the thread crash ?
The lock is necessary if the variable may be modifed while another thread is reading it. Is it the case here ?

This problem seems to occur when there are several instances of FFDShow running at the same time to extract a frame and build the thumbnails.
I think that the crash occurs because the threads termination is close to the thread creation (extracting one frame is very fast) and maybe before.
I am trying to output the error code that is generated by the assertion, it will explain the exact cause
I found latter half of TglyphThread is over written by garbage (0xcdcd in debug build) when it crashes. It may or may not be related to subtitles.

[haruhiko_yamagata]

I found I had to protect thread creation by mutex. Is this expected? I suspect a bug of boost.
Boost must start a thread after synchronization of CPU memory cache, mustn't it?

[clsid]

I haven't tested the latest revision yet, I am building as I write this. But the crashes happened with all filters disabled, so no subtitles were involved.

[haruhiko_yamagata]

The thread is created even if there is no subtitles.

[clsid]

I have just completed a quick test on Vista x64 and ffdshow did not cause a crash :) So the problem seems to have been solved. Nice work guys ;)

[albain]

Nice work to Haruhiko who fixed the problem

[rack04]

Have there been any reports of broken LPCM decoding with ffdshow? I converted a truehd audio track from an evo file to w64 using eac3to. I then muxed the w64 audio track to m2ts using tsMuxeR. Here is a sample:

http://www.megaupload.com/?d=QT69TJXU

The audio plays using MPC-HC internal LPCM decoder and LPCM Audio Decoder (low merit) but the volume is very low.

[tetsuo55]

Is this conclusion correct:

When using a 7.1 LPCM 24/192 capable hdmi output:

I can listen to lossless(un-downconverterd 24/192):
*Dolby TrueHD and older versions
*DTS in all flavours except MA? (or is the DTS-HD not supported at all)
*LPCM in all flavours

What combination of settings would i need for this?

[clsid]

I can confirm that LPCM decoding is broken.

[mark0077]

Just posting two bugs, one of which is not really a bug but something haruhiko talked about some time ago that I think should be fixed.

Bug 1) ffdshow not loaded bug when ffdshow allows no output type to be entered. If for example a user has yv12 selected as the only output type (common for me and others) and then goes to enable "Set interlace flag in output media type", then hit save. In this scenario, yv12 would have been darkened out as not usable, but the user can still save and will then be shocked to see ffdshow doesn't load..... This has caught me out a few times over the last few months. A fix would be to popup a dialog saying please select an output type if one either isn't chosen at all, or those that are chosen are not usable due to other output options.

Bug 2) ffdshow resize problem with DVD's. I think haruhiko already knows this one well enough, where DVD's may not look in the right aspect ratio when going through ffdshow's resize. haruhiko proposed enforcing 4:3 or 16:9 as the only possible DVD aspect ratios and converting incorrect ones to one or the other somehow (as far as I know renderers are doing this workaround for us already, but in this case ffdshow would have to do the same workaround). I would love to start using ffdshow's resize but can't because of this.

[clsid]

Please post both on the SourceForge bugtracker, so they won't be forgotten.

[Mercury_22]

have there been any reports of broken lpcm decoding with ffdshow? I converted a truehd audio track from an evo file to w64 using eac3to. I then muxed the w64 audio track to m2ts using tsmuxer. Here is a sample:

http://www.megaupload.com/?d=qt69tjxu

the audio plays using mpc-hc internal lpcm decoder and lpcm audio decoder (low merit) but the volume is very low.

yes !i can't play (some) files with lpcm sound stream(s) although mpc-hc's internal lpcm filter it's playing fine the same file(s). (ffdshow svn 2913 default settings)

is this a known problem (bug) ?

Can this be fixed ?

sample (http://rapidshare.com/files/226414424/lpcm_test.m2ts.html) or mirror (http://www.megaupload.com/?d=ao7343up)

[haruhiko_yamagata]

Just posting two bugs, one of which is not really a bug but something haruhiko talked about some time ago that I think should be fixed.

Bug 1) ffdshow not loaded bug when ffdshow allows no output type to be entered. If for example a user has yv12 selected as the only output type (common for me and others) and then goes to enable "Set interlace flag in output media type", then hit save. In this scenario, yv12 would have been darkened out as not usable, but the user can still save and will then be shocked to see ffdshow doesn't load..... This has caught me out a few times over the last few months. A fix would be to popup a dialog saying please select an output type if one either isn't chosen at all, or those that are chosen are not usable due to other output options.

Bug 2) ffdshow resize problem with DVD's. I think haruhiko already knows this one well enough, where DVD's may not look in the right aspect ratio when going through ffdshow's resize. haruhiko proposed enforcing 4:3 or 16:9 as the only possible DVD aspect ratios and converting incorrect ones to one or the other somehow (as far as I know renderers are doing this workaround for us already, but in this case ffdshow would have to do the same workaround). I would love to start using ffdshow's resize but can't because of this.
OK, I can reproduce.

[ikarad]

I knew it, please add delay (in the dialog 'Delay:' plus 'Keys & remote' numpad + and -) until I have time to fix it.
excuse me to disturb you
I would like to known if you expect to fix this problem soon or later (some months for example)?

[haruhiko_yamagata]

excuse me to disturb you
I would like to known if you expect to fix this problem soon or later (some months for example)?
Sorry, it is technically very difficult. Hopeless for now.

[fastplayer]

Sorry, it is technically very difficult. Hopeless for now.
Is reverting to pre-2833 behavior an option? Or are the changes made in 2833 required for the post-2833 revisions?

[jos99]

I have a problem with ffdshow and MPEG2 TV in DVBViewer- if I have "set interlace flag in output media type" selected (for hardware deinterlacing) and the aspect ratio changes, I get 2 problems-

(i) with libavcodec the video freezes but sound carries on

(ii) with libmpeg2 DVBViewer closes

This does not happen with any other MPEG2 decoder (using XP SP3, 64x2 5000+, 780G, 3200 HD gfx, 8.12 Catalyst, 2GB RAM).

[haruhiko_yamagata]

Is reverting to pre-2833 behavior an option?
Yes, I'll revert it.

[haruhiko_yamagata]

Bug 2) ffdshow resize problem with DVD's. I think haruhiko already knows this one well enough, where DVD's may not look in the right aspect ratio when going through ffdshow's resize. haruhiko proposed enforcing 4:3 or 16:9 as the only possible DVD aspect ratios and converting incorrect ones to one or the other somehow (as far as I know renderers are doing this workaround for us already, but in this case ffdshow would have to do the same workaround). I would love to start using ffdshow's resize but can't because of this.
If you are using Media Player Classic, uncheck “View” → “Video Frame” → “Keep Aspect Ratio”.
http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example

I've written that long before and forgotten about it.

[haruhiko_yamagata]

I have a problem with ffdshow and MPEG2 TV in DVBViewer- if I have "set interlace flag in output media type" selected (for hardware deinterlacing) and the aspect ratio changes, I get 2 problems-

(i) with libavcodec the video freezes but sound carries on

(ii) with libmpeg2 DVBViewer closes

This does not happen with any other MPEG2 decoder (using XP SP3, 64x2 5000+, 780G, 3200 HD gfx, 8.12 Catalyst, 2GB RAM).
Rev 2394 may fix the issues. Please try when it is available.

[jos99]

Rev 2394 may fix the issues. Please try when it is available.
Thanks- I take it you mean 2934?

[haruhiko_yamagata]

Oops, it's 2934.

[jffulcrum]

Build 2936 on xvidvideo.ru contains broken libmplayer.dll, all players crashes on them if its functions (like post-processing) used in ffdshow. Clisd build from sourceforge.net unaffected.

[haruhiko_yamagata]

Build 2936 on xvidvideo.ru contains broken libmplayer.dll, all players crashes on them if its functions (like post-processing) used in ffdshow. Clisd build from sourceforge.net unaffected.Perhaps it's GCC 4.4.1.
It works for me. What is your CPU? Mine is Core2 Quad Q6600.

[jos99]

Oops, it's 2934.
OK I tried with 2936 (clsid and icl10), and I still get the same issue- sample:

http://rapidshare.com/files/229019781/05-04_12-13-18_ITV3_1__eng__Kojak.ts

with libmpeg2 DVBViewer closes/crashes at about 11 seconds into the clip, with libavcodec the video freezes at about 11 seconds, but sound carries on.

[Rumbah]

I hope this hasn't been posted before so here's a little? feature request:

ffdshow already notices which application called it. So it would be great if you could specify options for a specific application, for example leave it as it is for general purpose and enabling HQ RGB output for the media players. Now I just have to switch between these two options and hope that I do not forget to switch back to YV12 for encoding purposes.

[haruhiko_yamagata]

OK I tried with 2936 (clsid and icl10), and I still get the same issue- sample:

http://rapidshare.com/files/229019781/05-04_12-13-18_ITV3_1__eng__Kojak.ts

with libmpeg2 DVBViewer closes/crashes at about 11 seconds into the clip, with libavcodec the video freezes at about 11 seconds, but sound carries on.
Thanks, I can reproduce now.

[mark0077]

haruhiko, I still get DVD menu freezes with a couple of my DVD's. I can upload a sample of a simple DVD menu that freezes about 75 percent of the time when clicking "English" at the first menu.

mpc-hc shows about 20 frame drops each time this menu is shown so those are the two problems I still see with DVD menu's. Let me know if you need a sample.

EDIT: Wow very strange, i actually clicked "English" a couple of minutes ago and left the dvd menu frozen while I wrote this post, then all of a sudden the second menu of this dvd appeared. The frame drops is still at 20 so its like part of the ffdshow code got stuck / slowed down ALOT when drawing the second menu. 5 minutes to draw a menu is certainly not expected.

[clsid]

I hope this hasn't been posted before so here's a little? feature request:

ffdshow already notices which application called it. So it would be great if you could specify options for a specific application, for example leave it as it is for general purpose and enabling HQ RGB output for the media players. Now I just have to switch between these two options and hope that I do not forget to switch back to YV12 for encoding purposes.
That is already possible. See Profiles/Presets in ffdshow configuration.

[haruhiko_yamagata]

haruhiko, I still get DVD menu freezes with a couple of my DVD's. I can upload a sample of a simple DVD menu that freezes about 75 percent of the time when clicking "English" at the first menu.

mpc-hc shows about 20 frame drops each time this menu is shown so those are the two problems I still see with DVD menu's. Let me know if you need a sample.

EDIT: Wow very strange, i actually clicked "English" a couple of minutes ago and left the dvd menu frozen while I wrote this post, then all of a sudden the second menu of this dvd appeared. The frame drops is still at 20 so its like part of the ffdshow code got stuck / slowed down ALOT when drawing the second menu. 5 minutes to draw a menu is certainly not expected.Thanks for testing. Please send me a sample.

[Rumbah]

That is already possible. See Profiles/Presets in ffdshow configuration.

Ah, great, sorry for requesting then, I never saw the option and even now it took me a minute or two to find it ;) . I hope there comes a day when graphic cards can convert correctly to the right colorspace ;) .

[alter4]

Originally Posted by jos99
OK I tried with 2936 (clsid and icl10), and I still get the same issue- sample:

http://rapidshare.com/files/22901978..._eng__Kojak.ts

with libmpeg2 DVBViewer closes/crashes at about 11 seconds into the clip, with libavcodec the video freezes at about 11 seconds, but sound carries on.

May be this have some clashes with this bug (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
I forgot to tell the video freezes, but sound carries on in my case too.
Also I can't reproduce this bug at my work computer - Core duo 6550, intel G31 motherboad. Very strange.
Are there any problem with CPU type optimization when compilation?

[STaRGaZeR]

haruhiko,

ffmpeg-mt
fix blocky noise (one of some blocky bugs)

Works like a charm with several Blu-ray samples that were corrupted before. Nicely spotted.

[SBeaver]

In case someone missed my post in the mpc-hc thread, here is my sample of mono lpcm in m2ts:
http://www.mediafire.com/?1xvbq0jg2jj

As this fails to decode properly in both mpc and ffdshow, I hope it will intrest you also.

[Mr VacBob]

Sorry for disappearing, I have free time now and will restart ffmpeg-mt work tomorrow.


ffmpeg-mt
fix blocky noise (one of some blocky bugs)


That's of course the right fix (why didn't I notice dequant_coeff pointers weren't constant...) but it should update them to point to h->dequant_buffer instead of the other thread context, otherwise I think there can be a race condition with 2 threads if the other thread changes PPS again. I'll do that one.

[haruhiko_yamagata]

it should update them to point to h->dequant_buffer instead of the other thread context, otherwise I think there can be a race condition with 2 threads if the other thread changes PPS again. I'll do that one.Yeah, I can understand the point though I've never imagined that.

I've been tackling an issue with a file I've sent to you (full_metal_panic_blocky_at_100-105sec_with_ffmpeg-mt.m2ts) for these three days. It's MBAFF, and only B frames (spatial direct) are affected. I have gotten no clue except that it's a race condition. It's really difficult.

[ikarad]

Yeah, I can understand the point though I've never imagined that.

I've been tackling an issue with a file I've sent to you (full_metal_panic_blocky_at_100-105sec_with_ffmpeg-mt.m2ts) for these three days. It's MBAFF, and only B frames (spatial direct) are affected. I have gotten no clue except that it's a race condition. It's really difficult.

Ffdshow 2939 corrects some bugs of my blu-rays. But there is still some bugs of decoding (some blocks appear)
Nevertheless, these bugs don't appeared at the same moment and if I rewind the video and I read anew, sometimes there isn't the problem. It appears at random.

I also notice that when I use Telecide with avisynth (inverse telecine filter), these bugs appear more often.
my parameters with Telecide
http://nsa08.casimages.com/img/2009/05/07/mini_090507063345702457.jpg (http://www.casimages.com/img.php?i=090507063345702457.jpg)

an example
http://www.zshare.net/info.html?59712928-5e4c8510a933ca124bc6be2a49b78b18

video captured by fraps to show the problem
http://www.zshare.net/info.html?59711834-c5dad6225a37f7583d272485d86e8f20

[STaRGaZeR]

I get those Avisynth problems too. They make TIVTC and Decomb pretty useless.

[therealjoeblow]

The "set interlace flag in output media type" (formerly called HW deinterlacing) doesn't appear to be working for me. I tried it in several recent builds from 2880 to the current 2918, and it doesn't seem to do anything (XP32-SP3, VMR7 or VMR9, using ffdshow libmpeg2 to decode DVD's). When enabled, the video is still showing combing just the same as when it's disabled, it doesn't seem to trigger the hardware deinterlacing routines. PDVD and TheaterTek do, when their "use hardware accelerator" tickbox is enabled DVD's get deinterlaced by the video card.

I tried on 2 different machines, one has a gigabyte mb with an AMD 780G chipset (Radeon HD3200) and the other one has an asus mb with an AMD 690 chipset (Radeon X1250). I tried on several different versions of Catalyst, including the latest v-9.4 and it doesn't seem to make any difference, hardware deinterlacing never actually gets invoked with this ffdshow setting.

I did some more testing on this on my Dell Laptop with an Intel Mobile 965 Express IGP, WinXP32 SP3, ZoomPlayer6 (entirely different hardware platform than my other 2 desktop machines) and the following occur:

1) With OUT resize: The hardware deinterlacing DOES get enabled properly with the Overlay Mixer or Haali renderer. It does NOT with VMR7 or VMR9 (windowed, windowless or renderless).

2) With resize enabled: Hardware deinterlacing does not get enabled with any of the renderers. Never.

[edit - just checked when I got home; it does exactly as above with both of the ATI machines too, works with overlay and no resize, not with VMR or resize on]

Any idea if this is a bug that can be fixed? I see this behavior on three different hardware platforms, and can't seem to get it to work as I hoped.

[STaRGaZeR]

Are you trying to resize prior to deinterlacing? Very bad idea.

[khagaroth]

HW deinterlacing only works if the output color space is NV12, YUY2 or UYVY (in order of preference). NV12 is the only color space that supports VectorAdaptive and MotionAdaptive deinterlacing that gives the best results.

[FredThompson]

HuffYUV appears broken in recent builds. When I try to read and write to the HuffYUV format with VirtualDub there is a vertical black stripe down the extreme right edge in the input. This does not happen with the actual HuffYUV filter from Ben.

[clsid]

It helps if you can find the latest build that did work. That makes it easier for us to find when/where the bug was introduced.

[FredThompson]

I'll be quite happy to find the most recent build which works for HuffYUV. How do I see a list of all builds? SourceForge wants to show only the most recent few builds.

[tetsuo55]

How do I see a list of all builds?http://www.xvidvideo.ru/content/category/1/4/5/ (builds are compiled every few revisions, but this at least narrows it down)

[clsid]

If you look closely on the SourceForge page, you'll find hundreds of old builds:
https://sourceforge.net/project/showfiles.php?group_id=173941&package_id=199416&release_id=436746

[avivahl]

Two questions after installing rev. 2941:
1. Which is better for DVD MPEG2 decoding- libmpeg2 or libavcodec? Performance-wise? Speed-wise (on Core 2 Duo E6750)? Any known bugs for any of them? A short answer will be great (not looking into getting to technical stuff).
2. Currently, subtitle rendering seems (as far as I can see) to be aligned in a weird way. Some lines are a bit higher than previous ones. Here are two (http://i40.tinypic.com/r1c2ea.jpg) shots (http://i41.tinypic.com/okonep.jpg) demonstrating the issue (I saved them with low quality to have smaller files; look at the end-of-line dot).

[lych_necross]

@avivahl
1.) On my system, libmpeg2 is better. Its very stable and has decent performance. If you are looking for speed, download ffdshow-mt (might not be very stable).

2.) If you use ffdshow for subtitles, you might need to adjust the alignment of the subtitles on the subtitle tab. Try adjusting the horizontal/vertical positions and the alignment and see if that helps. If that doesn't work, try using directvobsub instead of ffdshow for subtitles.

[FredThompson]

(Belay this comment...)

[Gleb Egorych]

Tried rev2940, crackling sound in LPCM after rewind still presents. Can't say the last good revision, because LPCM decoder started to work (for me) only recently. Beta6 doesn't work because of missing submediatype.

[travolter]

I like to play my videos matching my refresh rate of monitor (60hz). I use crystal player to do frame interpolation (a feature inside this program).. but I wanted to try other player with that feature.

Anyone know any player with that frame interpolation feature?

FFdshow have a interlaced framerate dubler.. but really I dont notice the difference with or without. MAybe Im not using correct settings?
I would like to check the FFDshow framerate doubler VS the one from crystal player.. I need to save cpu for 1080 movies

regards

[nm]

I like to play my videos matching my refresh rate of monitor (60hz). I use crystal player to do frame interpolation (a feature inside this program).. but I wanted to try other player with that feature.

Anyone know any player with that frame interpolation feature?
You could use AviSynth for that with potentially much better results than the blending used in Crystal Player, but only SD content can be processed in realtime, at best. A GPU-based filter would be nice.

FFdshow have a interlaced framerate dubler.. but really I dont notice the difference with or without. MAybe Im not using correct settings?
I would like to check the FFDshow framerate doubler VS the one from crystal player.. I need to save cpu for 1080 movies
Which GPU do you have? It would be better to let the graphics card handle deinterlacing if it can do it well enough. GeForce 6 and Radeon X1000 or newer have pretty good deinterlacers, although the lower-end models may not be powerful enough for 1080i deinterlacing with the best methods. In ffdshow, select "set interlace flag in output media type" and use NV12 as the output colorspace (deselect everything else).

[travolter]

You could use AviSynth for that with potentially much better results than the blending used in Crystal Player, but only SD content can be processed in realtime, at best. A GPU-based filter would be nice.

thanks for the info pal!! a pity that Avisynth does not have GUI.
Using it as player is not the best option


Which GPU do you have? It would be better to let the graphics card handle deinterlacing if it can do it well enough. GeForce 6 and Radeon X1000 or newer have pretty good deinterlacers, although the lower-end models may not be powerful enough for 1080i deinterlacing with the best methods. In ffdshow, select "set interlace flag in output media type" and use NV12 as the output colorspace (deselect everything else).

done.. settings working in my GeForce 7. Now I have to set interlace of FFdshow on and try with framerate doubler?

edit:- Ok I forgot that I can use avisynth in to FFDshow.. I loaded DoubleWeave and I notice the framerate doubling effect.. anyway crystal player moves smoother

[nm]

done.. settings working in my GeForce 7. Now I have to set interlace of FFdshow on and try with framerate doubler?
No, you are supposed to let the GPU handle the bobbing. No resizing, deinterlacing or framerate doubling should be done in ffdshow.

edit:- Ok I forgot that I can use avisynth in to FFDshow.. I loaded DoubleWeave and I notice the framerate doubling effect.. anyway crystal player moves smoother
DoubleWeave is not a bobber and it doesn't do anything useful in this context.

Are you playing interlaced video? From what I gather of the bits and pieces of information on the web, Crystal Player does not have any particularly good deinterlacing filters. Ffdshow's yadif would be much better, as would hardware deinterlacing by GeForce 7, once you manage to set it up properly. But bobbing will only improve the smoothness of interlaced sources. Increasing the framerate of progressive video is another matter that requires some AviSynth trickery.

Here are some links for starters:
http://forum.doom9.org/showthread.php?t=130332
http://www.compression.ru/video/frame_rate_conversion/index_en_msu.html
http://forums.guru3d.com/showthread.php?t=288017

[travolter]

here are some links for starters:

http://forums.guru3d.com/showthread.php?t=288017

pal!!! Totally amazing!!!! This is the effect that i was looking for, and it supports mt!!!

You are my hero ;)

[alter4]

I stand my ground. I made some investigations for my bug (http://forum.doom9.org/showthread.php?p=1279113#post1279113)
In 99% cases bug appears when the first frame of mpeg-1 video is not I-Frame

Gspot screenshot for "buggy" video http://www.mediafire.com/imageview.php?quickkey=lli10ce4zsz&thumb=5
Gspot screenshot for "normal" video http://www.mediafire.com/imageview.php?quickkey=9wt2yctdzdm&thumb=5

[FredThompson]

HuffYUV appears broken in recent builds. When I try to read and write to the HuffYUV format with VirtualDub there is a vertical black stripe down the extreme right edge in the input. This does not happen with the actual HuffYUV filter from Ben.

Using the SourceForge archives I found the last version which worked properly was:

ffdshow_rev2851_20090403_clsid

The error starts with version:

ffdshow_rev2857_20090404_clsid

I assume that means the bug came to live on April 3rd or 4th of this year...

[FredThompson]

I'm sorry. My report was incorrect. The last version to properly handle HuffYUV for both reading and writing was:

ffdshow_rev2639_20090127_clsid

It was broken in:

ffdshow_rev2649_20090201_clsid

[haruhiko_yamagata]

I'm sorry. My report was incorrect. The last version to properly handle HuffYUV for both reading and writing was:

ffdshow_rev2639_20090127_clsid

It was broken in:

ffdshow_rev2649_20090201_clsid

If something is changed between 2851 and 2857, rev 2942 may improve something.

Could you test some more builds in XvidVideo.RU to narrow the range?
http://www.xvidvideo.ru/content/category/1/4/4/50/50/

[haruhiko_yamagata]

I have a problem with ffdshow and MPEG2 TV in DVBViewer- if I have "set interlace flag in output media type" selected (for hardware deinterlacing) and the aspect ratio changes, I get 2 problems-

(i) with libavcodec the video freezes but sound carries on

(ii) with libmpeg2 DVBViewer closes

This does not happen with any other MPEG2 decoder (using XP SP3, 64x2 5000+, 780G, 3200 HD gfx, 8.12 Catalyst, 2GB RAM).libavcodec freeze issue has been fixed at rev 2943.
libmpeg2 works for me. If the hardware deinterlacing is related, it may be a bug of video driver. Please try different output color spaces or different video renderer.

[travolter]

This script is really amazing for framerate doubling
http://forums.guru3d.com/showthread.php?t=288017

Its possible a similar trick to enable slowmotion in videos?

[jos99]

libavcodec freeze issue has been fixed at rev 2943.
libmpeg2 works for me. If the hardware deinterlacing is related, it may be a bug of video driver. Please try different output color spaces or different video renderer.
Thanks for working on it Haruhiko! Can you point where 2943 may be downloaded- I can't find on xvid or sourceforge?

[flanger216]

I've been playing with ffdshow's audio filter for the last couple days, and I thought I'd offer a couple of observations from an audio engineering / audiophile perspective. All of this is based off of build 2396 in MPC-HC:

LFE CROSSOVER FILTER

BUG: The crossover filter works perfectly on any source that does not already have an LFE channel. However, if an LFE channel is present in the source, then the crossover filter actually replaces it with the high-passed freqencies from the left and right channel. Ideally, there should at least be an option to have these frequencies mixed into the LFE channel, instead of outright replacing it. Most 5/6/7.1 systems still need to have all <80hz frequencies routed to the subwoofer, regardless of whether an LFE channel is present or not. This is especially true for 5.1 music, which often has very little info in the LFE channel. As it stands, enabling the crossover filter on a 5.1 source effectively kills most, if not all bass.

CONVOLVER FILTER

REQUEST: This thing's awesome. I've currently got it using one of my 32/96 reverb filters, and the results are phenomenal with very little CPU usage. However, the latency is just stratospheric: upon loading an audio file, it usually takes upwards of 5 seconds to start playback. I realize latency is part-and-parcel with using a convolver filter, but perhaps it could be lessened somewhat? There are zero-latency convolvers on the market (SIR2 (http://www.knufinke.de/sir/sir2.html) is a good example), though I don't know if zero-latency could be achieved outside of a VST plugin in a DAW sequencer...

BUG: Regardless, there is a bug in the convolver filter. As I said, there's usually a 4-5 second delay when starting to play an audio file. Unfortunately, this results in 4-5 seconds being cut-off the end of the audio; it just abruptly cuts out, and the seek bar jumps to the end.

[FredThompson]

Could you test some more builds in XvidVideo.RU to narrow the range?
http://www.xvidvideo.ru/content/category/1/4/4/50/50/
ffdshow_rev2645_20090131_sse_icl10 works properly.
ffdshow_rev2940_20090506_clsid does not.

I don't see builds 2646, 2647 or 2648.

[nm]

This script is really amazing for framerate doubling
http://forums.guru3d.com/showthread.php?t=288017

Its possible a similar trick to enable slowmotion in videos?
Sure, MVFlowFps can give you as many new interpolated frames as you want and it's possible to further improve quality with slower MVAnalyse and MVFlowFps parameters. Read MVTools (http://avisynth.org.ru/mvtools/mvtools.html) / MVTools 2 (http://avisynth.org.ru/mvtools/mvtools2.html) documentation for usage details and ask on the AviSynth subforum (http://forum.doom9.org/forumdisplay.php?f=33) if you need help.

[Jeremy Duncan]

This script is really amazing for framerate doubling
http://forums.guru3d.com/showthread.php?t=288017

Its possible a similar trick to enable slowmotion in videos?

I have answered your question in this post I will link too: link (http://www.avsforum.com/avs-vb/showthread.php?p=16436418#post16436418)

The thing is ffdshow doesn't use avisynth in the audio decoder, so I can't resample the audio to match the slower video.
So that's why I'm posting this here so maybe leak or somebody else can add the ability to resample audio playback speed to ffdshow audio decoder.

Some more info: link (http://avisynth.org/mediawiki/AssumeSampleRate)

Leak already told me in a different thread that he disabled ffdshow avisynth audio for some reason: link (http://forum.doom9.org/showthread.php?p=1261495#post1261495)

[haruhiko_yamagata]

ffdshow_rev2645_20090131_sse_icl10 works properly.
ffdshow_rev2940_20090506_clsid does not.

I don't see builds 2646, 2647 or 2648.
What is the dimension of the video?

[haruhiko_yamagata]

Tried rev2940, crackling sound in LPCM after rewind still presents. Can't say the last good revision, because LPCM decoder started to work (for me) only recently. Beta6 doesn't work because of missing submediatype.OK, I'm on it. Hope it's not too difficult.

[Leak]

Leak already told me in a different thread that he disabled ffdshow avisynth audio for some reason: link (http://forum.doom9.org/showthread.php?p=1261495#post1261495)
You're confusing the ability to set the video's framerate to some arbitrary value with "disabling avisynth audio" - ffdshow has never touched audio at all.

And since the ffdshow audio and video decoder are separate codecs you can't just feed the audio data in to the AviSynth instance used by the video decoder. So audio processing isn't going to happen, unless someone else does it.

[Inventive Software]

I'm having a problem with H.264 video in any container completely locking up my laptop, and I cannot pin it down as to which program is causing it to do so - ffdshow or MPC-HC. I've very recently (as in, the last few minutes) updated ffdshow to the latest clsid build (2940), and MPC-HC is now up to date, and there is nothing else running which could interfere with it, but at some point through an episode of something or a catchup I've downloaded, it'll freeze catastrophically, i.e. only switching it off and on again cures it. It's a right pain in the backside, to say the least, and something I need cured soon. It only happens with H.264/AVC video, not with Xvid/DivX/MPEG-2/any other video type. The audio is either AC-3, MP3, Ogg Vorbis and that has no bearing on it. I cannot work it out.

Previously, I was running clsid's 2719 build. Has something changed since then, as I've been out of the loop a while as to bug fixes, and this seems like a bug somewhere that needs to be squashed, either in MPC-HC or (more likely) ffdshow-tryout.

[mark0077]

BUG 1: DVD Menu's hangning

haruhiko, I am still awaiting my broadband to be reenabled to upload dvd menu which shows the problems I spoke about.


BUG 2: Player crash in ts files at certain frame(s).
I have another problem however that only happens with ffdshow in the chain. Playing the dark knight blu-ray rip, ts file, and doing all decoding in mpc-hc with ffdshow doing nothing (even outputting original yv12 or rgb32) I get a crash of mpc-hc after frame 1274. Sometimes seeking around in the file and then playing over this frame can repeat the crash, sometimes it passes the frame without problems. Very strange indeed.

In the Info area of ffdshow I notice something strange with this file, I am not sure if this is what may be causing the problem or not, but the Input Bitrate constantly grows and grows, it never levels out. Its now above 800,000kbps and rising as I am playing this file. I can repeatedly crash mpc-hc though with ffdshow in the chain and playing directly from frame 0-1274. If there is anything you would like me to try / check let me know, otherwise I will provide the first 1300~ frames if its not too big of a sample.

This is with mpc-hc 1104 and tested with various renderers.

I have attached a screenshot which shows ffdshow info and the crash. This frame may be unique in a way because it is a completely black frame, followed by a very bright scene. Just a guess but maybe ffdshow has some internal code that looks for completely black frames? This is just a guess though, but seems strange that it hangs right as the black frame is about to be replaced by a very bright one.

http://img7.imageshack.us/img7/4853/strangebug.th.jpg (http://img7.imageshack.us/my.php?image=strangebug.jpg)

Problem 3: Bad yuy2 -> RGB32 conversion

haruhiko, I thought yuy2 to yv12 and vice versa could be done losslessly. I notice yuy2 -> RGB32HQ gives me less quality than yv12 -> RGB32HQ. I can post screenshots if necessary. Its slightly worse.

I am worried about this because various decoders like mpc-hc vc-1 decoder outputs in yuy2 by default. nvidia's yuy2 -> RGB32HQ is off the charts bad so I am hoping for two things

1) mpc-hc tell me why their decoder doesn't output yv12 by default
2) and hoping haruhiko might be able to improve yuy2 -> RGB32 as its a little bit worse than yv12 -> RGB32 at the moment.

Thanks.

[Mixer73]

MPC-HC is now up to date, and there is nothing else running which could interfere with it, but at some point through an episode of something or a catchup I've downloaded, it'll freeze catastrophically, i.e. only switching it off and on again cures it.

My suggestion on this would be simply your laptop's cooling. Really for the most part it will take a hardware issue to lock the computer up like this.

[Jeremy Duncan]

You're confusing the ability to set the video's framerate to some arbitrary value with "disabling avisynth audio" - ffdshow has never touched audio at all.

And since the ffdshow audio and video decoder are separate codecs you can't just feed the audio data in to the AviSynth instance used by the video decoder. So audio processing isn't going to happen, unless someone else does it.

Ok. I don't think ffdshow audio decoder needs avisynth anyway.
I asked because Travolter needed I thought but he said he didn't.

But he asked that there by a keyboard shortcut that lets him switch between frame doubling using avisynth and slo motion using avisynth.
So I think if there is two ffdshow video decoders , one playing frame interpolation and the other slo mo. The one used by mpc is changed by a keyboard shortcut would do what Travolter asked.

In order to do this both ffdshow video decoders would be at the same place in the graph and simply switched at the press of a button.

Is this possible? And if it is can you please do it? :)

Edit,

Maybe instead of two ffdshow video decoders the one ffdshow video decoder can have two avisynth tabs checkable. And with a keyboard shortcut the one avisynth box is checked and the other unchecked, and vice versa.

The only thing is maybe this option is hidden by default and shown in a advanced mode so it's only used by advanced users.

[Inventive Software]

My suggestion on this would be simply your laptop's cooling. Really for the most part it will take a hardware issue to lock the computer up like this.

I've done long stress tests before and never had it happen. It only happens with MPC-HC and ffdshow decoding H.264.

[fastplayer]

I've done long stress tests before and never had it happen. It only happens with MPC-HC and ffdshow decoding H.264.

ffdshow runs in user mode. It can't crash the entire machine.
Lots of under-the-hood changes have been made recently to MPC-HC thanks to Beliyaal. This introduced quite a few bugs... I suggest you get a pre-Beliyaal build (< 1048) and see if those crashes happen again.

[haruhiko_yamagata]

Problem 3: Bad yuy2 -> RGB32 conversion

haruhiko, I thought yuy2 to yv12 and vice versa. I notice yuy2 -> RGB32HQ gives me less quality than yv12 -> RGB32HQ. I can post screenshots if necessary. Its slightly worse.

I am worried about this because various decoders like mpc-hc vc-1 decoder outputs in yuy2 by default. nvidia's yuy2 -> RGB32HQ is off the charts bad so I am hoping for two things

1) mpc-hc tell me why their decoder doesn't output yv12 by default
2) and hoping haruhiko might be able to improve yuy2 -> RGB32 as its a little bit worse than yv12 -> RGB32 at the moment.

Thanks.
OK, I wait for the samples. (BUG 1,2)
Problem 3:
Is it native YUY2 format? If the decoder outputs YUY2, YV12->YUY2 conversion is done in the decoder. Probably it's losing the quality.

[haruhiko_yamagata]

I've done long stress tests before and never had it happen. It only happens with MPC-HC and ffdshow decoding H.264.
If you can find a regression, please tell us the last build that works and the oldest recent build that does not work.

[mark0077]

OK, I wait for the samples. (BUG 1,2)
Problem 3:
Is it native YUY2 format? If the decoder outputs YUY2, YV12->YUY2 conversion is done in the decoder. Probably it's losing the quality.

Thank you, I will post in the mpc-hc forums regarding yv12 -> yuy2 conversion. When I output a DVD from mpc-hc in yuy2, it must be doing the conversion incorrectly.

[leeperry]

I know there's several active ffdshow patchers..any chance someone could please get the winamp2 plugins to work in 32float? FFX-4 does say "32float" in its properties window....but Ozone4's bitmeter is locked on 16int.

apparently line 48 is at fault :
http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/audioFilters/winamp2/Twinamp2.h?view=markup&pathrev=2917

would swapping "int16_t" for "float" work ? anyone could patch it please? Ozone4 works internally in 64float & has fantastic noiseshaping options, that's a bummer ffdshow converts back and forth to 16int :(

:thanks:

[Mixer73]

I've done long stress tests before and never had it happen. It only happens with MPC-HC and ffdshow decoding H.264.

Synthetic tests are not worth a pinch of shit compared to playing h264, this stresses the CPU, GPU and memory all at once.

I suggest you monitor the temperatures and see what comes of it.

[FredThompson]

What is the dimension of the video?

800x600 HuffYUV created by ffdshow. I've used files which were created by various builds of ffdshow, going back a couple of years but I couldn't possibly determine which versions were used to make these files.

I load a HuffYUV file into Virtualdub, disable audio out and set video compression to ffdshow (set to HuffYUV.) As soon as Virtualdub starts to write the file, both the input and output panes of Virtualdub show a vertical black strip on the right-hand edge. My guess is that is a 16-pixel block but I've not measured them. This would also seem to be a problem with corruption to the read stream as it is served from ffdshow but that does not mean there is not a similar problem with writing because nulling something which is already null would appear to be just the first nullification.

[haruhiko_yamagata]

800x600 HuffYUV created by ffdshow. I've used files which were created by various builds of ffdshow, going back a couple of years but I couldn't possibly determine which versions were used to make these files.

I load a HuffYUV file into Virtualdub, disable audio out and set video compression to ffdshow (set to HuffYUV.) As soon as Virtualdub starts to write the file, both the input and output panes of Virtualdub show a vertical black strip on the right-hand edge. My guess is that is a 16-pixel block but I've not measured them. This would also seem to be a problem with corruption to the read stream as it is served from ffdshow but that does not mean there is not a similar problem with writing because nulling something which is already null would appear to be just the first nullification.
I can't reproduce. I can create/re-encode 800x600 huffyuv.
Could you send me a sample?

[FredThompson]

I can't reproduce. I can create/re-encode 800x600 huffyuv.
Could you send me a sample?http://www.corrugatedmachines.com/RE%204-Corner%20Tray%20v2%20HuffYUV%2030fps%20short.rar

That is one of the files that gives me headaches. It's about 30 Megabytes.

The only filters used were VirtualDub's null filter for cropping and Donald Graft's SmartResize. However, those shouldn't do anything to the display of the input stream. THey function perfectly well with earlier ffdshow builds so my assumption is they are a non-issue. VirtualDub build is 1.8.8 on Windows Tablet SP3 with all the updates.

[leeperry]

ok! from what this link says(madshi confirmed it) : http://archives.devshed.com/forums/bsd-93/host-to-network-byte-order-for-floats-and-float32-t-125578.html

at line 48 of this component : http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/audioFilters/winamp2/Twinamp2.h?view=markup&pathrev=2917

"int16_t" needs to be replaced by "float"...can I submit it? where? how? please help http://forum-images.hardware.fr/images/perso/emanu.gif

[haruhiko_yamagata]

http://www.corrugatedmachines.com/RE%204-Corner%20Tray%20v2%20HuffYUV%2030fps%20short.rar

That is one of the files that gives me headaches. It's about 30 Megabytes.

The only filters used were VirtualDub's null filter for cropping and Donald Graft's SmartResize. However, those shouldn't do anything to the display of the input stream. THey function perfectly well with earlier ffdshow builds so my assumption is they are a non-issue. VirtualDub build is 1.8.8 on Windows Tablet SP3 with all the updates.
It works without filters. I need more information about the settings of the filters.

[Jeremy Duncan]

I installed windows 7 RC. The 64 bit ultimate one. I then installed newest 64 bit media player classic and 64 bit ffdshow. I then resized with ffdshow and the player did not crash. I then used avisynth and the player crashed and gave me this dump file: link (http://www.megaupload.com/?d=B8J8UTH1)

I was using the 2.5.8 mt avisynth.dll version. Then uninstalled avisynth and reinstalled the official 2.5.8 and it still crashed.
I then installed the 32 bit versions: latest clsid mpc and generic ffdshow and even with both ffdshow resize and avisynth used it did not crash.

I'm posting this here because it was avisynth in ffdshow that caused the crash. That's why i'm not posting it in the software forum in the mpc thread. :)

Edit, The video it crashed on was my ntsc test disk, the progressive one. Look at my profile to see the test disk thread I made to get the vob.

[Shark007]

revision 2896 works ok (tested with 32bit compiles acquired from xvidvideo) but all recent builds since then cause an explorer crash when building thumbnails for MP4 files using Gabest's splitter from HC SVN builds. (OS: Vista 32bit)

EDIT: this issue was resolved by haruhiko_yamagata in Private Messaging
he exposed a rogue registry entry in hkcr\.mp4

[FredThompson]

It works without filters. I need more information about the settings of the filters.
http://www.CorrugatedMachines.com/Filters.zip

Smart Resize 1.1
RGB Equalizer 2.1 (I'd forgotten to mention that one)
null

Zip has the first 2 filters and the exprted filter chain.

I'll test again here.

If it works for you, maybe there is another filter that is getting in the way.

Edit: I've tested and still have the problem. It happens even with empty plugins and plugins32 directories. Build 2645 is the most recent which works properly. Tested with Athlon and XP Pro and also with Centrino Duo and Tablet PC.

[Sharktooth]

@Shark007: gabest MP4 splitter is pretty old and obsolete. use haali's media splitter.

[Jeremy Duncan]

I am using a avisynth script in ffdshow that needs more than 99 buffers.

Here is the script:

setmtmode(mode=2)
last=ffdshow_source()
super = MSuper(pel=2,hpad=4,vpad=4)
backward_vec = MAnalyse(super,chroma=false, blksize=16, isb = true, search=5, searchparam=1)
forward_vec = MAnalyse(super, chroma=false, blksize=16)
MFlowFps(super, backward_vec, forward_vec, num=0, mask=1, ml=49)
distributor()

Ädd ffdshow video source" is unchecked.

I have a ntsc dvd called "The bourne identity". And it has a sub root menu that asks me if I want to see the extended or regular version.

I use this sub root menu to identify how much the buffer need to be.
How I do this is if the buffer is high enough the sub root menu will show.
How I see if the buffer is high enough is if it is too low a buffer number I raise it by 2 and the sub root menu is not seen but the picture overlaying on it has moves a frame, I keep adding buffers by two and in so doing move the overlaying picture frame by frame until the sub root menu is visable.

In windows 7, if the buffer is too high the video has a paused frame every few seconds.

So I need to keep the buffer high enough I see the sub root menu and low enough the video is windows 7 doesn't pause every few seconds.

I have raised the buffers to 99, 99 and in so doing have almost gotten to the sub root menu but I need more buffers, over 100 to get there.

So I humbly ask Leak if he can kindly raise the buffers so they can go over 100.
And if he does not believe me that they need to be so high he can test the code I posted above on a dvd that has a sub root menu that asks if he wants to see the regular or extended version. :)

I'm almost got the buffers where I need them to be I just need over 100 buffers to get there and am stuck at 99 buffers.

[Kurtnoise]

@Shark007: gabest MP4 splitter is pretty old and obsolete. use haali's media splitter.
I wouln't say that...

[Mixer73]

@Shark007: gabest MP4 splitter is pretty old and obsolete. use haali's media splitter.

Personally I only use Gabest's splitter on my system because of the bugs in Haali media splitter.

Means I can't play MKV files but I'd rather be able to watch and then delete MP4 files from MCE which is something you cannot do with Haali Splitter installed.

[Sharktooth]

I wouln't say that...
latest gabest mp4 splitter is from 2007... unless i missed something...

@Mixer73: just disable shell integration while installing haali splitter. if it's not enough kill the buggy windows preview with regsvr32 /u shmedia.dll

[Kurtnoise]

latest gabest mp4 splitter is from 2007... unless i missed something...
http://www.xvidvideo.ru/component/option,com_docman/task,cat_view/Itemid,11/gid,19/orderby,dmdate_published/

[clsid]

I use both MP4 splitters. Haali by default. Gabest as backup when Haali fails. Its just a matter of toggling the preferred splitter through a registry setting.

[Mixer73]

@Mixer73: just disable shell integration while installing haali splitter. if it's not enough kill the buggy windows preview with regsvr32 /u shmedia.dll

Disabling it in the installer does NOT work if you have previously installed it.

I will try unregistering the file as you say, maybe I can use it finally...

[haruhiko_yamagata]

revision 2896 works ok (tested with 32bit compiles acquired from xvidvideo) but all recent builds since then cause an explorer crash when building thumbnails for MP4 files using Gabest's splitter from HC SVN builds. (OS: Vista 32bit)
I can't reproduce the crash, but the video doesn't play with gabest's mp4 splitter. I'll take a look.

[haruhiko_yamagata]

Tried rev2940, crackling sound in LPCM after rewind still presents. Can't say the last good revision, because LPCM decoder started to work (for me) only recently. Beta6 doesn't work because of missing submediatype.
I would like to try it again after MPC-HC has begun to work. In that case, we can exclude the possibility of a bug in the splitter.

[Kurtnoise]

I use both MP4 splitters. Haali by default. Gabest as backup when Haali fails. Its just a matter of toggling the preferred splitter through a registry setting.
Also, the Haali splitter is not yet updated for compliant AC3-in-MP4 streams.

[Leak]

I am using a avisynth script in ffdshow that needs more than 99 buffers.
Sorry, but I call bull**** on that one.

So I humbly ask Leak if he can kindly raise the buffers so they can go over 100.
No way in hell. If your script needs more than a second of video buffered in both directions please fix your script, you're doing it wrong.

np: Sleepy Town Manufacture - O Nei (Inspired By You)

[Sharktooth]

@kurtnoise: doh! i thought the gabest mp4 splitter from project guliverkli...

[RadScorpion]

I'm not sure if this is a known bug, but Gabest MP4 splitter ignores values in "EDTS\ELST" boxes and if both streams don't start with the same time stamp values, it introduces desync. Haali works fine with this.

[Jeremy Duncan]

Sorry, but I call bull**** on that one.


No way in hell. If your script needs more than a second of video buffered in both directions please fix your script, you're doing it wrong.

np: Sleepy Town Manufacture - O Nei (Inspired By You)

Thank you for the reply. I thought it was a bit weird too that I would need more than 99 buffers. The script is not mine but somebody elses in a different forum. I was trying it out.

[STaRGaZeR]

haruhiko, after testing your lastest fixes to ffmpeg-mt I've noticed that each revision has fixed various blocking issues with different sources. However the changes in rev2941 produce some garbage when seeking the Blu-ray MSG00 ED.mkv comes from. Your code from rev2939 works perfectly fine.

[haruhiko_yamagata]

revision 2896 works ok (tested with 32bit compiles acquired from xvidvideo) but all recent builds since then cause an explorer crash when building thumbnails for MP4 files using Gabest's splitter from HC SVN builds. (OS: Vista 32bit)

After all, I can't reproduce. It was old MPC-HC EVR-custom with interlaced flag output (double clicking mp4 happened to open that quite old MPC-HC on my Vista). The latest MPC-HC works just fine.
If you can find a file that makes ffdshow crash, please send it to me.

// EDIT: By the way, how can I create thumbnails for explorer in Vista? I have tried some registry hacks which didn't work. I can create thumbnails for FLV, but not for MP4. Thumbnails are created for MKV, but they are all white.
Windows Registry Editor Version 5.00

[HKEY_CLASSES_ROOT\.mp4]
@="mp4file"

[HKEY_CLASSES_ROOT\.mp4\ShellEx]

[HKEY_CLASSES_ROOT\.mp4\ShellEx\{BB2E617C-0920-11d1-9A0B-00C04FC2D6C1}]
@="{c5a40261-cd64-4ccf-84cb-c394da41d590} "

[Kurtnoise]

By the way, how can I create thumbnails for explorer in Vista? I have tried some registry hacks which didn't work. I can create thumbnails for FLV, but not for MP4. Thumbnails are created for MKV, but they are all white.
try to clean up the thumbnails cache (http://a8t8.spaces.live.com/blog/cns!2518DD508BB713E8!265.entry)...

[haruhiko_yamagata]

try to clean up the thumbnails cache (http://a8t8.spaces.live.com/blog/cns!2518DD508BB713E8!265.entry)...
Thank you. I have changed the value to {327669A0-59A7-4be9-B99E-1C9F3A57611A}. It started to created thumbnails. Though they are all white, I can't reproduce crashes.

[haruhiko_yamagata]

haruhiko, after testing your lastest fixes to ffmpeg-mt I've noticed that each revision has fixed various blocking issues with different sources. However the changes in rev2941 produce some garbage when seeking the Blu-ray MSG00 ED.mkv comes from. Your code from rev2939 works perfectly fine.I can't reproduce.
To which frame should I seek? How many threads are used?

[clsid]

I'm not sure if this is a known bug, but Gabest MP4 splitter ignores values in "EDTS\ELST" boxes and if both streams don't start with the same time stamp values, it introduces desync. Haali works fine with this.
I don't think that is a known bug. Please report it to Casimir666. Then it get hopefully be fixed soon.

[haruhiko_yamagata]

Edit: I've tested and still have the problem. It happens even with empty plugins and plugins32 directories. Build 2645 is the most recent which works properly. Tested with Athlon and XP Pro and also with Centrino Duo and Tablet PC.
Does this mean image filters are not related at all? Is it CPU specific?

[STaRGaZeR]

I can't reproduce.
To which frame should I seek? How many threads are used?

I can't reproduce in that short sample, but on the entire file in random locations. 4 threads are used, but it happens also with only 1 thread. As I say rev2939 works fine, isn't rev2941 a better way of doing the same?

[travolter]

Its possible Anaglyph filter into ffdshow? (remember... the red&cyan glasses!!)

http://www.stereoscopy.com/faq/anaglyph.jpg

Watch videos in 3D!!!

[FredThompson]

Does this mean image filters are not related at all? Is it CPU specific?I have the problem with no filters present. I tested on an Athlon running Windows XP Pro and on a Centrino Duo running Tablet XP. The error happens with both ffdshow_rev2946_20090515_clsid and ffdshow_rev2946_20090515_xxl releases.

I will have access to an Athon 64 running Windows XP Pro tomorrow and will test on it as well.

Are the HuffYUV routines the exact same for build 2645 and whatever revision was next? Could it be a compiler issue?

[lych_necross]

Its possible Anaglyph filter into ffdshow? (remember... the red&cyan glasses!!)
Watch videos in 3D!!!

Probably via an avisynth plugin, but I don't know if any anaglyph filters exist in avisynth.

[Jeremy Duncan]

I am using a avisynth script in ffdshow that needs more than 99 buffers.

Here is the script:

setmtmode(mode=2)
last=ffdshow_source()
super = MSuper(pel=2,hpad=4,vpad=4)
backward_vec = MAnalyse(super,chroma=false, blksize=16, isb = true, search=5, searchparam=1)
forward_vec = MAnalyse(super, chroma=false, blksize=16)
MFlowFps(super, backward_vec, forward_vec, num=0, mask=1, ml=49)
distributor()

Ädd ffdshow video source" is unchecked.

I have a ntsc dvd called "The bourne identity". And it has a sub root menu that asks me if I want to see the extended or regular version.

I use this sub root menu to identify how much the buffer need to be.
How I do this is if the buffer is high enough the sub root menu will show.
How I see if the buffer is high enough is if it is too low a buffer number I raise it by 2 and the sub root menu is not seen but the picture overlaying on it has moves a frame, I keep adding buffers by two and in so doing move the overlaying picture frame by frame until the sub root menu is visable.

In windows 7, if the buffer is too high the video has a paused frame every few seconds.

So I need to keep the buffer high enough I see the sub root menu and low enough the video is windows 7 doesn't pause every few seconds.

I have raised the buffers to 99, 99 and in so doing have almost gotten to the sub root menu but I need more buffers, over 100 to get there.

So I humbly ask Leak if he can kindly raise the buffers so they can go over 100.
And if he does not believe me that they need to be so high he can test the code I posted above on a dvd that has a sub root menu that asks if he wants to see the regular or extended version. :)

I'm almost got the buffers where I need them to be I just need over 100 buffers to get there and am stuck at 99 buffers.

After telling the person who gave me the code in the quote above. He said the code was not wrong.
So I thought about it and decided to try the mt method and frame doubling method in the mvtools readme.
This would guarantee that neither the use of mt or the frame doubling method was done wrong with bad code.

So I will now post the codes I tried. The first code I will post is the exact code the person asked me to try:

#Settings
tc=4 #general thread count (try 2 or some other until it fits)
bs=16 #general block size (8 for nicer but slower)
ch=false #chroma correction (try true for nicer)
ms=0 #mask filtering (0-2)
sr=3 #search radius (3-4)
sp=1 #search param (1-2)
rr=60 #refresh rate of your screen
#EndSettings
#
SetMTmode(2, tc)
last=ffdshow_source()
super=MSuper(pel=1, hpad=bs, vpad=bs, chroma=ch)
backward_vec=MAnalyse(super, blksize=bs, chroma=ch,\
search=sr, searchparam=sp, isb = true)
forward_vec=MAnalyse(super, blksize=bs, chroma=ch,\
search=sr, searchparam=sp)
MFlowFps(super, backward_vec, forward_vec, num=0, mask=ms)
ConvertFps(rr) #remove this line for more speed
distributor()

The next two codes are from the mvtools readme. Note that the frame doubler did not use mt in the readme but I followed the readme on how to and did so in the two codes below. All of the two code from the codes below is from the mvtools readme:

last=ffdshow_Source()
MT("""
super = MSuper(pel=1,hpad=16,vpad=16)
backward_vec = MAnalyse(super, blksize=16, isb = true, chroma=false, searchparam=1)
forward_vec = MAnalyse(super, blksize=16, isb = false, chroma=false, searchparam=1)
MFlowFps(super, backward_vec, forward_vec, num=2*FramerateNumerator(last), \
den=FramerateDenominator(last), mask=0)
""",2)


setmtmode(5)
last=ffdshow_Source()
setmtmode(2)
super = MSuper(pel=1,hpad=16,vpad=16)
backward_vec = MAnalyse(super, blksize=16, isb = true, chroma=false, searchparam=1)
forward_vec = MAnalyse(super, blksize=16, isb = false, chroma=false, searchparam=1)
MFlowFps(super, backward_vec, forward_vec, num=2*FramerateNumerator(last), \
den=FramerateDenominator(last), mask=0)

I will post the sample of the movie, the root menu that is having the problem.
Use mpc and open file and open the top file from the list.
sample (http://www.mediafire.com/?mun4hxxmcny)
On the sample rrot menu click Play and it will take you to the sub root menu.

I will post the software I'm using now:

FFDshow. Updated april 24, 2009 (http://sourceforge.net/project/downloading.php?group_id=173941&filename=ffdshow_rev2940_20090506_clsid.exe&a=75398883)
media player classic. updated december 31, 2008 (http://downloads.sourceforge.net/guliverkli2/mplayerc_20081210.zip?modtime=1229004652&big_mirror=0)
Avisynth. march 10, 09 (http://downloads.sourceforge.net/avisynth2/Avisynth_258.exe?use_mirror=voxel)
mvtools (http://www.megaupload.com/?d=5RQK6DR2)
MT (http://www.megaupload.com/?d=V6HRTOBX)


And here is my software settings:

________ advanced ffdshow: framerate doubler ________

Codecs tab
Set Mpeg2 to Libmpeg2, and check "DVD decoding".
Set Avisynth to Avisynth, Raw video to All supported

Subtitles tab, Unchecked
Uncheck "Decode closed captions"
Uncheck "Accept embedded subs"
Uncheck "Accept SSA, ASS, ASS2 Subtitle (experimental)
Vobsub subpage, uncheck Vobsub Enable.

Avisynth tab checked
YV12 checked,
Add FFdshow Video source is UNchecked,
3:2 Pulldown box: Ignore Pulldown checked,
check Buffer back/Ahead: 9, 9

# add one of the codes from above in this place

Resize & aspect tab checked
Multiply by:
2.743 (for 1920x1080 16:9 aspect ratio)
Process Pixel aspect ratio internally checked
No aspect ratio correction checked
spline
Luma gaussian blur: 0.50
Luma Sharpen: 0.00
Accurate rounding checked

Queue & Output tab
Queue output samples unchecked
Very unstable. May cause player to crash.
No box is checked.

Output tab
rgb32 checked
high quality yv12 to rgb conversion checked

rgb conversion tab
ycbcr: itu-bt 601
contrast: standard
Output levels: Tv/Projector

________ Media Player Classic ________

Click the View button. Check only the top three in the drop down list.
Click options.

Click the output tab, and set Directshow output to vrm9 renderless.
Directshow Audio: default

Click the External Filters Tab.
Click the Add Filter Button.
Double click the filters you want. You'll need;
FFdshow Audio decoder
FFdshow video decoder

[lych_necross]

@ Jeremy Duncan
That script makes no sense (99+ buffers?!?). Using MFlowFPS to double the framerate, then converting the framerate again using ConvertFPS is a horrible idea. As a general rule, don't convert the fps more than once if possible. FFDShow has a nice built-in framerate doubler (fast too). Also, using MT( or SetMTMode( with MFlowFPS probably won't help much. Just use the latest version of mvtools 2 and the example code for realtime framerate doubling and forget the MT stuff.

[haruhiko_yamagata]

I have the problem with no filters present. I tested on an Athlon running Windows XP Pro and on a Centrino Duo running Tablet XP. The error happens with both ffdshow_rev2946_20090515_clsid and ffdshow_rev2946_20090515_xxl releases.

I will have access to an Athon 64 running Windows XP Pro tomorrow and will test on it as well.

Are the HuffYUV routines the exact same for build 2645 and whatever revision was next? Could it be a compiler issue?
If it is related to rev 2666, MMXEXT and 3DNOW! may matter.
Could you uncheck them?

[Jeremy Duncan]

@ Jeremy Duncan


That script makes no sense (99+ buffers?!?).

Yes, it should not be that much. It must be a problem with ffdshow.


Using MFlowFPS to double the framerate, then converting the framerate again using ConvertFPS is a horrible idea. As a general rule, don't convert the fps more than once if possible.
I showed in my previous post I used the mvtools readme code which only increases the frame rate once.


FFDShow has a nice built-in framerate doubler (fast too). No, thank you.


Also, using MT( or SetMTMode( with MFlowFPS probably won't help much. Just use the latest version of mvtools 2 and the example code for realtime framerate doubling and forget the MT stuff. Actuall it helps activate the hyperthreading cores in the I7 cpu's I was told.

[haruhiko_yamagata]

I can't reproduce in that short sample, but on the entire file in random locations. 4 threads are used, but it happens also with only 1 thread. As I say rev2939 works fine, isn't rev2941 a better way of doing the same?
Rev 2939 is theoretically wrong, 2941 should be better...

[nuhi]

@ Jeremy Duncan
That script makes no sense (99+ buffers?!?). Using MFlowFPS to double the framerate, then converting the framerate again using ConvertFPS is a horrible idea. As a general rule, don't convert the fps more than once if possible. FFDShow has a nice built-in framerate doubler (fast too). Also, using MT( or SetMTMode( with MFlowFPS probably won't help much. Just use the latest version of mvtools 2 and the example code for realtime framerate doubling and forget the MT stuff.
Ok I disabled HT on my i7 just to prove the point of MT being necessary for heavy processing.
Usually there would be 8 cores listed in the Task Manager, now it is 4.

Here is the comparison between MFlowFPS default fast example for frame doubling.
Without SetMTmode:
http://img194.imageshack.us/img194/4377/nomt.jpg
and with SetMTmode/distributor:
http://img41.imageshack.us/img41/746/56850135.jpg

Before you say tell that to MVTools author we are talking about general benefits, MT plugin can be wrapped around almost anything to split into more threads. And upcoming Avisynth 2.6 (if it ever gets finished) has these functions in it so you will be seeing more of it :)

Anyway just trying to help, thank you guys.

[haruhiko_yamagata]

I can't reproduce in that short sample, but on the entire file in random locations. 4 threads are used, but it happens also with only 1 thread. As I say rev2939 works fine, isn't rev2941 a better way of doing the same?
If it is not a race condition, you may be able to reproduce by pausing at a certain frame (enable OSD "current frame"), cltl-G to seek, and play. If it is possible, please cut that part.
It is not likely a race condition, but DirectShow has bug-prone thread issue.
I checked the code again, but couldn't find any suspicious part. Please try again with rev 2939 to make sure you are not too lucky with that revision.

[STaRGaZeR]

If it is not a race condition, you may be able to reproduce by pausing at a certain frame (enable OSD "current frame"), cltl-G to seek, and play. If it is possible, please cut that part.
It is not likely a race condition, but DirectShow has bug-prone thread issue.
I checked the code again, but couldn't find any suspicious part. Please try again with rev 2939 to make sure you are not too lucky with that revision.

OK, will do. In the meantime, some screenshots:

http://thumbnails12.imagebam.com/3604/39b52d36031425.gif (http://www.imagebam.com/image/39b52d36031425) http://thumbnails12.imagebam.com/3604/98e8b436031426.gif (http://www.imagebam.com/image/98e8b436031426)

The garbage is part of the frame that was being displayed before the seek. That's why is difficult to see in small samples, the two frames may be almost identical. Also it only last 1 frame.

[lych_necross]

@Jeremy Duncan
From the original code in question:

MFlowFps(super, backward_vec, forward_vec, num=0, mask=ms)
ConvertFps(rr) #remove this line for more speed

Two frame rate conversions: MFlowFPS with num=0 doubles framerate. ConvertFPS converts the framerate again.

The code from your guide(s) only convert's the framerate once, yes. My point was that the script you tried originally from another forum, is illogical. If you are increasing the framerate to achieve smoother playback, it won't help much (it will only mask playback problems, not eliminate them). Reducing jitter and dropped frames is far better than doubling fps.

@nuhi
I said, using MT or SetMTMode with MFlowFPS probably won't help much. That doesn't mean that won't (on my system, it doesn't help much, yours may differ).

[Jeremy Duncan]

@Jeremy Duncan
From the original code in question:

Two frame rate conversions: MFlowFPS with num=0 doubles framerate. ConvertFPS converts the framerate again.

The code from your guide(s) only convert's the framerate once, yes. My point was that the script you tried originally from another forum, is illogical.

I will post a few replies I got from a different thread on the subject.
The point the quotes make is if your video is 23.976 fps then use use mvtools to double the framerate and then use changefps to change the fps from 48 to 60.

I meant that if you first apply ChangeFPS(30) to 23.976 fps video and then double framerate with MVTools, interpolation will insert 3d duplicated frame between already existing repeated frames. So it would be better instead of ChangeFPS(30) before MFlowFPS to use ChangeFPS(60) after it like this:

#23.976
super = MSuper(pel=1,hpad=16,vpad=16)
backward_vec = MAnalyse(super, blksize=16, isb = true, chroma=false, searchparam=1)
forward_vec = MAnalyse(super, blksize=16, isb = false, chroma=false, searchparam=1)
MFlowFps(super, backward_vec, forward_vec, num=2*FramerateNumerator(last), \
den=FramerateDenominator(last), mask=0)
ChangeFPS(60)
#then the inserted frames will not form longer chains


The usual application of MFlowFPS to get 60 from 23.976 looks like this:
#23.976
super = MSuper(pel=2)
backward_vec = MAnalyse(super, isb = true)
forward_vec = MAnalyse(super, isb = false)
MFlowFps(super, backward_vec, forward_vec, num=60, den=1, ml=100)

It works not as fast as with script for double framerate (probably you wanted realtime processing), but this way you could completely avoid duplicated frames. It changes all the frames, not just inserts new ones.



Before any framerate conversion, you need to do inverse telecining (restore film to 23.976 progressive). Your sample of studio logo shows 23.976 pics/s (29.97 fps) in GSpot. This means soft telecining (encoded as 23.976 mpeg with display flags added for 29.97). If main film shows the same in GSpot, you want to import it into AviSynth as 23.976 (e.g. with DGIndex in 'force film' mode), otherwise you'll have 2 interlaced frames in every 5-frame cycle. If the film VOBs show 29.97/29.97 pics/frames per second, you need inverse telecining in AviSynth e.g. with Decomb plugin.
Don't rely on fps numbers determined by DirectShow-based import filters, use AssumeFPS() where necessary.

link (http://forum.doom9.org/showthread.php?p=1261569#post1261569)

So, in fact it does make sense to use both mvtools to double the frame rate as well as changefps.

[STaRGaZeR]

If it is not a race condition, you may be able to reproduce by pausing at a certain frame (enable OSD "current frame"), cltl-G to seek, and play. If it is possible, please cut that part.
It is not likely a race condition, but DirectShow has bug-prone thread issue.
I checked the code again, but couldn't find any suspicious part. Please try again with rev 2939 to make sure you are not too lucky with that revision.

It's impossible to reproduce it in the same frame more than 1 time. When you pause, seek again or restart the player, and then go back to the frame that had garbage before, it's fine. All I can do is upload a big sample so you can test it yourself. After retesting: 2939 is unaffected, problem starts with 2941 and upwards. For testing I'm using my own builds, clsid's and xvidvideo.ru's, all show the same issue. Other Blu-ray's are affected too.

[Jeremy Duncan]

http://thumbnails12.imagebam.com/3604/98e8b436031426.gif (http://www.imagebam.com/image/98e8b436031426)

I see that corruption in the root menu when I use ffdshow avisynth and the buffer is too low.

So whatever your doing it's using buffers i think, and it's using buffers incorrectly.
Since he said 2939 works fine and 2940 only updates ffmpeg. it seems ffmpeg changed it's use of buffers. I can't find the changelog to confirm this though.

[clsid]

It's caused by 2941, which fixes a different blocking problem.
http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/ffmpeg-mt/libavcodec/h264.c?r1=2941&r2=2940

[lych_necross]

I will post a few replies I got from a different thread on the subject.
The point the quotes make is if your video is 23.976 fps then use use mvtools to double the framerate and then use changefps to change the fps from 48 to 60.
link (http://forum.doom9.org/showthread.php?p=1261569#post1261569)

So, in fact it does make sense to use both mvtools to double the frame rate as well as changefps.
Actually, from that link, it still does not make sense to use both, but I don't feel like arguing right now. If your happy with using both, fine; however, I still believe that the problem is your script and not ffdshow.

[leeperry]

Its possible Anaglyph filter into ffdshow? (remember... the red&cyan glasses!!)

http://www.stereoscopy.com/faq/anaglyph.jpg

Watch videos in 3D!!!
http://forum.doom9.org/showthread.php?t=143855

[Jeremy Duncan]

Actually, from that link, it still does not make sense to use both, but I don't feel like arguing right now. If your happy with using both, fine; however, I still believe that the problem is your script and not ffdshow.

Not every script I posted uses changefps. The ones that don't still have the problem with the sample I linked too.
No number of buffers fixes the problem, even for the scripts that don't use changefps but still use setmtmode.

Using setmtmode activated the ht cores on the I7 I was told. so the problem I showed and gave the sample to reproduce the problem should be fixed.

[Leak]

I see that corruption in the root menu when I use ffdshow avisynth and the buffer is too low.
Would you please stop spouting garbage already?

The above is clearly a decoding error in ffmpeg-mt and doesn't concern AviSynth at all. And if playback somehow stalls in the menu but the AviSynth filter produces correct output during the movie part it's not "using buffers incorrectly" but there's obviously some other part of the chain that needs to be fixed.

np: Secede - Vega Libre (Vega Libre)

[Jeremy Duncan]

Leak,

With all due respect.
I can reproduce the blocking problem I quoted by adjusting the buffers to do this.
I read clsid reply in his post on rev 2941link (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/ffmpeg-mt/libavcodec/h264.c?r1=2941&r2=2940) and this rev uses buffers differently than the one it replaced.

I know full well the problem has nothing to do with avisynth.

[lych_necross]

:D, Leak. I agree. ffmpeg-mt is still experimental and buggy (but getting better).

[Jeremy Duncan]

lych sounds like leak in my language. I wonder what it means in your language.

Anyway. to Add to this linked post; link (http://forum.doom9.org/showthread.php?p=1286413#post1286413). I tested with the 2.4.2, the newest mvtools2, and i used only the official avisynth 2.5.8 with no mt avisynth.dll added. I tried with mpc-hc and clsid's newest build. I tried with this code in avisynth ffdshow using the sample root menu I uploaded before:

super = MSuper(pel=2,hpad=4,vpad=4)
backward_vec = MAnalyse(super,chroma=false, blksize=16, isb = true, search=5, searchparam=1)
forward_vec = MAnalyse(super, chroma=false, blksize=16)
MFlowFps(super, backward_vec, forward_vec, num=0, mask=1, ml=49)

I could not even see the sub root menu where it asks me which version i want to see.

[haruhiko_yamagata]

The garbage is part of the frame that was being displayed before the seek. That's why is difficult to see in small samples, the two frames may be almost identical. Also it only last 1 frame.
From todo.txt in the git (http://gitorious.org/~astrange/ffmpeg/ffmpeg-mt),
Todo

Bug fixes:
h264:
- The frame num gap compensation code calls ff_report_*_progress with INT_MAX on every frame; it should only do this for every frame but the last.Maybe this is concerned. Rev 2941 is not likely to cause that kind of problem. It may alter the timing a bit.

[swalker]

Does the ffdshow (r2940) crop -> autocrop filter work? Only manually cropping works; AviSynth's autocrop plugin is broken as well on a Mobility Radeon X300 running Catalyst 9.3 drivers.

[clsid]

autocrop has been broken for a long time

[STaRGaZeR]

From todo.txt in the git (http://gitorious.org/~astrange/ffmpeg/ffmpeg-mt),
Maybe this is concerned. Rev 2941 is not likely to cause that kind of problem. It may alter the timing a bit.

Since 2939 works, can you revert it until is fixed?

[haruhiko_yamagata]

@STaRGaZeR
Could you turn on OSD "Frame type"? Is the first frame after seeking a B frame and garbled?
In that case, it's a bug in ffdshow custom part.

[tetsuo55]

There appears to be a bug with the output framerate when playing back DVD's

All samples act the same, output is always fixed at 25fps

Here is a screenshot:
http://img33.imageshack.us/img33/5011/incorrectframerate.th.png (http://img33.imageshack.us/my.php?image=incorrectframerate.png)

And here is the pin out information for that file (same problem with every renderer)

Filter : madVR - CLSID : {E1A8B82A-32CE-4B0D-BE0D-AA68C772E423}

- Connected to:

CLSID: {04FE9017-F873-410E-871E-AB91661A4EF7}
Filter: ffdshow Video Decoder
Pin: Out

- Connection media type:

Video: YV12 720x480 (4:3) 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_YV12 {32315659-0000-0010-8000-00AA00389B71}
formattype: FORMAT_VideoInfo2 {F72A76A0-EB0A-11D0-ACE4-0000C0CC16BA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 518400
cbFormat: 112

VIDEOINFOHEADER:
rcSource: (0,0)-(720,480)
rcTarget: (0,0)-(720,480)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 4
dwPictAspectRatioY: 3
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 720
biHeight: 480
biPlanes: 3
biBitCount: 12
biCompression: YV12
biSizeImage: 518400
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 d0 02 00 00 e0 01 00 00 ........Ð...à...
0010: 00 00 00 00 00 00 00 00 d0 02 00 00 e0 01 00 00 ........Ð...à...
0020: 00 00 00 00 00 00 00 00 80 1a 06 00 00 00 00 00 ........€.......
0030: 00 00 00 00 00 00 00 00 04 00 00 00 03 00 00 00 ................
0040: 00 00 00 00 00 00 00 00 28 00 00 00 d0 02 00 00 ........(...Ð...
0050: e0 01 00 00 03 00 0c 00 59 56 31 32 00 e9 07 00 à.......YV12.é..
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................

Also it appears that DTS over Spdif is not working for me anymore all of suddon (it used to work, i did not touch any of the ffdshow settings) now all i get is silence or pops.
AC3 is working as expected.

I am streaming over HDMI
Build im currently using is 2940

[Spec-Chum]

I've heard it's implemented, but can't find it...I'm using build 2951.

Am I going mad?

[clsid]

The new option is located on the "RGB conversion" page in the config.

[BatKnight]

The new option is located on the "RGB conversion" page in the config.
How does it work? Is it like the DeBand filter?
Is it supposed to give better picture quality when checked or the other way around?

Bat

[Spec-Chum]

The new option is located on the "RGB conversion" page in the config.

I really am going mad, it's not there...I'm using latest build from XvidVideo.RU. Using Core2 so SSE2 is available. Can anyone confirm my sanity?

[mark0077]

I really am going mad, it's not there...I'm using latest build from XvidVideo.RU. Using Core2 so SSE2 is available. Can anyone confirm my sanity?

Its there for me in 2951, in the "Method" section of "RGB Conversion" tab. Thanks for this addition guys, hopefully ffdshow can get upto the standard of madVR for resizing quality and colorspace conversion.

[Spec-Chum]

I really am going mad, it's not there...I'm using latest build from XvidVideo.RU. Using Core2 so SSE2 is available. Can anyone confirm my sanity?

LOL, got it, I didn't restart after updating (explorer.exe is now in the don't use section)...n00b!

[mark0077]

haruhiko, apologies for 3 bugs I posted a few posts ago, I found all reproducable without ffdshow in the chain.

Glad to see dithering in rgb conversion, just wondering could I ask have you thought about adding the workaround for DVD's when resizing. You may remember the problem I and others have had, where some DVD's show the wrong aspect ratio when passing through ffdshow's resize?

Also just while on the topic of resize, I found a bug (not a critical one) within "Border" section of Resize. After setting one of the dividing borders, for example "Horizontal" to Left 100, Right 0, and hitting apply. Then dragging the slider for "Vertical" resets the "Horizontal" slider to centered on 50:50 again. Just strange. I want dividing borders set to right and bottom because of the way the LED backlight works on this LCD. Keeping black bars to one side or fully to top or bottom can give me blacker blacks.

Thanks.

[STaRGaZeR]

@STaRGaZeR
Could you turn on OSD "Frame type"? Is the first frame after seeking a B frame and garbled?
In that case, it's a bug in ffdshow custom part.

All of them are B frames according to the OSD:

http://thumbnails14.imagebam.com/3617/e73f1336163833.gif (http://www.imagebam.com/image/e73f1336163833) http://thumbnails15.imagebam.com/3617/c45ac636163835.gif (http://www.imagebam.com/image/c45ac636163835) http://thumbnails13.imagebam.com/3617/d3a18536163837.gif (http://www.imagebam.com/image/d3a18536163837)

HD-DVD source BTW.

[swalker]

autocrop has been broken for a long time

Can it (I'd suggest along with anything else that is broken/severely malfunctioning if nobody is working on it presently or recently) get disabled or removed then since I doubt it's a priority with the limited amount of developers ffdshow has at the moment?

Relnoting what is or isn't functional would be nice, well if ffdshow actually had release notes, I suppose a wiki page and/or changelog item would have to suffice.

[haruhiko_yamagata]

How does it work? Is it like the DeBand filter?
Is it supposed to give better picture quality when checked or the other way around?

Bat

The calculation of YCbCr->RGB conversion is done in 11bit or higher, and the output is dithered to 8bit.
In the end of calculation, RGB values are 12bit for each. Dithering is done by adding 4bit single dice random noise (different values for R,G,B). Finally, the values are shifted to the right by 4 bits.
Thus 100.8125 is mapped to 101 in 81.25% probability and to 100 in 18.75%.

The objective is the same as deband, but the method is different. I actually don't know how deband works though.
The picture quality is better with dithering checked.

Thanks to madTestPatternSource (http://forum.doom9.org/showthread.php?t=146203) by madshi, the image comparison is easy.

grayramp.ytp none dithered
http://ffdshow-tryout.sourceforge.net/doom9/gray_none_dithered.png

grayramp.ytp dithered
http://ffdshow-tryout.sourceforge.net/doom9/gray_dithered.png

colors.ytp none dithered
http://ffdshow-tryout.sourceforge.net/doom9/color_none_dithered.png

colors.ytp dithered
http://ffdshow-tryout.sourceforge.net/doom9/color_dithered.png

Hope you can find any real world videos that you can perceive the visual change.

[haruhiko_yamagata]

haruhiko, apologies for 3 bugs I posted a few posts ago, I found all reproducable without ffdshow in the chain.no problem.

Glad to see dithering in rgb conversion, just wondering could I ask have you thought about adding the workaround for DVD's when resizing. You may remember the problem I and others have had, where some DVD's show the wrong aspect ratio when passing through ffdshow's resize?
Did you try this (http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example)?

Also just while on the topic of resize, I found a bug (not a critical one) within "Border" section of Resize. After setting one of the dividing borders, for example "Horizontal" to Left 100, Right 0, and hitting apply. Then dragging the slider for "Vertical" resets the "Horizontal" slider to centered on 50:50 again. Just strange. I want dividing borders set to right and bottom because of the way the LED backlight works on this LCD. Keeping black bars to one side or fully to top or bottom can give me blacker blacks.

Thanks.
I can't reproduce. The sliders work for me.

[mark0077]

no problem.

Did you try this (http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example)?

I can't reproduce. The sliders work for me.

Hi Haruhiko.

Disabling mpc-hc's own aspect ratio correction seemed to do the trick. I don't think any combination of these settings should break the final aspect ratio though if all of the flags are sent and interpreted correctly.

Do you think mpc-hc doesn't interpret the flags sent by ffdshow correctly, if so I can make a bug report. Or does ffdshow not yet send the flags correctly. I would love to see the resize feature work without any manual changing of settings like this as it will only add to the manual settings to configure when setting up a machine.

Thanks for the link.

[haruhiko_yamagata]

MPC-HC expects the pixel aspect ratio in the original DVD is not changed in ffdshow. However, resize with "Process pixel aspect ratio internally" changes PAR. It's the spec of MPC-HC and ffdshow.
You could submit a report to MPC-HC. The solution would be an new API which both should implement.

[mark0077]

Ah I understand. Its like both are trying to do the same fix and neither knows whether the other has already performed the fix or not.?

I have posted a bug in mpc-hc bug tracker. Hopefully some discussion will come up with a solution.
http://sourceforge.net/tracker/?func=detail&aid=2793920&group_id=170561&atid=854651

[ipanema]

I'm trying to connect from my own source filter to the ffdshow video decoder which has H.264/AVC enabled, but I can't seem to discover the correct media type.

The following subtype which I've not found a symbolic name for

8D2D71CB-243F-45E3-B2D8-5FD7967EC09B

seems to work well with other decoders, but EnumMatchingFilters() can't even find the ffdshow video decoder, so I guess ffdshow does not support this subtype. :(

This Microsoft page:

http://msdn.microsoft.com/en-us/library/dd757808(VS.85).aspx

says that for a "H.264 Bitstream without Start Codes" you should use MEDIASUBTYPE_AVC1. I've noticed that when Media Player Classic uses the ffdshow video decoder to play m2ts files, the input media type of the ffdshow video decoder is MEDIASUBTYPE_AVC1 and this is confirmed by the sequence-header part of the media type having no start codes.

I assume that to use MEDIASUBTYPE_AVC1 you would have to parse the whole bitstream, replacing start codes and shifting lots of data around in memory before it is sent to the ffdshow decoder. I would rather just send the data as-is (with start codes) which does work fine with other decoders.

What AVC media type (WITH start codes) works best with the ffdshow video decoder?

[leeperry]

has there been any recent change in the Avisynth filter? coz I'm getting BSOD after BSOD when I enable it :(

usually it simply stutters a bit and reboots the system! I can run 10 passes of IntelBurnTest and my system is not o/c.

it used to work fine, I just updated to the latest build from xvidvideo.ru, I'll try older builds...but anyway, when I disable the Avisynth filter everything works fine! using whatever KMP or MPC, HR or Overlay, Sonic 4.2 or AC3Filter, Reclock or DirectSound....I believe I've narrowed down the problem to the Avisynth filter(I use Avisynth-MT 2.57 on XP SP3)

[leeperry]

ok I'm running a 2 months old build, and everything seems to work just fine!

I can't even enable CUDA in CoreAVC(on top of the Avisynth filter) w/ the latest builds...I get a IRQL_NOT_LESS_OR_EQUAL stop msg right away.

sounds like there's been some changes in the internal buffering of ffdshow lately? anyway, w/ this old build I can get CoreAVC CUDA and Avisynth working apparently.

that's one of the numerous stop msg I got :

0x0000000A
~5
~1C
0x804DC2C5

does the adress of the instruction ring any bell? http://support.microsoft.com/kb/314063

my PC works fine(3 days of foobar in KS, 10 passes of IntelBurnTest, etc etc), just the very latest builds of ffdshow give crazy BSOD's :(

I'd say the problem arised in the <7 days builds.

[STaRGaZeR]

Haruhiko, corruption fixed. However I don't know why it doesn't work with 2941, when the code you reverted was like it's now.

[Inspector.Gadget]

I can't even enable CUDA in CoreAVC(on top of the Avisynth filter) w/ the latest builds...I get a IRQ_LESS_OR EQUAL stop msg right away.

I had this problem plus random lock-ups after updating to NVIDIA 185.85 drivers, by the way.

[Thunderbolt8]

in how far does yadif's frame duplication option use CPU and GPU sources (e.g. more CPU than GPU based?)

[FredThompson]

If it is related to rev 2666, MMXEXT and 3DNOW! may matter.
Could you uncheck them?I assume you mean the VirtualDub CPU functions? I turned all of those off and the problem still exists for build 2954.

[Leak]

in how far does yadif's frame duplication option use CPU and GPU sources (e.g. more CPU than GPU based?)
Considering that all ffdshow uses is the CPU - all CPU, no GPU?

(Unless you consider that the renderer will have to upload twice the amount of frames to the graphics card for display, of course...)

[swalker]

Given the following from the March 10, 2009 news on http://www.ffmpeg.org/.
# AAC with ADTS support and >2x the speed of FAAD! (no HE AAC support yet)
Should the libavcodec AAC decoding choice be uncommented in the installer script and be offered along with libfaad2, noting libavcodec's lack of HE AAC support?
# AC-3 that is faster than liba52 in 5.1, up to 2x faster in stereo and also supports E-AC-3! Hence liba52 is now obsolete. Should a default switch be made to using libavcodec for AC-3 decoding with eventual removal of liba52 if no problems are found?

Are there that many bugs/efficiency differences in each MPEG-1 audio decoder to warrant having a choice of three of them?

[tetsuo55]

I use libvacodec for everything, i have encountered no problems (any issue i find can be reproduced with the alternative codec)

I agree that we should think about moving out as many of the extra codecs as possible and focus on getting libvacodec as compatible as possible (most of the alternatives are no longer being updated)

[Thunderbolt8]

Considering that all ffdshow uses is the CPU - all CPU, no GPU?

(Unless you consider that the renderer will have to upload twice the amount of frames to the graphics card for display, of course...)is it optimized for multi core cpus?

[haruhiko_yamagata]

I'm trying to connect from my own source filter to the ffdshow video decoder which has H.264/AVC enabled, but I can't seem to discover the correct media type.

Please read
http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=1138
You could write to the thread for more discussion.
8D2D71CB-243F-45E3-B2D8-5FD7967EC09B
What does this subtype indicate?

I assume that to use MEDIASUBTYPE_AVC1 you would have to parse the whole bitstream, replacing start codes and shifting lots of data around in memory before it is sent to the ffdshow decoder. I would rather just send the data as-is (with start codes) which does work fine with other decoders.
Yes, MPC and haali's splitter does this, as older ffdshow couldn't do without it. Current version doesn't require full access unit parsing by the splitter.

[haruhiko_yamagata]

I assume you mean the VirtualDub CPU functions? I turned all of those off and the problem still exists for build 2954.
Nope, I mean CPU in ffdshow.
It is placed in "ffdshow (DirectShow) video decoder configuration", Info & CPU page.

[ipanema]

Thanks haruhiko.


What does this subtype 8D2D71CB-243F-45E3-B2D8-5FD7967EC09B indicate?


I don't really know. I found it mentioned a few times when searching for H.264 media types. There didn't seem to be consistent "name" for it but it does seem to work OK with other decoders. Perhaps it is just a duplication of MEDIASUBTYPE_H264.

Current version doesn't require full access unit parsing by the splitter.

That's fine. So if I use the H264 subtype instead of the AVC1 subtype then I shouldn't have to replace start codes with lengths or shift data around. :)

Anyway, I've managed to get the connection to the ffdshow video decoder to succeed now using either the above 8D2D... subtype or MEDIASUBTYPE_H264. But when I attempt to play I get no frames delivered for a while and then occasional decoded frames which look fine apart from a small amount of corruption.

I've concluded that the connection media type is OK, so it's likely that the way the data is being presented to the ffdshow video decoder is the problem.

As I was familiar with the Microsoft MPEG-2 Demultiplexer and it works OK with other H.264 decoders, that's what I am using to feed the ffdshow decoder. The Microsoft Demux doesn't look at the PES payload contents - it just seems to extract the video PES payload bytes from consective TS packets, concatenate them and output it in samples of 8192 bytes in length. It carries on doing this until it reaches the next frame and then starts sending a new series of 8192 byte samples. The first 8192 byte sample of a frame has a start and end stream time (but no media times) and Sync Point set; and subsequent samples have no timestamps at all and Sync Point is not set.

You mentioned in the other thread

If you use 'H264' as fourcc, ffdshow can parse access units if the stream has AU delimiter. If the stream does not have AU delimiter, please parse access units at your side.

What do you mean by a AU delimiter? Does the above description of Microsoft's demux behaviour seem to satisfy ffdshow's requirements? I think it may not because of the problems I'm having getting perfect decoded frames.

[tetsuo55]

Has anyone tried bitstreaming True-HD or DTS-HD over hdmi?

As ffdshow does not need HDCP or PAP at all it might just work...

[mark0077]

tetsuo, do you know if ffdshow decodes dts-hd ma and true-hd without downsampling?

[tetsuo55]

tetsuo, do you know if ffdshow decodes dts-hd ma and true-hd without downsampling?

i hope it does, but i currently do not have any movies with higher than 48khz tracks so i cannot be sure
So the question is

Do the DTS(-HD) and the (e-)AC-3 decoders support 24 bit and 192khz decoding with full support for all 7.1 channels? And if so does it stay that way through all of the filters until it finally gets output to the audio renderer?

What about the other audio codecs? The complete ffdshow audio path should support 32bit/192khz imho (and all decoders should be able to natively decode 32/192 where sanely possible)

[e-t172]

do you know if ffdshow decodes dts-hd ma and true-hd without downsampling?

- ffdshow does not decode DTS HD MA, just the DTS core (it's even worse than downsampling).
- ffdshow decode TrueHD without downsampling.

[mark0077]

Ah thanks alot, hopefully we will see DTS-HD MA soon ;)

[e-t172]

Ah thanks alot, hopefully we will see DTS-HD MA soon ;)

I don't think so. Sadly, there hasn't been any news on DTS-HD MA in ffdshow for what... months, years ? By the way, I don't understand why TrueHD has been implemented so quickly in libavcodec but not DTS-HD MA...

[mark0077]

Well I'm surprised in two ways, that trueHD can be decoded without always sampling down to 16bit 48khz, thought only some TMT decoders could do that, and I'm also surprised like you that dts-hd ma hasn't had a decoder yet.

[tetsuo55]

the reason is very simple, DTS-HD is far more difficult to reverse engineer, probably due to lack of publicly available information on how it works

[mark0077]

the reason is very simple, DTS-HD is far more difficult to reverse engineer, probably due to lack of publicly available information on how it works

Is there some sort of money or licence involved? I can't imagine anyone like DTS, Dolby etc coming up with something like this and not giving out information on how it works...

[tetsuo55]

Is there some sort of money or licence involved? I can't imagine anyone like DTS, Dolby etc coming up with something like this and not giving out information on how it works...

A company wishing to implement DTS decoding has to pay a license fee and probably sign a NDA contract

[therealjoeblow]

There appears to be a bug with the output framerate when playing back DVD's

All samples act the same, output is always fixed at 25fps

...

Also it appears that DTS over Spdif is not working for me anymore all of suddon (it used to work, i did not touch any of the ffdshow settings) now all i get is silence or pops.
AC3 is working as expected.

I am streaming over HDMI
Build im currently using is 2940

I had problems a few builds back with DTS over SPDIF creating a jerk (like a dropped frame) in the video every 6 seconds. I ended up using AC3Filter for DTS and it doesn't do that. My post about the issue I don't think were ever responded to. I haven't tried a recent build to see if it was fixed, but maybe someone was working on it and didn't complete the fix?

Cheers
The REAL Joe

[mark0077]

MPC-HC expects the pixel aspect ratio in the original DVD is not changed in ffdshow. However, resize with "Process pixel aspect ratio internally" changes PAR. It's the spec of MPC-HC and ffdshow.
You could submit a report to MPC-HC. The solution would be an new API which both should implement.

DVD Aspect Ratios
haruhiko, I posted over on the mpc-hc forum and made a bug / feature request for this.

https://sourceforge.net/tracker/?func=detail&aid=2793920&group_id=170561&atid=854651

However beliyaal replied to the bug report with a comment saying "ffdshow should output the aspect ration
corresponding to the desired PAR.". I am not familiar with how par and sar work together etc but he doesn't seem to think disabling mpc-hc's "Keep Aspect Ratio" should be needed. i am eager to keep this "problem" in upper minds of you guys because I know this effects ALOT of people who are especially new to ffdshow and are trying its resize functionality with their DVD's.

DVD Menu Hang - Still some remaining problem DVD Menus
haruhiko, as promised I have provided a sample of a dvd menu that is able to make ffdshow either stall or act very slowly. In the following dvd menu from star wars episode 2, navigating into the THX tests menu can work very fast without ffdshow in evr-cp in mpc-hc (using any of the latest builds, currently 1120). With ffdshow however, navigating can be extremely slow and often take maybe 5-10 seconds to navigate from one frame to the next. I am simply decoding using libmpeg2 in ffdshow and outputting in RGB32 with no de-interlacing.

Here is the sample that illustrates the remaining problem.
http://www.megaupload.com/?d=GAY1HQAD

To reproduce: Start dvd using video_ts.ifo file and wait for initial white images and menus to pass. These were removed to reduce file size. It might take a minute. You will then see a dvd menu appearing. I have only included one menu path which illustrates the problem. Instructions to navigate through this path is below.

"DVD Main Menu" -> "Language Selection" -> "THX Optimizer Logo" -> "Audio Tests" and then back to "THX Optimizer Menu"

When doing this, move the mouse around a little bit over some of the possible options, then select the next item in the path. This usually lets me reproduce the problem easily. You will notice a large slowdown and sometimes even upto 10 or more seconds delay before you are taken to the next menu. It does not happen every time. One time it may hang and be slow, next time it will be perfect but after a couple of navigations through the menu you should eventually see the problem.

Please let me know if you can reproduce. Other than this dvd menu navigation is almost perfect now with ffdshow and is actually already better than mpc-hc on its own. Thanks.

[haruhiko_yamagata]

What do you mean by a AU delimiter? Does the above description of Microsoft's demux behaviour seem to satisfy ffdshow's requirements? I think it may not because of the problems I'm having getting perfect decoded frames.
Access unit delimiter.
code & 0xffffff1f == 0x00000109

[ipanema]

Access unit delimiter.
code & 0xffffff1f == 0x00000109

Oh right. Yes the demux always starts a new series of 8192 sized samples at a 0, 0, 0, 1, 9 sequence. So the AU code always occurs at the start of a sample.

But I still can't get correctly decoded frames from the ffdshow video apart from occasional frames with a small amount of visible corruption.

There must still be something which ffdshow H.264 decoder is expecting which it is not getting.

Below is a summary of the samples that the demux sends. The start and end times given are stream times (there are no media times set), and they are only set for the first sample of a series.


Received sample 0 (21974974ms) First sample of 1st series
Start time (0x332A48A79A)
End time (0x332A48A79B)
Sync point (1)
Preroll (0)
Discontinuity (1)
Actual data length (8192)
Type changed (0)
0 0 0 1 9 10 0 0 0 1 27 64 0 28 ad 88
11 21 48 30 41 46 85 31 90 89 10 22 4c 9c 99 22
48 82 48 d2 74 91 c4 14 c4 6d 1d 8 23 31 d8 e1
24 b9 ca 31 26 98 a1 3 79 c 53 98 71 83 b0 90
....

Received sample 1 (21974976ms)
Sync point (0)
Preroll (0)
Discontinuity (0)
Actual data length (8192)
Type changed (0)
33 a3 8a 3e 80 39 8 17 1b aa fe bc 24 96 3a 5a
f5 17 62 10 f2 60 5b 9 69 0 5a 52 19 c2 e2 fe
4c 58 d1 a6 cb 2 e8 3c 28 eb 12 0 fa 5 b2 58
....

More samples similar to sample 1.....

Received sample 15 (21974986ms) Final sample of series
Sync point (0)
Preroll (0)
Discontinuity (0)
Actual data length (766)
Type changed (0)
9f 27 c5 3d b6 3c 6c 28 8d 3 41 7c 3a b3 2d f3
28 4e a9 aa 17 49 62 5b 4b 92 17 ff be dc d0 8a
6d 27 47 22 31 d0 2f 19 3 f0 b4 bc c2 53 eb e8
....

Received sample 16 (21974987ms) First sample of 2nd series
Start time (0x332A3E78D5)
End time (0x332A3E78D6)
Sync point (1)
Preroll (0)
Discontinuity (0)
Actual data length (4303)
Type changed (0)
0 0 0 1 9 50 0 0 0 1 28 ea 43 ce 20 92
14 83 8 51 a1 42 c6 14 2a 44 12 4c 85 c8 59 a
24 88 54 d2 74 91 c4 14 c4 6d 1d 8 23 21 ca 38
49 2e 72 8c 49 a2 8a 30 de 43 14 e6 18 60 fc 20
....

Received sample 17 (21974988ms) First sample of 3rd series
Start time (0x332A439038)
End time (0x332A439039)
Sync point (1)
Preroll (0)
Discontinuity (0)
Actual data length (8192)
Type changed (0)
0 0 0 1 9 50 0 0 0 1 6 1 9 0 4 0
6c 68 0 0 3 0 1 80 0 0 0 1 1 9e 37 a2
33 ff d8 e3 9b e1 b1 dc c4 4c 70 18 16 e7 a2 2
....

Received sample 18 (21974988ms)
Sync point (0)
Preroll (0)
Discontinuity (0)
Actual data length (533)
Type changed (0)
e1 26 56 77 2c e0 f 59 35 aa 0 31 4c f1 2b 51
fc 24 9a ce 80 48 8b 0 71 a4 a6 80 2f 98 4f 24
8d a5 2b 7b 5a 10 8a 6a ec 3b 1b 53 ec 15 8e 37
....

Received sample 19 (21974989ms) First sample of 4th series
Start time (0x332A52D660)
End time (0x332A52D661)
Sync point (1)
Preroll (0)
Discontinuity (0)
Actual data length (8192)
Type changed (0)
0 0 0 1 9 30 0 0 0 1 6 1 9 0 6 10
6c 68 0 0 3 0 1 80 0 0 0 1 21 9a 31 4
77 d1 25 71 ad b6 a2 c5 ff af a1 46 b5 63 8d ef
9c 66 e8 fc ed b5 ce 67 e3 7c 80 61 f9 4f 17 c8
....

And so on ..................


Other AVC decoders seem happy with this.

Is there anything in this formatting which the ffdshow H.264 might not like?

[haruhiko_yamagata]

I couldn't find anything wrong.

I can get working filter graph "TS File Source"->Microsoft MPEG-2 Demultiplexer->ffdshow->video renderer in graphedit.exe. It should basically be OK with Microsoft's MPEG-2 Demultiplexer.

Before rev 2955, you had to call IPin::NewSegment. Please try the latest. (Rev 2955 available here (http://www.xvidvideo.ru/content/category/1/4/5/)).
There may be other compatibility issues...

[tetsuo55]

The ffmpeg links to SVN in the start post no longer work, it looks like the entire SVN was deleted

[Leak]

The ffmpeg links to SVN in the start post no longer work, it looks like the entire SVN was deleted
It's svn://svn.ffmpeg.org/ffmpeg/trunk according to the ffmpeg website - i.e. directly using SVN, not through HTTP.

Looks more like they threw out the apache SVN modules and went to serving SVN directly - actually, it looks more like they've thrown out Apache wholly and replaced it with lighttpd.

np: Kettel - Whom (Re: Through Friendly Waters)

[clsid]

Their SVN web-interface is simply not activated anymore since they moved to a different server. I have removed the links.

[Reimar]

Their SVN web-interface is simply not activated anymore since they moved to a different server. I have removed the links.

That's not quite accurate. The web interface was shut down because it was buggy in ways that caused trouble for the server. It was supposed to go live again, but I am not up-to-date on that. This has nothing to do with a server move.
While you're at it, would you mind changing ffmpeg.mplayerhq.hu to ffmpeg.org, the later is the official domain nowadays?
(Also, Apache never ran on the server, and at no time was SVN available over http for FFmpeg. The git repository might be available via http though).

[clsid]

With web-interface I meant the ViewVC interface, like we have for ffdshow.
http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/

git repo is indeed viewable online. For those interested here is the link:
http://git.ffmpeg.org/?p=ffmpeg;a=shortlog

[Casshern]

I can confirm this. I switched to AC3 Filter for outputting DTS (and the DTS core) over sp/dif. Ffdshow instroduces way to much jitter leading to packt drops/repeats in Reclock. This started about a year ago. Mpc hcs internal decoder has the same probs as ffdshow (does not matter whether libdts or libav is used). 2956 still has the bug.
I had problems a few builds back with DTS over SPDIF creating a jerk (like a dropped frame) in the video every 6 seconds. I ended up using AC3Filter for DTS and it doesn't do that. My post about the issue I don't think were ever responded to. I haven't tried a recent build to see if it was fixed, but maybe someone was working on it and didn't complete the fix?

Cheers
The REAL Joe

[haruhiko_yamagata]

DVD Menu Hang - Still some remaining problem DVD Menus
haruhiko, as promised I have provided a sample of a dvd menu that is able to make ffdshow either stall or act very slowly. In the following dvd menu from star wars episode 2, navigating into the THX tests menu can work very fast without ffdshow in evr-cp in mpc-hc (using any of the latest builds, currently 1120). With ffdshow however, navigating can be extremely slow and often take maybe 5-10 seconds to navigate from one frame to the next. I am simply decoding using libmpeg2 in ffdshow and outputting in RGB32 with no de-interlacing.

Here is the sample that illustrates the remaining problem.
http://www.megaupload.com/?d=GAY1HQAD

To reproduce: Start dvd using video_ts.ifo file and wait for initial white images and menus to pass. These were removed to reduce file size. It might take a minute. You will then see a dvd menu appearing. I have only included one menu path which illustrates the problem. Instructions to navigate through this path is below.

"DVD Main Menu" -> "Language Selection" -> "THX Optimizer Logo" -> "Audio Tests" and then back to "THX Optimizer Menu"

When doing this, move the mouse around a little bit over some of the possible options, then select the next item in the path. This usually lets me reproduce the problem easily. You will notice a large slowdown and sometimes even upto 10 or more seconds delay before you are taken to the next menu. It does not happen every time. One time it may hang and be slow, next time it will be perfect but after a couple of navigations through the menu you should eventually see the problem.

Please let me know if you can reproduce. Other than this dvd menu navigation is almost perfect now with ffdshow and is actually already better than mpc-hc on its own. Thanks.
It works for me. The highlight may not follow mouse cursor just after moving to a new page, but if I move mouse again, it responds.
libavcodec doesn't work for that sample. Though it may be due to editing, libmpeg2 seems to be more stable for such streams.

[ipanema]

Please try the latest. (Rev 2955)

I installed rev 2955 (was previously using the latest stable version dated December 2008), and it does behave differently. I seem to get more frames being displayed but there is extreme corruption - lots of pixelation and grey areas, but with small areas of perfect image.

I can get working filter graph "TS File Source"->Microsoft MPEG-2 Demultiplexer->ffdshow->video renderer in graphedit.exe. It should basically be OK with Microsoft's MPEG-2 Demultiplexer.

When I eventually managed to get a connection to ffdshow video decoder in Graphedit it played for a duration much shorter than the file and the renderer window stayed black.

I assumed this might be because there isn't a mechanism in Graphedit for setting the exact demux output pin media type (with the SPS and PPS).

I also tried TS File Source v2.2.0.7 in Graphedit but couldn't get a connection between the demux and ffdshow filter at all. Strangely the TS File Source property page doesn't show any PID info for m2ts files.

Were you using a m2ts file as the source? Also what settings did you use the the GUI properties page for the demux video output pin?

[mark0077]

It works for me. The highlight may not follow mouse cursor just after moving to a new page, but if I move mouse again, it responds.
libavcodec doesn't work for that sample. Though it may be due to editing, libmpeg2 seems to be more stable for such streams.

Yeah thats strange. For me the highlights work perfectly, but the actual delay from clicking an item, to the next page appearing can vary from instant to maybe 10 seconds. I am using evr-cp. Thanks for testing, its my only problem with ffdshow's handling of dvd menu's. I can survive without it but would have been nice to find out why the long delay.

[Mixer73]

Hey guys

I'm having horrible trouble with 2527 build on my MCE box, basically if I have played a couple of xvids eventually MCE will crash with FFDShow.ax as the cause.

I'm running more recent builds on my desktop without any issues so I guess I should run up the latest SVN and see if the the problem persists, and if it does provide some info about the crash?

I noted lots of fixes that seem to impact MCE recently so fingers crossed, love the great work you all do!

[haruhiko_yamagata]

Were you using a m2ts file as the source? Also what settings did you use the the GUI properties page for the demux video output pin?
I was using TS files which worked OK.
It doesn't work with m2ts. ffdshow has connection but black window (no image). Other decoders crash on connection.

[tetsuo55]

There appears to be a bug with the output framerate when playing back DVD's

All samples act the same, output is always fixed at 25fps

Here is a screenshot:
http://img33.imageshack.us/img33/5011/incorrectframerate.th.png (http://img33.imageshack.us/my.php?image=incorrectframerate.png)

And here is the pin out information for that file (same problem with every renderer)

Filter : madVR - CLSID : {E1A8B82A-32CE-4B0D-BE0D-AA68C772E423}

- Connected to:

CLSID: {04FE9017-F873-410E-871E-AB91661A4EF7}
Filter: ffdshow Video Decoder
Pin: Out

- Connection media type:

Video: YV12 720x480 (4:3) 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_YV12 {32315659-0000-0010-8000-00AA00389B71}
formattype: FORMAT_VideoInfo2 {F72A76A0-EB0A-11D0-ACE4-0000C0CC16BA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 518400
cbFormat: 112

VIDEOINFOHEADER:
rcSource: (0,0)-(720,480)
rcTarget: (0,0)-(720,480)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 4
dwPictAspectRatioY: 3
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 720
biHeight: 480
biPlanes: 3
biBitCount: 12
biCompression: YV12
biSizeImage: 518400
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 d0 02 00 00 e0 01 00 00 ........Ð...à...
0010: 00 00 00 00 00 00 00 00 d0 02 00 00 e0 01 00 00 ........Ð...à...
0020: 00 00 00 00 00 00 00 00 80 1a 06 00 00 00 00 00 ........€.......
0030: 00 00 00 00 00 00 00 00 04 00 00 00 03 00 00 00 ................
0040: 00 00 00 00 00 00 00 00 28 00 00 00 d0 02 00 00 ........(...Ð...
0050: e0 01 00 00 03 00 0c 00 59 56 31 32 00 e9 07 00 à.......YV12.é..
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................

Also it appears that DTS over Spdif is not working for me anymore all of suddon (it used to work, i did not touch any of the ffdshow settings) now all i get is silence or pops.
AC3 is working as expected.

I am streaming over HDMI
Build im currently using is 2940It has been confirmed that this is a bug with the "DVD Navigator" so this can be ignored.

[ipanema]

It doesn't work with m2ts. ffdshow has connection but black window (no image). Other decoders crash on connection.

Ah right. I'm using my own m2ts source filter. The only difference between m2ts and m2t is the extra 4 bytes at the start of each packet so it's just as easy to source m2ts files.

I'm having good results using this and the Microsoft MPEG-2 demux filter with other H.264 filters. It's a pity that ffdshow H.264 decoder is still evading a solution. There's a link below to a couple of images showing what I'm getting from the H.254 ffdshow decoder. The first pic shows almost perfect decoding apart from the grey areas, and the second pic shows much more corruption.

http://img7.imageshack.us/gal.php?g=pic1dwk.jpg

I'm obviously very close to a solution but it's difficult to know what more ffdshow is requiring.

[ipanema]

I think I may have found the cause of the problem.

If you use 'H264' as fourcc, ffdshow can parse access units if the stream has AU delimiter. If the stream does not have AU delimiter, please parse access units at your side.

If I understand you correctly, this means that any occurance of the Access UD byte sequence 0 0 0 1 9 must only appear at the start of a new directshow sample arriving at the ffdshow H.264 decoder.

The MPEG-2 demux only starts a new output sample for Access UDs at the start of a frame. It seems there are usually 2 or maybe more Access UD byte sequences occuring for each frame, so perhaps ffdshow is missing these extra Access UDs which could appear anywhere in the samples. I guess other decoders don't care about the alignment of Access UDs on sample boundaries and that's why they are successful.

I don't think it's possible to persuade the MPEG-2 decoder to break samples on Access UDs, so I may have to give up on H.264 ffdshow.

Is there any prospect of a future version of ffdshow where the H.264 decoder doesn't require Access UDs to be always at the start of samples?

[haruhiko_yamagata]

ffdshow does not require access unit delimiters to be placed at the start of DirectShow sample boundary.
Maybe the sample does not have access unit delimiter for all access unit.

[haruhiko_yamagata]

Is there any prospect of a future version of ffdshow where the H.264 decoder doesn't require Access UDs to be always at the start of samples?
Here's another related thread.
http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=968
Some stream does not have access unit delimiter. sensing has kindly offered us a patch, which is unfinished.
I know what I should do to parse such streams, but haven't have time until now. Speaking of future plan, yes. But please don't expect too much, there are too many things in my private todo.

[ikarad]

Is it possible to add an option which allow to define core affinity of different filters or plug-in in ffdshow?

For example, when I want to see interlaced blu-ray movie at 30 fps, I would like to use TIVTC (avisynth plug-in) but IVTC isn't multithread and my cpu (Q6600@3.0ghz) isn't powerfull because on the different cores, yet, some percents (nearly 30 %) is used for video decoding.
My idea would be to use only 3 cores to decode video (we can do with decoder option in ffdshow) and the core 4 would be used only for IVTC.

The problem is when I select 3 threads in decoder option of ffdshow, if I activate TIVTC, it's not the core 4 which is used but one of three cores yet used for decoding.

I don't know if it's possible but if it is possible, it would be wonderfull.

[haruhiko_yamagata]

Is it possible to add an option which allow to define core affinity of different filters or plug-in in ffdshow?
No, it isn't.

[ipanema]

Thanks. That thread looks promising. But I think that the sample files I have always seem to have the complete 00 00 00 01 09 byte sequences.

For example here is the NAL sequence from a typical file:


--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 10)
Sequence parameter set
Picture Parameter set
SEI
SEI
SEI
SEI
SEI
Coded slice of an IDR picture
Access UD (00 00 00 01 09 30)
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 50)
Picture parameter set
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 50)
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 50)
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 50)
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 30)
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 30)
Picture parameter set
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 50)
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 50)
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 30)
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 30)
SEI
Coded slice of a non-IDR picture

--- New MPEG-2 Demux output sample starts here ---
Access UD (00 00 00 01 09 50)
Picture parameter set
SEI
Coded slice of a non-IDR picture
Access UD (00 00 00 01 09 50)


"New MPEG-2 Demux output sample starts here" indicates where a new output sample from the MPEG-2 mux begins, and this always seems to coincide with the start of a new frame. Note that there seem to be 2 Access UDs in each frame - one at the start of the frame (and therefore appears at the start of the demux output sample) and the other could appear anywhere in a sample.

Between the "New MPEG-2 Mux output sample starts here" markers, the output sample will also be split when the byte count exceeds 8192 bytes (that is the maximum sample size from the demux is 8192 bytes). This split occurs with no regard to the sample contents - it just splits at the 8192th byte.

The important thing to notice is that there is always a 00 00 00 01 09 seqence present, and yet I'm still getting the corruption from these sort of files.

As the patch you mentioned seems to address Access UD markers being missing, I'm not sure that it will solve this problem - unless I've missed something from the thread you referenced.

[haruhiko_yamagata]

@ipanema
Can I have a testing environment?
If not, please tell us the full media type on pin connection. SPS and PPS are not required there.

[ipanema]

Can I have a testing environment?
If not, please tell us the full media type on pin connection. SPS and PPS are not required there.

I'm using VC++ and it would be a massive job to extract the code and construct a standalone test case. So here is the media type you asked for. This is when reading the file that I listed the NALs for above.

I've just connected the video output pin of the Demux to the ffdshow video decoder, and then called ConnectionMediaType() on the Demux's video output pin.

This is the media type listed from the Visual Studio debugger with a small amount of reformatting to make it easier to read. Note that I'm using the 8D2D... subtype here but the H264 subtype (34363248-0000-0010-8000-00AA00389B71) gives the same results.


// AM_MEDIA_TYPE
tmpMediaType
majortype {73646976-0000-0010-8000-00AA00389B71} _GUID
subtype {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B} _GUID
bFixedSizeSamples 1 int
bTemporalCompression 0 int
lSampleSize 1 unsigned long
formattype {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA} _GUID (FORMAT_MPEG2_VIDEO)
pUnk 0x00000000 IUnknown *
cbFormat 275 unsigned long
pbFormat 0x00d05318 "" unsigned char *

// pbFormat expanded
(MPEG2VIDEOINFO *)(tmpMediaType.pbFormat)
hdr
rcSource {top=0 bottom=1088 left=0 right=1920} tagRECT
rcTarget {top=0 bottom=1088 left=0 right=1920} tagRECT
dwBitRate 0 unsigned long
dwBitErrorRate 0 unsigned long
AvgTimePerFrame 0 __int64
dwInterlaceFlags 0 unsigned long
dwCopyProtectFlags 0 unsigned long
dwPictAspectRatioX 16 unsigned long
dwPictAspectRatioY 9 unsigned long
dwControlFlags 0 unsigned long
dwReserved1 0 unsigned long
dwReserved2 0 unsigned long
bmiHeader
biSize 40 unsigned long
biWidth 0 long
biHeight 0 long
biPlanes 0 unsigned short
biBitCount 0 unsigned short
biCompression 875967080 unsigned long ('H264')
biSizeImage 0 unsigned long
biXPelsPerMeter 0 long
biYPelsPerMeter 0 long
biClrUsed 0 unsigned long
biClrImportant 0 unsigned long
dwStartTimeCode 0 unsigned long
cbSequenceHeader 139 unsigned long
dwProfile 100 unsigned long
dwLevel 40 unsigned long
dwFlags 0 unsigned long
dwSequenceHeader 0x00d0539c unsigned long [1]

// Sequence header dump (hexadecimal)
pSeqHdr,139 0x00d0539c "" unsigned char *
[0x0] 0x00 unsigned char
[0x1] 0x00 unsigned char
[0x2] 0x00 unsigned char
[0x3] 0x01 unsigned char
[0x4] 0x27 unsigned char
[0x5] 0x64 unsigned char
[0x6] 0x00 unsigned char
[0x7] 0x28 unsigned char
[0x8] 0xad unsigned char
[0x9] 0x88 unsigned char
[0xa] 0x11 unsigned char
[0xb] 0x21 unsigned char
[0xc] 0x48 unsigned char
[0xd] 0x30 unsigned char
[0xe] 0x41 unsigned char
[0xf] 0x46 unsigned char
[0x10] 0x85 unsigned char
[0x11] 0x31 unsigned char
[0x12] 0x90 unsigned char
[0x13] 0x89 unsigned char
[0x14] 0x10 unsigned char
[0x15] 0x22 unsigned char
[0x16] 0x4c unsigned char
[0x17] 0x9c unsigned char
[0x18] 0x99 unsigned char
[0x19] 0x22 unsigned char
[0x1a] 0x48 unsigned char
[0x1b] 0x82 unsigned char
[0x1c] 0x48 unsigned char
[0x1d] 0xd2 unsigned char
[0x1e] 0x74 unsigned char
[0x1f] 0x91 unsigned char
[0x20] 0xc4 unsigned char
[0x21] 0x14 unsigned char
[0x22] 0xc4 unsigned char
[0x23] 0x6d unsigned char
[0x24] 0x1d unsigned char
[0x25] 0x08 unsigned char
[0x26] 0x23 unsigned char
[0x27] 0x31 unsigned char
[0x28] 0xd8 unsigned char
[0x29] 0xe1 unsigned char
[0x2a] 0x24 unsigned char
[0x2b] 0xb9 unsigned char
[0x2c] 0xca unsigned char
[0x2d] 0x31 unsigned char
[0x2e] 0x26 unsigned char
[0x2f] 0x98 unsigned char
[0x30] 0xa1 unsigned char
[0x31] 0x03 unsigned char
[0x32] 0x79 unsigned char
[0x33] 0x0c unsigned char
[0x34] 0x53 unsigned char
[0x35] 0x98 unsigned char
[0x36] 0x71 unsigned char
[0x37] 0x83 unsigned char
[0x38] 0xb0 unsigned char
[0x39] 0x90 unsigned char
[0x3a] 0x90 unsigned char
[0x3b] 0x98 unsigned char
[0x3c] 0x68 unsigned char
[0x3d] 0x58 unsigned char
[0x3e] 0x90 unsigned char
[0x3f] 0x94 unsigned char
[0x40] 0x41 unsigned char
[0x41] 0x24 unsigned char
[0x42] 0xd7 unsigned char
[0x43] 0x5e unsigned char
[0x44] 0xbe unsigned char
[0x45] 0xbf unsigned char
[0x46] 0x18 unsigned char
[0x47] 0x2f unsigned char
[0x48] 0xe4 unsigned char
[0x49] 0x45 unsigned char
[0x4a] 0xf1 unsigned char
[0x4b] 0x48 unsigned char
[0x4c] 0x5f unsigned char
[0x4d] 0x22 unsigned char
[0x4e] 0x2b unsigned char
[0x4f] 0x43 unsigned char
[0x50] 0x10 unsigned char
[0x51] 0x4e unsigned char
[0x52] 0x30 unsigned char
[0x53] 0xe2 unsigned char
[0x54] 0xc2 unsigned char
[0x55] 0x47 unsigned char
[0x56] 0x38 unsigned char
[0x57] 0xed unsigned char
[0x58] 0x30 unsigned char
[0x59] 0x1e unsigned char
[0x5a] 0x01 unsigned char
[0x5b] 0x13 unsigned char
[0x5c] 0xf7 unsigned char
[0x5d] 0x80 unsigned char
[0x5e] 0x88 unsigned char
[0x5f] 0x00 unsigned char
[0x60] 0x00 unsigned char
[0x61] 0x7d unsigned char
[0x62] 0x20 unsigned char
[0x63] 0x00 unsigned char
[0x64] 0x1d unsigned char
[0x65] 0x4c unsigned char
[0x66] 0x07 unsigned char
[0x67] 0x43 unsigned char
[0x68] 0x00 unsigned char
[0x69] 0x07 unsigned char
[0x6a] 0xa1 unsigned char
[0x6b] 0x20 unsigned char
[0x6c] 0x00 unsigned char
[0x6d] 0x08 unsigned char
[0x6e] 0x95 unsigned char
[0x6f] 0x45 unsigned char
[0x70] 0x77 unsigned char
[0x71] 0x97 unsigned char
[0x72] 0x1a unsigned char
[0x73] 0x18 unsigned char
[0x74] 0x00 unsigned char
[0x75] 0x3d unsigned char
[0x76] 0x09 unsigned char
[0x77] 0x00 unsigned char
[0x78] 0x00 unsigned char
[0x79] 0x44 unsigned char
[0x7a] 0xaa unsigned char
[0x7b] 0x2b unsigned char
[0x7c] 0xbc unsigned char
[0x7d] 0xb8 unsigned char
[0x7e] 0x7c unsigned char
[0x7f] 0x22 unsigned char
[0x80] 0x11 unsigned char
[0x81] 0x45 unsigned char
[0x82] 0x80 unsigned char
[0x83] 0x00 unsigned char
[0x84] 0x00 unsigned char
[0x85] 0x00 unsigned char
[0x86] 0x01 unsigned char
[0x87] 0x28 unsigned char
[0x88] 0xea unsigned char
[0x89] 0x43 unsigned char
[0x8a] 0xcb unsigned char


Note that when setting up the media type for the demux video pin, other decoders seem to be happy with most of the structure all zero-ed apart from the dwSequenceHeader - I assume that they get all the info they need from the SPS and PPS which would agree with the Microsoft page

http://msdn.microsoft.com/en-us/library/dd757808(VS.85).aspx

which says:

When the bitstream contains start codes, any of the format types listed here is sufficient, because the decoder does not require any additional information to parse the stream. The bitstream already contains all of the information needed by the decoder, and the start codes enable the decoder to locate the start of each NALU.

But I found that ffdshow needs some fields to be filled in such as rcSource otherwise pins will not connect later on in the graph.

The values above seem to be the same as the values I assigned - that is, no additional fields seem to get filled in during pin connection negotiation.

[haruhiko_yamagata]

OK, I'll see if I can improve anything around this. As you say, it's ffdshow's limitation that it requires some values in the media type.

[ipanema]

Thanks Haruhiko. If you need any more info let me know. Note that the values that I did fill in above I had extracted from the SPS (they aren't just hard-coded guesses).

[haruhiko_yamagata]

Oh, you have already done that, then why not fill bmiHeader::biWidth and biHeight? I'm not sure if it takes any effect though.

[ipanema]

I've been trying a variety of BMI and MPEG2VIDEO fields filled-in over the last couple of days but with no luck in getting rid of the corruption in decoded frames. Sorry I thought that the debuging snapshot I posted had all of the fields filled in that I had tried, but I must have commented out the BMI width and height fields again before that run.

Just to be absolutely sure I just un-commented them (BMI width and height now being set to 1920 and 1088 respectively) and re-ran and there is no difference. I still get corrupted decoded frames.

I think that the size of the video is being picked up correctly in the decoder because the decoded frames seem to be the right size and the small areas that are not corrupted look OK (see the link to screen shots I posted earlier).

[mark0077]

no problem.

Did you try this (http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example)?

I can't reproduce. The sliders work for me.

Unfortunately using the guide at http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example I have many many DVD's which show incorrect aspect ratio when passing through ffdshow's resize, even with mpc-hc's own "Keep Aspect Ratio" disabled. I am resizing to my screen resolution of 1920 x 1080, with "Keep original aspect ratio 1.78:1" selected, this 16:9 movie appears to be squashed to something like 4:3.

With "No aspect ratio correction" these dvd's work fine and are not squashed.

I am still struggling to find a configuration that will allow me to reliably use ffdshow's resize and keep aspect ratio's correct. In this scenario where the DVD is 16:9 (1.78:1 with black bars top and bottom) maybe you might know what is going wrong? Is ffdshow able to handle these type of DVD's yet, ones not marked as 16:9 or 4:3?

I was not clear if you had implemented the forcing of 16:9 or 4:3 for dvd's so maybe this 1.78:1 isn't handled by ffdshow in the current builds but here is my ffdshow settings incase it helps at all.

http://img198.imageshack.us/img198/5350/ffdshowsettings.th.jpg (http://img198.imageshack.us/my.php?image=ffdshowsettings.jpg)

This is the squashed image it produces.
http://img39.imageshack.us/img39/7414/wrongaspectratio.th.jpg (http://img39.imageshack.us/my.php?image=wrongaspectratio.jpg)

I am wondering if
1) There is another flag / aspect ratio variable that needs to be read for DVD's which would give the correct AR, ie 16:9 or 4:3
2) If forcing of 16:9 or 4:3 for DVD content can be implemented in an upcoming version if not already done.

I look forward to getting this issue resolved with your help. Thanks.

[therealjoeblow]

Unfortunately using the guide at http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example I have many many DVD's which show incorrect aspect ratio when passing through ffdshow's resize, even with mpc-hc's own "Keep Aspect Ratio" disabled. I am resizing to my screen resolution of 1920 x 1080, with "Keep original aspect ratio 1.78:1" selected, this 16:9 movie appears to be squashed to something like 4:3.

With "No aspect ratio correction" these dvd's work fine and are not squashed.

I am still struggling to find a configuration that will allow me to reliably use ffdshow's resize and keep aspect ratio's correct. In this scenario where the DVD is 16:9 (1.78:1 with black bars top and bottom) maybe you might know what is going wrong? Is ffdshow able to handle these type of DVD's yet, ones not marked as 16:9 or 4:3?

I was not clear if you had implemented the forcing of 16:9 or 4:3 for dvd's so maybe this 1.78:1 isn't handled by ffdshow in the current builds but here is my ffdshow settings incase it helps at all.


I am wondering if
1) There is another flag / aspect ratio variable that needs to be read for DVD's which would give the correct AR, ie 16:9 or 4:3
2) If forcing of 16:9 or 4:3 for DVD content can be implemented in an upcoming version if not already done.

I look forward to getting this issue resolved with your help. Thanks.

I use ffdshow's resize for DVD's with ZoomPlayer 6 *all* the time, and I don't recall it ever producing an aspect ratio problem. I've watched hundreds of DVD's, anamorphic 16:9, letterboxed 16:9, 4:3, PAL, NTSC, Interlaced, Progressive, homebrew from my Pioneer DVD DVR-650H Recorder, homebrew with ConvertX2DVD, and never an aspect ratio problem that I can recall.

My settings are exactly the same as yours with one main exception:

-under the "Resize" section at the top left, I have HORIZONTAL specified only, 1920 (you have "Resize to screen resolution checked")

-also under "Resize if" I have "Resize Always" checked, but that's not likely to be the problem, it's only the trigger to enable or disable the filter.

Like I said, no issues with aspect ratio ever.

I don't use MPC of any variety - I tried it and found it to be really disfunctional compared to ZP. Maybe try another player and see if it's MPC that's the root of your problems.

[mark0077]

I use ffdshow's resize for DVD's with ZoomPlayer 6 *all* the time, and I don't recall it ever producing an aspect ratio problem. I've watched hundreds of DVD's, anamorphic 16:9, letterboxed 16:9, 4:3, PAL, NTSC, Interlaced, Progressive, homebrew from my Pioneer DVD DVR-650H Recorder, homebrew with ConvertX2DVD, and never an aspect ratio problem that I can recall.

My settings are exactly the same as yours with one main exception:

-under the "Resize" section at the top left, I have HORIZONTAL specified only, 1920 (you have "Resize to screen resolution checked")

-also under "Resize if" I have "Resize Always" checked, but that's not likely to be the problem, it's only the trigger to enable or disable the filter.

Like I said, no issues with aspect ratio ever.

I don't use MPC of any variety - I tried it and found it to be really disfunctional compared to ZP. Maybe try another player and see if it's MPC that's the root of your problems.

See in this case mpc-hc does no aspect ratio adjustment at all (have its internal aspect ratio control disabled), it just shows what its given. In this situation ffdshow is actually squasing up the image horizontally where it should be just outputting the entire frame at 16:9. This is why I feel there must be an error in calculation somewhere.

[MatMaul]

here is a patch to update the spectral extension code of the eac3 decoder :
http://www.etud.insa-toulouse.fr/~mvelten/bordel/ffdshow_eac3_spx.diff

it comes from here :
http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2009-May/069726.html

[clsid]

applied.

[mark0077]

To update on resize problem with various DVD's. Here are details shown on the ffdshow osd of a DVD movie that gets output in a window that looks like 4:3 (mpc-hc's own aspect ratio controls disabled).

With both "No aspect ratio correction" AND "Keep original aspect ratio 1.78:1" the same details below get output
Input 720 x 576 SAR 64/45 DAR 16/9
Output SAR 1/1 DAR 16:9

BUT

With "No aspect ratio correction", this particular disk outputs perfectly in 16:9, but with "Keep original....." it gets output squashed. I think this indicates some problem with "Keep original aspect ratio" setting somewhere as both options give same output on OSD, but massive difference in ACTUAL output ratio.

Grey bars below show what ffdshow is adding to the sides of the content when "keep original aspect ratio" is enabled. Something wrong somewhere.
http://img33.imageshack.us/img33/6266/weirdj.th.jpg (http://img33.imageshack.us/my.php?image=weirdj.jpg)

Thanks.

[her34]

wmv files broken in rev2954 and rev2968

works in rev2946

sample file:
http://www.gametrailers.com/video/preview-hd-prototype/49703

[jmartinr]

To update on resize problem with various DVD's. Here are details shown on the ffdshow osd of a DVD movie that gets output in a window that looks like 4:3 (mpc-hc's own aspect ratio controls disabled).

With both "No aspect ratio correction" AND "Keep original aspect ratio 1.78:1" the same details below get output
Input 720 x 576 SAR 64/45 DAR 16/9
Output SAR 1/1 DAR 16:9

BUT

With "No aspect ratio correction", this particular disk outputs perfectly in 16:9, but with "Keep original....." it gets output squashed. I think this indicates some problem with "Keep original aspect ratio" setting somewhere as both options give same output on OSD, but massive difference in ACTUAL output ratio.

Grey bars below show what ffdshow is adding to the sides of the content when "keep original aspect ratio" is enabled. Something wrong somewhere.
http://img33.imageshack.us/img33/6266/weirdj.th.jpg (http://img33.imageshack.us/my.php?image=weirdj.jpg)

Thanks.

Just checking the obvious: you have enabled "process pixel aspect ratio internally"?

[mark0077]

Yeah here is a screenshot with my settings
http://img198.imageshack.us/img198/5350/ffdshowsettings.th.jpg (http://img198.imageshack.us/my.php?image=ffdshowsettings.jpg)

This is a very simple case above where the output should be 16:9 but ffdshow adds bars to each side of the image.

[tetsuo55]

Yeah here is a screenshot with my settings
http://img198.imageshack.us/img198/5350/ffdshowsettings.th.jpg (http://img198.imageshack.us/my.php?image=ffdshowsettings.jpg)

This is a very simple case above where the output should be 16:9 but ffdshow adds bars to each side of the image.
Judging from your settings this might not be a bug.

What happens when you use these settings instead:
-Multiply by 3
-No aspect ratio correction

[EpsilonX]

To update on resize problem with various DVD's. Here are details shown on the ffdshow osd of a DVD movie that gets output in a window that looks like 4:3 (mpc-hc's own aspect ratio controls disabled).

With both "No aspect ratio correction" AND "Keep original aspect ratio 1.78:1" the same details below get output
Input 720 x 576 SAR 64/45 DAR 16/9
Output SAR 1/1 DAR 16:9

BUT

With "No aspect ratio correction", this particular disk outputs perfectly in 16:9, but with "Keep original....." it gets output squashed. I think this indicates some problem with "Keep original aspect ratio" setting somewhere as both options give same output on OSD, but massive difference in ACTUAL output ratio.

Grey bars below show what ffdshow is adding to the sides of the content when "keep original aspect ratio" is enabled. Something wrong somewhere.
http://img33.imageshack.us/img33/6266/weirdj.th.jpg (http://img33.imageshack.us/my.php?image=weirdj.jpg)

Thanks.

I am using ZP6...
I got the same "squashed" DVD menu...
Both "No aspect ratio correction" and "keep original aspect ratio" resulted the same...
It only happened if I use ZP6 default and recommended aspect ratio setting, which is "Derived"...
Changed ZP aspect ratio setting to "Widescreen 16:9" and the picture is fine...
In MPC, to make the picture correct it required "keep original aspect ratio" in FFDShow and uncheck “Keep Aspect Ratio” in MPC...
Just like what the wiki example...

http://ffdshow-tryout.sourceforge.net/wiki/video:resize_aspect#example

[tetsuo55]

i just ran some tests to try and reproduce this problem

And as EpsilonX said it's a conflict between the MPC aspect ratio settings and the FFdshow aspect ratio settings.

The best setting i was able to find was:
MPC:
-Touch windows from the inside
-Keep aspect ratio

FFdshow: Keep aspect ratio, or no aspect ratio correction, other settings have no influence.

[mark0077]

i just ran some tests to try and reproduce this problem

And as EpsilonX said it's a conflict between the MPC aspect ratio settings and the FFdshow aspect ratio settings.

The best setting i was able to find was:
MPC:
-Touch windows from the inside
-Keep aspect ratio

FFdshow: Keep aspect ratio, or no aspect ratio correction, other settings have no influence.

Thanks for that. I was using haruhiko's suggestion from his link of disabling mpc-hc's keep aspect ratio. I am just trying to nudge the workings of both to be very consistent. At the moment I can find many combinations of settings that doesn't give the right aspect ratio. The only way I find to get aspect ratio correct ALL of the time is to just use mpc-hc keep aspect ratio and ignore ffdshow.

The image I posted above with bars on the side is definitely not right through. In this case ffdshow is the only piece of software I have that adds black bars to the side of this 16:9 movie. Maybe another flag needs to be ignored / forced to a certain value because of the fact that its a DVD?

[tetsuo55]

Thanks for that. I was using haruhiko's suggestion from his link of disabling mpc-hc's keep aspect ratio. I am just trying to nudge the workings of both to be very consistent. At the moment I can find many combinations of settings that doesn't give the right aspect ratio. The only way I find to get aspect ratio correct ALL of the time is to just use mpc-hc keep aspect ratio and ignore ffdshow.

The image I posted above with bars on the side is definitely not right through. In this case ffdshow is the only piece of software I have that adds black bars to the side of this 16:9 movie. Maybe another flag needs to be ignored / forced to a certain value because of the fact that its a DVD?

I did some more tests:
Using the default recommended MPC settings i posted above.
No aspect ratio correction: Image is always correct
Keep aspect ratio: there is a 50% chance of the image being squashed
Manual: Image is always correct(conforms to the input value)

Now if you want to disable MPC's correction and do everything manually with FFdshow you have to match the settings exactly to both the display and the source image (MPC will make a empty canvas the size of the window(or full screen) and then it's up to FFdshow to fill that canvas any way it want's to)

----------

Based on these results i would say the "keep aspect ratio" function is broken.


EDIT:

Can you reproduce this problem with any other container format?
The DVD navigator in windows can be a real problem for some options, also different windows versions have different navigators that all behave differentely, even same version can be behave differentely depending on output renderer.

[haruhiko_yamagata]

To update on resize problem with various DVD's.
...
Does it work if you uncheck resize?

[EpsilonX]

I did some more tests:
Using the default recommended MPC settings i posted above.
No aspect ratio correction: Image is always correct
Keep aspect ratio: there is a 50% chance of the image being squashed
Manual: Image is always correct(conforms to the input value)

Now if you want to disable MPC's correction and do everything manually with FFdshow you have to match the settings exactly to both the display and the source image (MPC will make a empty canvas the size of the window(or full screen) and then it's up to FFdshow to fill that canvas any way it want's to)

----------

Based on these results i would say the "keep aspect ratio" function is broken.


EDIT:

Can you reproduce this problem with any other container format?
The DVD navigator in windows can be a real problem for some options, also different windows versions have different navigators that all behave differentely, even same version can be behave differentely depending on output renderer.

Which "keep aspect ratio" setting..?
FFDShow or MPC's..?
About other container...
Assuming it's encoded correctly, no problem whatsoever...
I use FFDShow Resize's "Keep aspect ratio" and ZP6 in "Derived" setting for those...
It's only DVD for me, cause I have to force ZP6's "Widescreen" aspect ratio....
To sum it up...

Vista SP2 32-bit
FFDShow Tryout rev2972 ICL10
ZoomPlayer 6
Media Player Classic Home Cinema 1.2.1125

FFDShow "Keep original aspect ratio" + ZP6 "Derived" = Squashed
FFDShow “No aspect ratio correction” + ZP6 "Derived" = Squashed
FFDShow "Keep original aspect ratio" + ZP6 "Widescreen" = Normal
FFDShow “No aspect ratio correction” + ZP6 "Widescreen" = Depends on the menu

FFDShow "Keep original aspect ratio" + MPC “Keep Aspect Ratio” enabled = Squashed
FFDShow “No aspect ratio correction” + MPC “Keep Aspect Ratio” enabled = Squashed
FFDShow "Keep original aspect ratio" + MPC “Keep Aspect Ratio” disabled = Normal
FFDShow “No aspect ratio correction” + MPC “Keep Aspect Ratio” disabled = Depends on the menu

*Depends on the menu = some menu are stretched

For now I set FFDShow's to "Keep original aspect ratio", ZP6 to "Widescreen" and uncheck MPC's “Keep Aspect Ratio”...
All my videos seems fine...

[tetsuo55]

Which "keep aspect ratio" setting..?
FFDShow or MPC's..?
About other container...
Assuming it's encoded correctly, no problem whatsoever...
I use FFDShow Resize's "Keep aspect ratio" and ZP6 in "Derived" setting for those...
It's only DVD for me, cause I have to force ZP6's "Widescreen" aspect ratio....
To sum it up...

Vista SP2 32-bit
FFDShow Tryout rev2972 ICL10
ZoomPlayer 6
Media Player Classic Home Cinema 1.2.1125

FFDShow "Keep original aspect ratio" + ZP6 "Derived" = Squashed
FFDShow “No aspect ratio correction” + ZP6 "Derived" = Squashed
FFDShow "Keep original aspect ratio" + ZP6 "Widescreen" = Normal
FFDShow “No aspect ratio correction” + ZP6 "Widescreen" = Depends on the menu

FFDShow "Keep original aspect ratio" + MPC “Keep Aspect Ratio” enabled = Squashed
FFDShow “No aspect ratio correction” + MPC “Keep Aspect Ratio” enabled = Squashed
FFDShow "Keep original aspect ratio" + MPC “Keep Aspect Ratio” disabled = Normal
FFDShow “No aspect ratio correction” + MPC “Keep Aspect Ratio” disabled = Depends on the menu

*Depends on the menu = some menu are stretched

For now I set FFDShow's to "Keep original aspect ratio", ZP6 to "Widescreen" and uncheck MPC's “Keep Aspect Ratio”...
All my videos seems fine...
I mean the ffdshow setting.

Have you also enabled "touch window from the inside" in MPC?

That other containers work properly basically confirms that the problem is caused by the dvd navigator

[EpsilonX]

I mean the ffdshow setting.

Have you also enabled "touch window from the inside" in MPC?

That other containers work properly basically confirms that the problem is caused by the dvd navigator

Yes I did... :D
Along with “Process pixel aspect ratio internally”...
And I resize it to "1360x768"...
Any alternative DVD Navigator to use with ZP or MPC..?
IIRC I've tried Cyberlink's and it failed to connect...

[tetsuo55]

Not that i know of, a lot of people trying to find one though.

What we basically need is an opensource dvd navigator compatible with directshow

[therealjoeblow]

Do the settings for "YCbCr Specification", "Input Levels" and "Output Levels" actually work for anyone? I've tried changing the various options and never see the picture getting darker or lighter as would be expected for any of XviD, H264 or DVD's. The similar options available in CoreAVC produce the expected darkening/lightening of the picture when they are adjusted; as do the similar options when the Haali Renderer is used, but they definitely do nothing in ffdshow?

I tried on two of my machines, both are running WinXP32 SP3; one has an ATI Radeon 3200 (780G) chipset with Catalyst 9.4; the other is a Dell Laptop with an Intel Mobile 965 Express graphics chipset. Same results on both.

[mark0077]

Found another bug relating to aspect ratios. I have just converted all of my DVD TV Show Episodes to mkv's. Using ffdshow, without resize (as I am having too many problems), using ffdshow's mpeg2 decoders, none of these mkv's play at the correct AR. They are squashed with black bars at the top and bottom. Using mpc-hc's own mpeg2 decoder everything is fine. I am getting confused now as this only happens when using one of ffdshow's mpeg2 decoders. Below ar

Using ffdshow Decoders
Original DVD
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

mkv
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 256/135, DAR: 64/27

Original DVD
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

mkv
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

This time, the only combination that gives the incorrect AR is using ffdshow's internal mpeg2 decoders. This is with ffdshow resize off.

AND SURPRISE SURPRISE

Now all of my previous problem dvd's (the 3 or 4 that I have tested) are magically working perfectly even with ffdshow's resize but only when ffdshow's mpeg2 decoders are disabled. haruhiko: Is this where you talked about doing the enforcement of certain aspect ratios for DVD's? Something is going wrong only when using ffdshows mpeg2 decoders. Disabling the ffdshow mpeg2 decoders seemed to do the trick so this might give a clue to whats giving the strange output numbers I showed above. I suppose to test any fixes that may come, we can just check (or at least for this disk I am testing with) if the Output in the ffdshow OSD is "SAR: 64/45, DAR: 16/9" when its internal decoders are enabled

[yesgrey]

Do the settings for "YCbCr Specification", "Input Levels" and "Output Levels" actually work for anyone? I've tried changing the various options and never see the picture getting darker or lighter as would be expected for any of XviD, H264 or DVD's.
Have you checked only RGB32 output in the "supported output colorspaces"?

[tetsuo55]

Something is going wrong only when using ffdshows mpeg2 decoders. Disabling the ffdshow mpeg2 decoders seemed to do the trickI think your conclusion is correct, this would also explain why other codecs don't have this problem.

I even found a disc that would crash with both mpeg2 decoders, but works fine with the one in MPC-HC.
I don't have a clue where the disc is (we changed the living room and the disc got cleaned up somewhere :()

EDIT:
The movie is "Pulse 2"

[cca]

I don't know if anyone else here uses ReClock, but I am unable to use versions higher than rev. 2955 when ffmpeg-mt is used as a decoder. It causes a crash in ReClockHelper.dll .From what I know that helper is trying to construct a directshow graph of the source file in order to extract the fps of the video. If libavcodec is selected as a decoder the crash does not occur, it also does not occur if I have ffmpeg-mt selected with number of threads 2 or 1.

[haruhiko_yamagata]

Found another bug relating to aspect ratios. I have just converted all of my DVD TV Show Episodes to mkv's. Using ffdshow, without resize (as I am having too many problems), using ffdshow's mpeg2 decoders, none of these mkv's play at the correct AR. They are squashed with black bars at the top and bottom. Using mpc-hc's own mpeg2 decoder everything is fine. I am getting confused now as this only happens when using one of ffdshow's mpeg2 decoders. Below ar

Using ffdshow Decoders
Original DVD
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

mkv
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 256/135, DAR: 64/27

Original DVD
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

mkv
Input: Resolution: 720 x 576, SAR: 64/45, DAR 16/9
Output: SAR: 64/45, DAR: 16/9

This time, the only combination that gives the incorrect AR is using ffdshow's internal mpeg2 decoders. This is with ffdshow resize off.

AND SURPRISE SURPRISE

Now all of my previous problem dvd's (the 3 or 4 that I have tested) are magically working perfectly even with ffdshow's resize but only when ffdshow's mpeg2 decoders are disabled. haruhiko: Is this where you talked about doing the enforcement of certain aspect ratios for DVD's? Something is going wrong only when using ffdshows mpeg2 decoders. Disabling the ffdshow mpeg2 decoders seemed to do the trick so this might give a clue to whats giving the strange output numbers I showed above. I suppose to test any fixes that may come, we can just check (or at least for this disk I am testing with) if the Output in the ffdshow OSD is "SAR: 64/45, DAR: 16/9" when its internal decoders are enabled
Then the stream is wrong. MPC-HC gets the AR from container, not the stream. But in that case, it won't be able to decode such streams whose PAR change during playback.
And yes, it's the enforcement that I have spoken of. The enforcement is restricted to DVD playback.
Sample (http://ffdshow-tryout.sourceforge.net/doom9/ARchangeDuringPlayback.ts)

[haruhiko_yamagata]

I don't know if anyone else here uses ReClock, but I am unable to use versions higher than rev. 2955 when ffmpeg-mt is used as a decoder. It causes a crash in ReClockHelper.dll .From what I know that helper is trying to construct a directshow graph of the source file in order to extract the fps of the video. If libavcodec is selected as a decoder the crash does not occur, it also does not occur if I have ffmpeg-mt selected with number of threads 2 or 1.
I'm not sure if it is a bug of ffdshow.
Could you tell me the last working version?

[mark0077]

Then the stream is wrong. MPC-HC gets the AR from container, not the stream. But in that case, it won't be able to decode such streams whose PAR change during playback.
And yes, it's the enforcement that I have spoken of. The enforcement is restricted to DVD playback.
Sample (http://ffdshow-tryout.sourceforge.net/doom9/ARchangeDuringPlayback.ts)

OK Thanks for the clarification. But this particular mkv, when opened in mkvinfo, shows the mpeg2 stream has the values

Pixel width: 720
Pixel height: 576
Display height: 576
Display width: 1024

I think this is correct...., basically saying the image should be stretched to 1024 x 576 right? Are there even more variables like this somewhere else in there that you think are wrong?

[cca]

I'm not sure if it is a bug of ffdshow.
Could you tell me the last working version?

Last known working version is revision 2955.

[haruhiko_yamagata]

OK Thanks for the clarification. But this particular mkv, when opened in mkvinfo, shows the mpeg2 stream has the values

Pixel width: 720
Pixel height: 576
Display height: 576
Display width: 1024

I think this is correct...., basically saying the image should be stretched to 1024 x 576 right? Are there even more variables like this somewhere else in there that you think are wrong?
It's the container AR, I think.
As I said earlier, there are two ways of calculating PAR from the bit-stream, one is wrong of course. So we can only guess the PAR.

[haruhiko_yamagata]

Last known working version is revision 2955.
Then which is the oldest version that does not work? 2956?

[cca]

Then which is the oldest version that does not work? 2956?

I get my builds from www.xvidvideo.ru, the first version after 2955 was 2964, so something that was changed between 2955 and 2964. In the changelog I see many updates to ffmpeg-mt, I guess something in one of those.

[STaRGaZeR]

I don't know if anyone else here uses ReClock, but I am unable to use versions higher than rev. 2955 when ffmpeg-mt is used as a decoder. It causes a crash in ReClockHelper.dll .From what I know that helper is trying to construct a directshow graph of the source file in order to extract the fps of the video. If libavcodec is selected as a decoder the crash does not occur, it also does not occur if I have ffmpeg-mt selected with number of threads 2 or 1.

I have this problem too, but only when the used profile has Avisynth enabled. It's not related to ffmpeg-mt as it happens with any file. However I've experienced it one or two times with the default profile.

[haruhiko_yamagata]

I get my builds from www.xvidvideo.ru, the first version after 2955 was 2964, so something that was changed between 2955 and 2964. In the changelog I see many updates to ffmpeg-mt, I guess something in one of those.
Then it alters timing. I think it is not a bug of ffdshow.
The graph state change from pause to playing when first video frame is delivered. It may be triggering a bug in ReClock. The timing first frame is delivered differs a great deal between libavcodec and ffmpeg-mt.
As the crashing module is ReClockHelper.dll, it is most likely a bug of ReClock. If ffmpeg-mt is causing delayed crash (ie breaking heap), it should crash even without ReClock.

[cca]

Then it alters timing. I think it is not a bug of ffdshow.
The graph state change from pause to playing when first video frame is delivered. It may be triggering a bug in ReClock. The timing first frame is delivered differs a great deal between libavcodec and ffmpeg-mt.
As the crashing module is ReClockHelper.dll, it is most likely a bug of ReClock. If ffmpeg-mt is causing delayed crash (ie breaking heap), it should crash even without ReClock.

I will forward this to the ReClock forums then, see what the developer will say.

[mark0077]

It's the container AR, I think.
As I said earlier, there are two ways of calculating PAR from the bit-stream, one is wrong of course. So we can only guess the PAR.

Ah I see, so its the mpeg2 stream (from the original DVD) thats incorrect, and only when playing from original disk does ffdshow do the workaround for DVD mpeg2 streams? I think I understand now.

I guess it is not recommended to convert DVD to mkv then unless you know the mkv creator performs these corrections to the incorrect DVD aspect ratio flags... I am just confused how both the mkv container AND the stream can be different. The makemkv tool I used should have read the AR (even if incorrect) from the mpeg2 stream and saved that in the mkv container... but instead both must be different which baffles me.

I will try find a tool to edit the mpeg2 streams of all of my newly created mkv's anyways. Thanks.

[tetsuo55]

Ah I see, so its the mpeg2 stream (from the original DVD) thats incorrect, and only when playing from original disk does ffdshow do the workaround for DVD mpeg2 streams? I think I understand now.

I guess it is not recommended to convert DVD to mkv then unless you know the mkv creator performs these corrections to the incorrect DVD aspect ratio flags... I am just confused how both the mkv container AND the stream can be different. The makemkv tool I used should have read the AR (even if incorrect) from the mpeg2 stream and saved that in the mkv container... but instead both must be different which baffles me.

I will try find a tool to edit the mpeg2 streams of all of my newly created mkv's anyways. Thanks.I'm in the proces of converting all my DVD's to MKV using the same tool.
So far i have been lucky, all show correct aspect ratio in the MKV container.

I've dicided to completely avoid all those unfixable/hacky-solution DVD problems.
I had already converted all my other containers to MKV for similar reasons in the past
The only thing i don't have in MKV now is Realmedia (which is stored in the RMVB container), but with the full realmedia support of ffmpeg i technically should be able to do a problem free conversion to MKV

[mark0077]

I'm in the proces of converting all my DVD's to MKV using the same tool.
So far i have been lucky, all show correct aspect ratio in the MKV container.

I've dicided to completely avoid all those unfixable/hacky-solution DVD problems.
I had already converted all my other containers to MKV for similar reasons in the past
The only thing i don't have in MKV now is Realmedia (which is stored in the RMVB container), but with the full realmedia support of ffmpeg i technically should be able to do a problem free conversion to MKV

Good stuff, but are you using mpc-hc's mpeg2 decoder or ffdshow's. If using mpc-hc's you may have badly marked mpeg2 streams also which I think is the case with me. I would rather have them all set correctly so I can use any decoder.

[STaRGaZeR]

haruhiko, rev2976 makes seeking slower. Are dropped frames on seeking really a problem?

[tetsuo55]

Good stuff, but are you using mpc-hc's mpeg2 decoder or ffdshow's. If using mpc-hc's you may have badly marked mpeg2 streams also which I think is the case with me. I would rather have them all set correctly so I can use any decoder. I use libvacodec decoder in ffdshow

[mark0077]

I know its off topic but do you know of a tool to edit the mpeg2 screen AR flags that must be set incorrectly in alot of my mkv's now. Will be nice to fix them all.

[jmartinr]

I know its off topic but do you know of a tool to edit the mpeg2 screen AR flags that must be set incorrectly in alot of my mkv's now. Will be nice to fix them all.

Dvdpatcher (http://www.wincesoft.de/html/dvdpatcher.html) maybe? You'll have to demux the mkv's first of course.

[tetsuo55]

Dvdpatcher (http://www.wincesoft.de/html/dvdpatcher.html) maybe? You'll have to demux the mkv's first of course.That tool looks good, i hope it works.

It's not 100% off topic we basically concluded that to get the file working withing FFdshow it has to be physically changed.

[haruhiko_yamagata]

haruhiko, rev2976 makes seeking slower. Are dropped frames on seeking really a problem?
Yeah, it's slower. But it was really a problem.
Try this (http://x264.nl/h.264.samples/premiere-paff.ts) with Haali's splitter, with 8threads. Haali's splitter drops non ref frames until the first image is delivered to the video renderer. If you use 8 threads, frames are heavily dropped.

[mark0077]

Dvdpatcher (http://www.wincesoft.de/html/dvdpatcher.html) maybe? You'll have to demux the mkv's first of course.

Excellent, well just a final note on this, dvdpatcher set to patch the "entire file" did the trick. Strangely though the only combination of "Aspect Ratio:" and "Size:" that gave fullscreen 16\9 with a DVD I am using to test, was 4:3 and 720*576.

I'm a bit skeptical about these settings... it works and looks fine when passing through ffdshow... but this need to set to 4:3 and an optional bitrate setting in dvdpatcher has me worried I will ruin all of these mpeg2 files of mine :P I would assume the bitrate is just for information purposes rather than effecting quality? Thanks for the input guys, learning alot, bloody DVD's and their messed up AR's.

EDIT: Again when setting the entire mpeg2 file (no mkv containers involved), to 16/9 and 720x576, ffdshows decoders gets the AR wrong, it adds black bars top and bottom. Using mpc-hc's decoders, this plays perfectly fullscreen. I still think ffdshow has a bug somewhere. Setting the entire mpeg2 file to 16/9 and 720x576 with no mkv container shouldn't show any black bars in ffdshow decoders but it does.

haruhiko, with the entire mpeg2 stream set to 16/9 and 720x576 with no mkv containers to mess things up, and with ffdshow mpeg2 decoders, shouldn't I get 16/9 fullscreen? It should know from this that the 720x576 needs to be stretched from 1.25:1 to 16/9 right? Instead it gets displayed at what looks like 1.25:1 WITHIN 16/9 container which adds the black bars, basically ignoring the stretch to fill 16/9. Basically if its 16/9 throughout the mpeg2 file, and I use 16/9 screen I shouldn't see black bars if they arn't already encoded into the stream of course. Seriously looks like a bug to me. Let me know if you need a sample if you want to look into it. Thanks for your time on this issue.

[tetsuo55]

I have been reading a lot about mkv today and it seems mkv passes on whatever the video stream tells the muxer.

So the stream info and mkv container info will always match.
If DVD needs special treatment the mpeg2 files will have to be edited before muxing to mkv.

However what mark just mentioned does sound like a bug to me

[Snowknight26]

So the stream info and mkv container info will always match.

Not always the case.

[the_corona]

Is there a way to do Motion Interpolation in FFDShow, kinda like MotionFlow on TV's? Obviously not to 120hz, but maybe up to 60hz or 50hz.

Only way I found was using AVISynth scripts, but I would guess a "native" method would be alot faster?

Ever since I saw that Tech at a friends TV I've been trying to recreate it, but without much sucess.

PS: don't want to hear the old "but thats not how its supposed to look preachings" :-)

[tetsuo55]

Not always the case.it will only not match if you used a muxer that allows manual override(and you entered a value yourself) or the muxer had a bug.

Default behaviour is copy stream info

[STaRGaZeR]

Yeah, it's slower. But it was really a problem.
Try this (http://x264.nl/h.264.samples/premiere-paff.ts) with Haali's splitter, with 8threads. Haali's splitter drops non ref frames until the first image is delivered to the video renderer. If you use 8 threads, frames are heavily dropped.

I have no dropped frames with that sample with Haali's splitter and ffmpeg-mt with 4 or 8 threads, playback is just fine using rev2975. In fact with rev2976 I get garbled frames at the beginning, and after second 10 or so there is a lot of stuttering and dropped frames with VMR9 and EVR Custom.

[haruhiko_yamagata]

I have no dropped frames with that sample with Haali's splitter and ffmpeg-mt with 4 or 8 threads, playback is just fine using rev2975.After seeking, within first one or two seconds, several frames are dropped.

In fact with rev2976 I get garbled frames at the beginning, and after second 10 or so there is a lot of stuttering and dropped frames with VMR9 and EVR Custom.
Please rebuild ffdshow.ax.

[STaRGaZeR]

After seeking, within first one or two seconds, several frames are dropped.

I can't reproduce. Only 1 or 2 frames are dropped here with 2975 when seeking, and not always. No big deal IMHO. Same with other files. With 2976 0 are dropped.

Please rebuild ffdshow.ax.

That worked, thanks.

[Snowknight26]

ffdshow seems to crash 3-4 seconds into playing this VC-1 stream from the Cloverfield Blu-ray using libavcodec. wmv9 is not affected; r2975.

http://stfcc.org/misc/cloverfield.sample.m2ts - first 20,000,000 bytes.

[ikarad]

1st question: 3 months ago, only mlp decoder existed in ffdshow and allowed to decode trueHD streams. But I notice that now trueHD decoder also exists and I d'ont know the difference between the two decoders.

what is the difference between "mlp" decoder and "trueHD" decoder? Who is the best?

2nd question
to decode ac3 there are two decoders in ffdshow audio: liba52 and libavcodec.
What is the difference between the two decoders?

[clsid]

MLP and TrueHD are two different (but very similar) audio formats.

[ikarad]

MLP and TrueHD are two different (but very similar) audio formats.

What is the best (the best sound quality)?

[FredThompson]

Nope, I mean CPU in ffdshow.
It is placed in "ffdshow (DirectShow) video decoder configuration", Info & CPU page.

I have tested simultaneous reading/writing of HuffYUV with ffdshow and various combinations of the CPU flags as you suggested. It appears MMX must be enabled for decoding HuffYUV. When that is the ONLY flag enabled, the erroneous right-hand "empty" blocks are created with ffdshow_rev2975_20090528_clsid. Is that helpful information?

[clsid]

What is the best (the best sound quality)?http://en.wikipedia.org/wiki/Dolby_TrueHD

[haruhiko_yamagata]

I have tested simultaneous reading/writing of HuffYUV with ffdshow and various combinations of the CPU flags as you suggested. It appears MMX must be enabled for decoding HuffYUV. When that is the ONLY flag enabled, the erroneous right-hand "empty" blocks are created with ffdshow_rev2975_20090528_clsid. Is that helpful information?I still can't reproduce.
Could you try with this sample (http://ffdshow-tryout.sourceforge.net/doom9/huffyuv800x600.avi)?
The file plays fine for me.

[FredThompson]

I still can't reproduce.
Could you try with this sample (http://ffdshow-tryout.sourceforge.net/doom9/huffyuv800x600.avi)?
The file plays fine for me.I still have the problem. This happens even if I use GSpot to raise the ffdshow VFW preference to preferred.

Maybe there is a conflict somewhere. I'm a little puzzled why ffdshow_rev2645_20090131_sse_icl10 works and later versions do not.

Have you looked at the code base to see if anything might have changed unintentionally; typo, compiler change or something like that?

Here is my list of installed codecs.

[clsid]

I don't know what exactly you did in GSpot, but VFW does not have a merit system.

[haruhiko_yamagata]

I still have the problem. This happens even if I use GSpot to raise the ffdshow VFW preference to preferred.

Maybe there is a conflict somewhere. I'm a little puzzled why ffdshow_rev2645_20090131_sse_icl10 works and later versions do not.

Have you looked at the code base to see if anything might have changed unintentionally; typo, compiler change or something like that?

Here is my list of installed codecs.
Thank you for testing. But excuse me, CPU flags control from GUI was broken. I have fixed it at rev 2982. Please wait for a next build and try again.
Just after rev 2645, SIMD code has been changed.
To confirm that ffdshow vfw decoder is being used, you could enable level filter and make it very bright or something.

[FredThompson]

Thank you for testing. But excuse me, CPU flags control from GUI was broken. I have fixed it at rev 2982. Please wait for a next build and try again.
Just after rev 2645, SIMD code has been changed.
To confirm that ffdshow vfw decoder is being used, you could enable level filter and make it very bright or something.OK, I will test again. VirtualDub 1.8.8 only supports VFW so there should be no question that VFW is the method. We're looking for verification of a problem in SIMD code?

[FredThompson]

I don't know what exactly you did in GSpot, but VFW does not have a merit system.If you look at the list of installed codec with GSpot there is a VFw category. Those were all set to 20. I set ffdshow to default, 80, I think, something like that. I've had to do this with quartz and past builds of Sound Forge which trumped it, thus preventing PowerPoint and other old VFW methods of embedded video from working. Haven't had to do that in a long time so I guess Sony fixed that problem. It was just something I thought of from years ago that helped with clashes. Basically, I went looking for conflicts which might be there. Found that stupid Pixie pack and removed it, too. TeamViewer, SUPER or MediaCoder installs that piece of junk.

[haruhiko_yamagata]

If you want to confirm ffdshow is creating the black blocks, the easiest way is to disable huffyuv in VFW decoder configuration.

[FredThompson]

If you want to confirm ffdshow is creating the black blocks, the easiest way is to disable huffyuv in VFW decoder configuration.I don't have the HuffYUV codec installed because ffdshow supports the format. Good idea, though.

[clsid]

If you look at the list of installed codec with GSpot there is a VFw category. Those were all set to 20. I set ffdshow to default, 80, I think, something like that. I've had to do this with quartz and past builds of Sound Forge which trumped it, thus preventing PowerPoint and other old VFW methods of embedded video from working. Haven't had to do that in a long time so I guess Sony fixed that problem. It was just something I thought of from years ago that helped with clashes. Basically, I went looking for conflicts which might be there. Found that stupid Pixie pack and removed it, too. TeamViewer, SUPER or MediaCoder installs that piece of junk.
That must be a hidden option in the Registry then. I know ACM codecs can be set in a preference order in the Device manager, but there is no such option in there for VFW codecs.

[Blight]

Hi Devs,
I'm trying to implement ffdshow as a subtitle filter in Zoom Player.
I'm using the COM API (IffDecoder) in order to interact with ffdshow settings.

To get the basics down first, what I'm trying to do is play a video from one directory and then instruct ffdshow to load a subtitle file from a different directory.

What I've done so far:
* Enable subtitle support in ffdshow (working):
IffDecoder.putParam(IDFF_isSubtitles,Integer(True));

* Set ffdshow to file mode (working):
IffDecoder.putParam(IDFF_subAutoFlnm,Integer(False));

* And now the difficult part, passing along the subtitle file name to FFDShow. Since this is a string rather than a boolean, I've tried using "IffDecoder.putParamStr(IDFF_subFilename,SubFileName)" function, but the specs don't mention the character format. I tried ANSI, UNICODE and UTF8, all fail (the file field isn't filled with the file name and no subtitle is displayed).

I also tried calling this after setting the file name:
IffDecoder.putParam(IDFF_subFlnmChanged,Integer(True));

But it really didn't make a difference. What am I missing?

[Blight]

clsid:
Is it by design that the daily builds have a lower version number than beta 6?

ffdshow_rev2981_20090601_clsid.exe = 1.0.5.2945
ffdshow_beta6_rev2527_20081219.exe = 1.0.6.2524

[clsid]

Version is indeed wrong. I'll fix that.

[Blight]

clsid:
What about loading the subtitles, how can it be done?

[clsid]

Sorry, I can't help you with that.

[haruhiko_yamagata]

Hi Devs,
I'm trying to implement ffdshow as a subtitle filter in Zoom Player.
I'm using the COM API (IffDecoder) in order to interact with ffdshow settings.

To get the basics down first, what I'm trying to do is play a video from one directory and then instruct ffdshow to load a subtitle file from a different directory.

What I've done so far:
* Enable subtitle support in ffdshow (working):
IffDecoder.putParam(IDFF_isSubtitles,Integer(True));

* Set ffdshow to file mode (working):
IffDecoder.putParam(IDFF_subAutoFlnm,Integer(False));

* And now the difficult part, passing along the subtitle file name to FFDShow. Since this is a string rather than a boolean, I've tried using "IffDecoder.putParamStr(IDFF_subFilename,SubFileName)" function, but the specs don't mention the character format. I tried ANSI, UNICODE and UTF8, all fail (the file field isn't filled with the file name and no subtitle is displayed).

I also tried calling this after setting the file name:
IffDecoder.putParam(IDFF_subFlnmChanged,Integer(True));

But it really didn't make a difference. What am I missing?
I think you are using IffDecoder_com.pas in delphi directory.
Binary compatibility of that interface has been broken since 1126 (and again at rev 2852 which has been fixed at rev 2942). I'll fix this.

@developers: please do not add new lines in the midst of an interface. New lines must be added at the end of the interface (or more ideally create a new interface with a new IID).

[Blight]

haruhiko:
How about a very limited interface just for getting/putting config parameter values, basic control over the settings, something that wont ever break on newer builds.

In essence it would only need 2 functions (set/get) which will pass an identifier and a pointer to a pre-allocated memory block. If the pointer is null, the returned function value can be the buffer size required.

Then just write a doc listing what value each identifier expects (UTF8 String, unicode string, word, dword, float, etc...).

This will make things easier and would ensure forward compatibility. And the concept is nearly identical to the existing interface.

In any case, please let me know which build has this fixed so that I can implement the subtitle control. Also, does the code accept UTF8/Unicode strings, it's important for file names, but if it can't, I need to know so that I can pass the 8.3 filename for compatibility.

[albain]

Hi Devs,
I'm trying to implement ffdshow as a subtitle filter in Zoom Player.
I'm using the COM API (IffDecoder) in order to interact with ffdshow settings.

To get the basics down first, what I'm trying to do is play a video from one directory and then instruct ffdshow to load a subtitle file from a different directory.

What I've done so far:
* Enable subtitle support in ffdshow (working):
IffDecoder.putParam(IDFF_isSubtitles,Integer(True));

* Set ffdshow to file mode (working):
IffDecoder.putParam(IDFF_subAutoFlnm,Integer(False));

* And now the difficult part, passing along the subtitle file name to FFDShow. Since this is a string rather than a boolean, I've tried using "IffDecoder.putParamStr(IDFF_subFilename,SubFileName)" function, but the specs don't mention the character format. I tried ANSI, UNICODE and UTF8, all fail (the file field isn't filled with the file name and no subtitle is displayed).

I also tried calling this after setting the file name:
IffDecoder.putParam(IDFF_subFlnmChanged,Integer(True));

But it really didn't make a difference. What am I missing?

Hi,

the string format depends on how ffdshow has been compiled, so it should be unicode for windows 2000/XP/Vista/7.
I guess that it should work again when Haruhiko will have fixed the problem

Otherwise you can use the FFDShowAPI (based on windows messages) or FFDShowAPI.Net which contain all the primitives

[Soulhunter]

I know its already possible to switch profiles via hotkeys...

But, could you maybe add a option to assign/bind hotkeys to profiles?

[haruhiko_yamagata]

haruhiko:
How about a very limited interface just for getting/putting config parameter values, basic control over the settings, something that wont ever break on newer builds.
It's a good idea. Even if I fix the broken compatibility, it breaks compatibility for applications that are built for beta6 (non delphi). If a new interface exists, the application can know what to do.

In essence it would only need 2 functions (set/get) which will pass an identifier and a pointer to a pre-allocated memory block. If the pointer is null, the returned function value can be the buffer size required.

Then just write a doc listing what value each identifier expects (UTF8 String, unicode string, word, dword, float, etc...).

This will make things easier and would ensure forward compatibility. And the concept is nearly identical to the existing interface.

In any case, please let me know which build has this fixed so that I can implement the subtitle control. Also, does the code accept UTF8/Unicode strings, it's important for file names, but if it can't, I need to know so that I can pass the 8.3 filename for compatibility.
wchar_t* should work.

[haruhiko_yamagata]

Also, does the code accept UTF8/Unicode strings, it's important for file names, but if it can't, I need to know so that I can pass the 8.3 filename for compatibility.
Lol, it accepts only char*. I'll create a new one that accepts unicode.

[_xxl]

I have a question:
Why does ffdshow subtitles filter use 60-80 Mb of ram?
Tested with XP sp3, Amd Xp 2000+, fx5200, 512Mb ram, avi Xvid + mp3 sample + .srt subtitle, latest ffdshow 2985 ,MPC-HC, default video renderer.
Subtitles on:
http://i44.tinypic.com/2sacygh.jpg
Subtitles off:
http://i44.tinypic.com/33eiikz.jpg

[avivahl]

I have a question:
Why does ffdshow subtitles filter use 60-80 Mb of ram?
Tested with XP sp3, Amd Xp 2000+, fx5200, 512Mb ram, avi Xvid + mp3 sample + .srt subtitle, latest ffdshow 2985 ,MPC-HC, default video renderer.
Subtitles on:
http://i44.tinypic.com/2sacygh.jpg
Subtitles off:
http://i44.tinypic.com/33eiikz.jpg
Wow, I just checked it and same here. Vista+SP1. WMP11. SRT subtitle.

[Mirabel]

I'm sorry, no offense, but recent ffdshow builds really maul :eek::eek: Windows Explorer, Windows Media Center and WMP. They crash when it comes to thumbnail rendering (I have thumbs for FLV, MKV etc. enabled).

I tried it two times, the last build is ffdshow_rev2984_20090605_sse_icl10, I don't know which build was the first one that didn't worked. I assume it's from the 2nd part of May. I'm now using "ffdshow tryouts revision 2941 May 9 2009 02:47:13 (icl 10, x86, unicode, r)". Both times I had to use system restore, because installing an older build broke the entire thumbnail rendering, even of WMV files.

Explorer just stops rendering thumbs or crashes, the other mentioned apps crash.

Please fix it or I'll have to stick with what I current use. I use ffdshow for ages now, and never experienced that problems. I guess there are other apps which will encounter those problems.

I'm running Vista SP2, and I have this computer not longer than a year and I did not blow it up with codec packs or other stuff which may have broke the thumbnail stuff - reverting to the old ffdshow build fixes it as described. In my eyes it is safe to assume that the problem lies within ffdshow, sadly.

[Atak_Snajpera]

Could you check Generic r2975?

[haruhiko_yamagata]

haruhiko:
How about a very limited interface just for getting/putting config parameter values, basic control over the settings, something that wont ever break on newer builds.

In essence it would only need 2 functions (set/get) which will pass an identifier and a pointer to a pre-allocated memory block. If the pointer is null, the returned function value can be the buffer size required.

Then just write a doc listing what value each identifier expects (UTF8 String, unicode string, word, dword, float, etc...).

This will make things easier and would ensure forward compatibility. And the concept is nearly identical to the existing interface.

In any case, please let me know which build has this fixed so that I can implement the subtitle control. Also, does the code accept UTF8/Unicode strings, it's important for file names, but if it can't, I need to know so that I can pass the 8.3 filename for compatibility.
I have fixed IffDecoder to match delphi. It breaks binary compatibility for recent applications though.
I have added IID_IffDecoder2 which can be queried on an instance of DirectShow video decoder and returns IffDecoder. It can be used to check the version of IffDecoder.
If IID_IffDecoder2 is implemented, getIffDecoder2Version, getParamStrW and putParamStrW should work.
Test build and patch (http://ffdshow-tryout.sourceforge.net/doom9/ffdshow_rev2987_20090608_Q.zip).
Index: src/IffDecoder.h
===================================================================
--- src/IffDecoder.h (revision 2987)
+++ src/IffDecoder.h (working copy)
@@ -7,6 +7,8 @@

// {00F99063-70D5-4bcc-9D88-3801F3E3881B}
DEFINE_GUID(IID_IffDecoder , 0x00f99063, 0x70d5, 0x4bcc, 0x9d, 0x88, 0x38, 0x01, 0xf3, 0xe3, 0x88, 0x1b);
+// {5DD3A966-7365-46d0-B79F-D4973BD88E8D}
+DEFINE_GUID(IID_IffDecoder2, 0x5dd3a966, 0x7365, 0x46d0, 0xb7, 0x9f, 0xd4, 0x97, 0x3b, 0xd8, 0x8e, 0x8d);

struct Tpreset;
struct TglobalSettingsDecVideo;
@@ -66,8 +68,6 @@
STDMETHOD (compat_cpuSupportsMMXEXT)(void) PURE;
STDMETHOD (compat_cpuSupportsSSE)(void) PURE;
STDMETHOD (compat_cpuSupportsSSE2)(void) PURE;
- STDMETHOD (compat_cpuSupportsSSE3)(void) PURE;
- STDMETHOD (compat_cpuSupportsSSSE3)(void) PURE;
STDMETHOD (compat_cpuSupports3DNOW)(void) PURE;
STDMETHOD (compat_cpuSupports3DNOWEXT)(void) PURE;
STDMETHOD (compat_getAVIfps1000_2)(void) PURE;
@@ -132,6 +132,11 @@
STDMETHOD (compat_cpuSupportsSSE42)(void) PURE;
STDMETHOD (compat_cpuSupportsSSE4A)(void) PURE;
STDMETHOD (compat_cpuSupportsSSE5)(void) PURE;
+ STDMETHOD (compat_cpuSupportsSSE3)(void) PURE;
+ STDMETHOD (compat_cpuSupportsSSSE3)(void) PURE;
+ STDMETHOD (compat_getIffDecoder2Version)(void) PURE;
+ STDMETHOD (compat_getParamStrW)(unsigned int paramID,wchar_t *buf,unsigned int buflen) PURE;
+ STDMETHOD (compat_putParamStrW)(unsigned int paramID,const wchar_t *buf) PURE;
};

#ifdef __cplusplus

[haruhiko_yamagata]

I have a question:
Why does ffdshow subtitles filter use 60-80 Mb of ram?
Tested with XP sp3, Amd Xp 2000+, fx5200, 512Mb ram, avi Xvid + mp3 sample + .srt subtitle, latest ffdshow 2985 ,MPC-HC, default video renderer.
Subtitles on:
http://i44.tinypic.com/2sacygh.jpg
Subtitles off:
http://i44.tinypic.com/33eiikz.jpg

40MB is used for buffering image. I'm not sure about the rest, but ffdshow uses 10x or more of the text size. It's not very efficient, but we don't have trouble with 100MB on current PC.

[haruhiko_yamagata]

I'm sorry, no offense, but recent ffdshow builds really maul :eek::eek: Windows Explorer, Windows Media Center and WMP. They crash when it comes to thumbnail rendering (I have thumbs for FLV, MKV etc. enabled).

I tried it two times, the last build is ffdshow_rev2984_20090605_sse_icl10, I don't know which build was the first one that didn't worked. I assume it's from the 2nd part of May. I'm now using "ffdshow tryouts revision 2941 May 9 2009 02:47:13 (icl 10, x86, unicode, r)". Both times I had to use system restore, because installing an older build broke the entire thumbnail rendering, even of WMV files.

Explorer just stops rendering thumbs or crashes, the other mentioned apps crash.

Please fix it or I'll have to stick with what I current use. I use ffdshow for ages now, and never experienced that problems. I guess there are other apps which will encounter those problems.

I'm running Vista SP2, and I have this computer not longer than a year and I did not blow it up with codec packs or other stuff which may have broke the thumbnail stuff - reverting to the old ffdshow build fixes it as described. In my eyes it is safe to assume that the problem lies within ffdshow, sadly.
If libavcodec is selected for VC1, try disabling it.

[haruhiko_yamagata]

I had problems a few builds back with DTS over SPDIF creating a jerk (like a dropped frame) in the video every 6 seconds. I ended up using AC3Filter for DTS and it doesn't do that. My post about the issue I don't think were ever responded to. I haven't tried a recent build to see if it was fixed, but maybe someone was working on it and didn't complete the fix?

Cheers
The REAL Joe@tetsuo55 and therealjoeblow
I have fixed at rev 2990. Thank you for report and sorry for late reply.

[tetsuo55]

@tetsuo55 and therealjoeblow
I have fixed at rev 2990. Thank you for report and sorry for late reply.thanks, as soon as a 2990 build is out i will test. Also want to test the AR fixes asap.

[therealjoeblow]

@tetsuo55 and therealjoeblow
I have fixed at rev 2990. Thank you for report and sorry for late reply.

Thank you haruhiko, I'll test it a bit later.

Cheers,
The REAL Joe

[Thunderbolt8]

Then it alters timing. I think it is not a bug of ffdshow.
The graph state change from pause to playing when first video frame is delivered. It may be triggering a bug in ReClock. The timing first frame is delivered differs a great deal between libavcodec and ffmpeg-mt.
As the crashing module is ReClockHelper.dll, it is most likely a bug of ReClock. If ffmpeg-mt is causing delayed crash (ie breaking heap), it should crash even without ReClock.
I posted this on slysoft forums and the answer there was

"Just disable the DShow estimator."

so how can I do that and does this have any disadvantages?

[haruhiko_yamagata]

I posted this on slysoft forums and the answer there was

"Just disable the DShow estimator."

so how can I do that and does this have any disadvantages?
What is "DShow estimator"?

[cca]

What is "DShow estimator"?

That would be "Directshow estimator", it is a way to extract the fps of the video stream by building a directshow filter chain and query the appropriate filter for the fps. At least this is my understanding of it.

[e-t172]

I posted this on slysoft forums and the answer there was

"Just disable the DShow estimator."

so how can I do that and does this have any disadvantages?

You can disable it in the Reclock options, Video Settings. I don't think there will be serious disadvantages.

[cca]

You can disable it in the Reclock options, Video Settings. I don't think there will be serious disadvantages.

That is completely subjective and depends on the video content used. For me for example there would be a big disadvantage since I use VFR videos 80% of the time, the other estimator of ReClock can't cope with them. But this is getting offtopic.

[haruhiko_yamagata]

I'm using VC++ and it would be a massive job to extract the code and construct a standalone test case. So here is the media type you asked for. This is when reading the file that I listed the NALs for above.

I've just connected the video output pin of the Demux to the ffdshow video decoder, and then called ConnectionMediaType() on the Demux's video output pin.

This is the media type listed from the Visual Studio debugger with a small amount of reformatting to make it easier to read. Note that I'm using the 8D2D... subtype here but the H264 subtype (34363248-0000-0010-8000-00AA00389B71) gives the same results.


// AM_MEDIA_TYPE
tmpMediaType
majortype {73646976-0000-0010-8000-00AA00389B71} _GUID
subtype {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B} _GUID
bFixedSizeSamples 1 int
bTemporalCompression 0 int
lSampleSize 1 unsigned long
formattype {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA} _GUID (FORMAT_MPEG2_VIDEO)
pUnk 0x00000000 IUnknown *
cbFormat 275 unsigned long
pbFormat 0x00d05318 "" unsigned char *

// pbFormat expanded
(MPEG2VIDEOINFO *)(tmpMediaType.pbFormat)
hdr
rcSource {top=0 bottom=1088 left=0 right=1920} tagRECT
rcTarget {top=0 bottom=1088 left=0 right=1920} tagRECT
dwBitRate 0 unsigned long
dwBitErrorRate 0 unsigned long
AvgTimePerFrame 0 __int64
dwInterlaceFlags 0 unsigned long
dwCopyProtectFlags 0 unsigned long
dwPictAspectRatioX 16 unsigned long
dwPictAspectRatioY 9 unsigned long
dwControlFlags 0 unsigned long
dwReserved1 0 unsigned long
dwReserved2 0 unsigned long
bmiHeader
biSize 40 unsigned long
biWidth 0 long
biHeight 0 long
biPlanes 0 unsigned short
biBitCount 0 unsigned short
biCompression 875967080 unsigned long ('H264')
biSizeImage 0 unsigned long
biXPelsPerMeter 0 long
biYPelsPerMeter 0 long
biClrUsed 0 unsigned long
biClrImportant 0 unsigned long
dwStartTimeCode 0 unsigned long
cbSequenceHeader 139 unsigned long
dwProfile 100 unsigned long
dwLevel 40 unsigned long
dwFlags 0 unsigned long
dwSequenceHeader 0x00d0539c unsigned long [1]

// Sequence header dump (hexadecimal)
pSeqHdr,139 0x00d0539c "" unsigned char *
[0x0] 0x00 unsigned char
[0x1] 0x00 unsigned char
[0x2] 0x00 unsigned char
[0x3] 0x01 unsigned char
[0x4] 0x27 unsigned char
[0x5] 0x64 unsigned char
[0x6] 0x00 unsigned char
[0x7] 0x28 unsigned char
[0x8] 0xad unsigned char
[0x9] 0x88 unsigned char
[0xa] 0x11 unsigned char
[0xb] 0x21 unsigned char
[0xc] 0x48 unsigned char
[0xd] 0x30 unsigned char
[0xe] 0x41 unsigned char
[0xf] 0x46 unsigned char
[0x10] 0x85 unsigned char
[0x11] 0x31 unsigned char
[0x12] 0x90 unsigned char
[0x13] 0x89 unsigned char
[0x14] 0x10 unsigned char
[0x15] 0x22 unsigned char
[0x16] 0x4c unsigned char
[0x17] 0x9c unsigned char
[0x18] 0x99 unsigned char
[0x19] 0x22 unsigned char
[0x1a] 0x48 unsigned char
[0x1b] 0x82 unsigned char
[0x1c] 0x48 unsigned char
[0x1d] 0xd2 unsigned char
[0x1e] 0x74 unsigned char
[0x1f] 0x91 unsigned char
[0x20] 0xc4 unsigned char
[0x21] 0x14 unsigned char
[0x22] 0xc4 unsigned char
[0x23] 0x6d unsigned char
[0x24] 0x1d unsigned char
[0x25] 0x08 unsigned char
[0x26] 0x23 unsigned char
[0x27] 0x31 unsigned char
[0x28] 0xd8 unsigned char
[0x29] 0xe1 unsigned char
[0x2a] 0x24 unsigned char
[0x2b] 0xb9 unsigned char
[0x2c] 0xca unsigned char
[0x2d] 0x31 unsigned char
[0x2e] 0x26 unsigned char
[0x2f] 0x98 unsigned char
[0x30] 0xa1 unsigned char
[0x31] 0x03 unsigned char
[0x32] 0x79 unsigned char
[0x33] 0x0c unsigned char
[0x34] 0x53 unsigned char
[0x35] 0x98 unsigned char
[0x36] 0x71 unsigned char
[0x37] 0x83 unsigned char
[0x38] 0xb0 unsigned char
[0x39] 0x90 unsigned char
[0x3a] 0x90 unsigned char
[0x3b] 0x98 unsigned char
[0x3c] 0x68 unsigned char
[0x3d] 0x58 unsigned char
[0x3e] 0x90 unsigned char
[0x3f] 0x94 unsigned char
[0x40] 0x41 unsigned char
[0x41] 0x24 unsigned char
[0x42] 0xd7 unsigned char
[0x43] 0x5e unsigned char
[0x44] 0xbe unsigned char
[0x45] 0xbf unsigned char
[0x46] 0x18 unsigned char
[0x47] 0x2f unsigned char
[0x48] 0xe4 unsigned char
[0x49] 0x45 unsigned char
[0x4a] 0xf1 unsigned char
[0x4b] 0x48 unsigned char
[0x4c] 0x5f unsigned char
[0x4d] 0x22 unsigned char
[0x4e] 0x2b unsigned char
[0x4f] 0x43 unsigned char
[0x50] 0x10 unsigned char
[0x51] 0x4e unsigned char
[0x52] 0x30 unsigned char
[0x53] 0xe2 unsigned char
[0x54] 0xc2 unsigned char
[0x55] 0x47 unsigned char
[0x56] 0x38 unsigned char
[0x57] 0xed unsigned char
[0x58] 0x30 unsigned char
[0x59] 0x1e unsigned char
[0x5a] 0x01 unsigned char
[0x5b] 0x13 unsigned char
[0x5c] 0xf7 unsigned char
[0x5d] 0x80 unsigned char
[0x5e] 0x88 unsigned char
[0x5f] 0x00 unsigned char
[0x60] 0x00 unsigned char
[0x61] 0x7d unsigned char
[0x62] 0x20 unsigned char
[0x63] 0x00 unsigned char
[0x64] 0x1d unsigned char
[0x65] 0x4c unsigned char
[0x66] 0x07 unsigned char
[0x67] 0x43 unsigned char
[0x68] 0x00 unsigned char
[0x69] 0x07 unsigned char
[0x6a] 0xa1 unsigned char
[0x6b] 0x20 unsigned char
[0x6c] 0x00 unsigned char
[0x6d] 0x08 unsigned char
[0x6e] 0x95 unsigned char
[0x6f] 0x45 unsigned char
[0x70] 0x77 unsigned char
[0x71] 0x97 unsigned char
[0x72] 0x1a unsigned char
[0x73] 0x18 unsigned char
[0x74] 0x00 unsigned char
[0x75] 0x3d unsigned char
[0x76] 0x09 unsigned char
[0x77] 0x00 unsigned char
[0x78] 0x00 unsigned char
[0x79] 0x44 unsigned char
[0x7a] 0xaa unsigned char
[0x7b] 0x2b unsigned char
[0x7c] 0xbc unsigned char
[0x7d] 0xb8 unsigned char
[0x7e] 0x7c unsigned char
[0x7f] 0x22 unsigned char
[0x80] 0x11 unsigned char
[0x81] 0x45 unsigned char
[0x82] 0x80 unsigned char
[0x83] 0x00 unsigned char
[0x84] 0x00 unsigned char
[0x85] 0x00 unsigned char
[0x86] 0x01 unsigned char
[0x87] 0x28 unsigned char
[0x88] 0xea unsigned char
[0x89] 0x43 unsigned char
[0x8a] 0xcb unsigned char


Note that when setting up the media type for the demux video pin, other decoders seem to be happy with most of the structure all zero-ed apart from the dwSequenceHeader - I assume that they get all the info they need from the SPS and PPS which would agree with the Microsoft page

http://msdn.microsoft.com/en-us/library/dd757808(VS.85).aspx

which says:



But I found that ffdshow needs some fields to be filled in such as rcSource otherwise pins will not connect later on in the graph.

The values above seem to be the same as the values I assigned - that is, no additional fields seem to get filled in during pin connection negotiation.
I have fixed the limitation that ffdshow required dimension of the video at rev 2994.
Could you remove SPS and PPS from the dwSequenceHeader as a test? I'm not sure if it makes a difference though.

[clsid]

@Albain

AutoCrop is still a bit buggy after your changes in rev 2996. At tolerance 40 (default) it will crop parts of the video. To be more precise, it actually replaced the top/bottom parts of the video with black bars. The output resolution did not change.
The problem does not happen with tolerance 10.
I tested with a video that did not have any black bars, so AutoCrop should have done nothing in that case.

[STaRGaZeR]

How autocrop is supposed to work? I've a good amount of videos with big black bars, both horizontal and vertical, none of them are detected. It's seems the filter is doing nothing. Cropping manually works of course.

[clsid]

You are probably not yet using revision 2996.

[STaRGaZeR]

You are probably not yet using revision 2996.

I'm using it. Whatever I do with the autocrop configuration and/or sources, the output is just the same. No other filters are used.

[Octo-puss]

Will rev 3000 be anything special? It should! :)

[Blight]

Maybe rev 3000 can be ffdshow v1.00 final :P

[swalker]

@tetsuo55 and therealjoeblow
I have fixed at rev 2990. Thank you for report and sorry for late reply.

tetsuo55 and therealjoeblow, have either of you seen the same issue with AC-3 over SPDIF?

[Kurtnoise]

@devs: could you sum up somewhere (in the wiki for example ?) which lib(s) must be compiled mandatory w/ gcc ?

[albain]

@Albain

AutoCrop is still a bit buggy after your changes in rev 2996. At tolerance 40 (default) it will crop parts of the video. To be more precise, it actually replaced the top/bottom parts of the video with black bars. The output resolution did not change.
The problem does not happen with tolerance 10.
I tested with a video that did not have any black bars, so AutoCrop should have done nothing in that case.

I'm using it. Whatever I do with the autocrop configuration and/or sources, the output is just the same. No other filters are used.


Hi,

do you have samples so that I can test with ?

Thanks

Damien

[tetsuo55]

tetsuo55 and therealjoeblow, have either of you seen the same issue with AC-3 over SPDIF?
i have not

[STaRGaZeR]

Sure, here you have one: http://www.megaupload.com/?d=OF3EIVF2 (144MB)

I'm unable to make it work with any source, not specifically this one. It doesn't crop anything.

[haruhiko_yamagata]

@devs: could you sum up somewhere (in the wiki for example ?) which lib(s) must be compiled mandatory w/ gcc ?
Please read our wiki (http://ffdshow-tryout.sourceforge.net/wiki/devel:building#which_compiler_to_use).

[clsid]

For kerneldeint and TomsMoComp I recommend using GCC as well. Gives better performance than ICL10.

[Kurtnoise]

ok, thanks...I was blind this morning.

[haruhiko_yamagata]

For kerneldeint and TomsMoComp I recommend using GCC as well. Gives better performance than ICL10.
OK, updated the wiki.

[albain]

Hi,

do you have samples so that I can test with ?

Thanks

Damien

Okay, I have tested with your sample and it works (with both tolerances 10 & 40) : the slight vertical black bars are removed. Maybe you don't have the right revision ??

@clsid : indeed when the tolerance is too high it will cut some parts of the video (horizontally and/or vertically) but it won't add black bars, if you resize the window of your player you will see those black bars disappear.
The right sample is to have blue bars in the video. I will try to upload a such sample.
Sample :
Sample (http://damienbt.free.fr/Sequence02.avi)

So a lower tolerance should be set by default and also there are 2 kinds of tolerances which I calculate and compare to the tolerance parameter :
1/ Tolerance inside a line : it should have the same luminance on all its length (standard deviation should be low). However sometimes there may be a logo on the black bars.
2/ Tolerance between 2 lines : both should have the same luminance (average luminance should be equals)

Maybe those 2 should not be compared in the same way or maybe an additional parameter is necessary to tune autocrop.

[STaRGaZeR]

Yes albain, I'm using the right revision. I was using my own build but I just downloaded xvidvideo.ru's to be sure and it still does nothing. I've it configured like this:

http://thumbnails6.imagebam.com/3880/fc493c38793625.gif (http://www.imagebam.com/image/fc493c38793625)

That sample has wide vertical and horizontal black bars. I get nothing from any of the options. It must be something in my side if it works for you, but what?

[clsid]

@clsid : indeed when the tolerance is too high it will cut some parts of the video (horizontally and/or vertically) but it won't add black bars, if you resize the window of your player you will see those black bars disappear.
The right sample is to have blue bars in the video. I will try to upload a such sample.
So a lower tolerance should be set by default and also there are 2 kinds of tolerances which I calculate and compare to the tolerance parameter :
1/ Tolerance inside a line : it should have the same luminance on all its length (standard deviation should be low). However sometimes there may be a logo on the black bars.
2/ Tolerance between 2 lines : both should have the same luminance (average luminance should be equals)

Maybe those 2 should not be compared in the same way or maybe an additional parameter is necessary to tune autocrop.
A few comments and ideas:
* I think the tolerance option should be removed. The average user will have no clue what value to set there. So lets not complicate things for them. The less unnecessary options the better.
* Some fixed tolerances should be set internally. I am sure that the feedback of some of the regular testers here can help to find good values.
* Different constraints should use different tolerances.
* Perhaps chroma values should also be taken into consideration? The pixels in a line should have approximately the same color, not just the same luminance.
* Is your algorithm designed to crop black bars only, or does it support other colors? Restricting the algorithm to black only, could probably simplify the algo. Instead of the pixels having almost same color, the algo could then check if they are all (almost) black.
* Logo detection: perhaps allow a small range of the pixels in a line to have different color (non-black)? For example, take the first and last pixel in the line that have a 'different' color than the rest. If the distance between those two pixels is less than 10-15% of the length of the whole line, then there might be a logo.

[therealjoeblow]

tetsuo55 and therealjoeblow, have either of you seen the same issue with AC-3 over SPDIF?

I'm pretty sure I only saw it with DTS. That's why it took me so long to figure out what the issue really was, because it only occurred from time to time and I wasn't paying particular attention to the source.

[albain]

A few comments and ideas:
* I think the tolerance option should be removed. The average user will have no clue what value to set there. So lets not complicate things for them. The less unnecessary options the better.
* Some fixed tolerances should be set internally. I am sure that the feedback of some of the regular testers here can help to find good values.
* Different constraints should use different tolerances.
* Perhaps chroma values should also be taken into consideration? The pixels in a line should have approximately the same color, not just the same luminance.
* Is your algorithm designed to crop black bars only, or does it support other colors? Restricting the algorithm to black only, could probably simplify the algo. Instead of the pixels having almost same color, the algo could then check if they are all (almost) black.
* Logo detection: perhaps allow a small range of the pixels in a line to have different color (non-black)? For example, take the first and last pixel in the line that have a 'different' color than the rest. If the distance between those two pixels is less than 10-15% of the length of the whole line, then there might be a logo.

The algorithm will work for coloured bars too (which can occur even if it is not common)
Okay to remove the tolerance, but as you said tests will be needed with samples : bad sample (such as vhs capture), a good sample, some samples with a logo

[Inventive Software]

Maybe rev 3000 can be ffdshow v1.00 final :P

Nah, they're saving that for rev 4000! :p

[haruhiko_yamagata]

Maybe rev 3000 can be ffdshow v1.00 final :P
Congratulations, you have gotten it.

[clsid]

There seems to be a bug in ffdshow that causes Explorer to crash on Windows 7 (build 7231).
Last known good revision is 2992. First known bad revision is 2996. I will try to narrow it down further.
Any ideas on what might be the cause?

[Blight]

clsid:
Possibly due to the iffdecoder com object api fixes introduced in build 2987? I know it's earlier than 2992, but it's pretty close and a 3rd party app using the api may crash, taking explorer with it. Perhaps some plugin that uses ffdshow to get a thumbnail image.

[Spec-Chum]

I know this question has been asked before, but I can't seem to find a definative answer, does ffdshow apply the +10db to the LFE channel when decoding ac3/dts/flac?

Also, is it true that bass redirection replaces the LFE channel instead of adding to it? My av amp adds either 0-15db to the SW input, but I just can't get it to sound right.

Many thanks

[clsid]

clsid:
Possibly due to the iffdecoder com object api fixes introduced in build 2987? I know it's earlier than 2992, but it's pretty close and a 3rd party app using the api may crash, taking explorer with it. Perhaps some plugin that uses ffdshow to get a thumbnail image.
I just finished installing Windows 7 build 7231 and I don't get any crashes myself. The crash was reported to me by someone else. I'll try to get more details from that user.

[haruhiko_yamagata]

There seems to be a bug in ffdshow that causes Explorer to crash on Windows 7 (build 7231).
Last known good revision is 2992. First known bad revision is 2996. I will try to narrow it down further.
Any ideas on what might be the cause?
Rev 2994 is most likely.
I think the user has a file that makes ffdshow crash. I need that file.

[clsid]

I think it may have been a false alarm.

[albain]

Same thing here, I had crashes with earlier revisions but not anymore on recent ones (since the boost library updates)

[her34]

wmv files broken in rev2954 and rev2968

works in rev2946

sample file:
http://www.gametrailers.com/video/preview-hd-prototype/49703

wmv still broken in rev2992

[ipanema]

I have fixed the limitation that ffdshow required dimension of the video at rev 2994.
Could you remove SPS and PPS from the dwSequenceHeader as a test? I'm not sure if it makes a difference though.

Sorry fo the delay in replying. I'm not sure what you mean. The apparent requirement for the video dimension to be filled in was just to get the pins to connect correctly - it was just an extra observation.

Do you mean that I should set the dwSequenceHeader to NULL? But that would deprive the decoder of all the info in the SPS and PPS which decoders seem to require (and largely ignore corresponding info in the main part of the MPEG2VIDEOINFO structure).

The problem of the decoded H.264 being corrupted still exists in rev3004 that I've just installed and tried out.

[haruhiko_yamagata]

Sorry fo the delay in replying. I'm not sure what you mean. The apparent requirement for the video dimension to be filled in was just to get the pins to connect correctly - it was just an extra observation.
Yes.

Do you mean that I should set the dwSequenceHeader to NULL? But that would deprive the decoder of all the info in the SPS and PPS which decoders seem to require (and largely ignore corresponding info in the main part of the MPEG2VIDEOINFO structure).

Yes, and in that case, decoders can extract the dimensions from the stream as far as the stream has SPS and PPS. Just as a test.

[Gleb Egorych]

There is a problem with ffdshow and DVBViewer. Looks like a bug in DVBViewer source filter that confuses ffdshow (from beta6 to r3004).

On mono channels (MPEG2 video + MP2 audio) with DVBViewer I had no sound, so I made a discovery and supposed that DVBViewer sent wrong channel number to decoder.
I tested all available options for MP2 decoding, here are the results:

libmad: detects input as stereo and gives no output, info page is blank.
mp3lib: detects input as stereo and outputs as stereo.
libavcodec: detects input as mono and outputs as mono.

Looks like libavcodec is the most foolproof, but by default libmad is used for MP2. So I believe libmad has higher quality (?).

With another DVB program mono channels are perfeclty detected and played with all three libraries.

[Gleb Egorych]

http://www.gametrailers.com/video/preview-hd-prototype/49703
...
wmv still broken in rev2992
I confirm, broken in 3004, works in 2940.

[haruhiko_yamagata]

There is a problem with ffdshow and DVBViewer. Looks like a bug in DVBViewer source filter that confuses ffdshow (from beta6 to r3004).

On mono channels (MPEG2 video + MP2 audio) with DVBViewer I had no sound, so I made a discovery and supposed that DVBViewer sent wrong channel number to decoder.
I tested all available options for MP2 decoding, here are the results:

libmad: detects input as stereo and gives no output, info page is blank.
mp3lib: detects input as stereo and outputs as stereo.
libavcodec: detects input as mono and outputs as mono.

Looks like libavcodec is the most foolproof, but by default libmad is used for MP2. So I believe libmad has higher quality (?).

With another DVB program mono channels are perfeclty detected and played with all three libraries.
If you have off line file that can reproduce the issue, please send us one.

[ipanema]

OK I set the dwSequenceHeader to NULL and filled in some of the fields in the main part of the media type structure. The values should be correct because I'v read them from the SPS and PPS of the file.

It seems to play pretty much the same. Most of the frame is grey blockiness but there are patches of near-perfectly decoded video.

Here is a dump of the media type structure in case it helps:


- mt
+ majortype {73646976-0000-0010-8000-00AA00389B71} _GUID
+ subtype {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B} _GUID
bFixedSizeSamples 1 int
bTemporalCompression 0 int
lSampleSize 1 unsigned long
+ formattype {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA} _GUID
+ pUnk 0x00000000 IUnknown *
cbFormat 136 unsigned long
+ pbFormat 0x0239af20 "" unsigned char *


- pMVI
- hdr
+ rcSource {top=0 bottom=1088 left=0 right=1440} tagRECT
+ rcTarget {top=0 bottom=1088 left=0 right=1440} tagRECT
dwBitRate 0 unsigned long
dwBitErrorRate 0 unsigned long
AvgTimePerFrame 0 __int64
dwInterlaceFlags 0 unsigned long
dwCopyProtectFlags 0 unsigned long
dwPictAspectRatioX 16 unsigned long
dwPictAspectRatioY 9 unsigned long
dwControlFlags 0 unsigned long
dwReserved1 0 unsigned long
dwReserved2 0 unsigned long
- bmiHeader
biSize 40 unsigned long
biWidth 1440 long
biHeight 1088 long
biPlanes 0 unsigned short
biBitCount 0 unsigned short
biCompression 875967080 unsigned long
biSizeImage 0 unsigned long
biXPelsPerMeter 0 long
biYPelsPerMeter 0 long
biClrUsed 0 unsigned long
biClrImportant 0 unsigned long
dwStartTimeCode 0 unsigned long
cbSequenceHeader 0 unsigned long
dwProfile 77 unsigned long
dwLevel 40 unsigned long
dwFlags 0 unsigned long
- dwSequenceHeader 0x0239afa4 unsigned long [1]
[0] 0 unsigned long

[Gleb Egorych]

If you have off line file that can reproduce the issue, please send us one.
No, sample is useless, because only live playback is affected. If record TS file and play it even in DVBViewer itself then there is no problem.

[haruhiko_yamagata]

OK I set the dwSequenceHeader to NULL and filled in some of the fields in the main part of the media type structure. The values should be correct because I'v read them from the SPS and PPS of the file.

It seems to play pretty much the same. Most of the frame is grey blockiness but there are patches of near-perfectly decoded video.

Here is a dump of the media type structure in case it helps:
Thank you for testing anyway. I'll think about this.

[haruhiko_yamagata]

No, sample is useless, because only live playback is affected. If record TS file and play it even in DVBViewer itself then there is no problem.
OK, please send the file to me anyway. MPEG-2 + mono audio is not so common in my collection. I'll try to simulate imaginable mediatypes.

[clsid]

I propose releasing Beta 7 sometime next month. Of course all regressions since beta 6 should be fixed first.

Known regressions:
- WMV3 decoding crash. This is a bug in FFmpeg code. Will hopefully be fixed soon.
- SPDIF related issues? Can anyone who suffers from this problem summarize the problem, including last known good revision and first known bad revision?

If there is anything I forgot, please let me know.

Improving AutoCrop before releasing beta 7 would be nice. But since it was broken before, not a requirement.

@Haruhiko, do you have any plans for changes that you would like to include in beta 7?

[haruhiko_yamagata]

@Haruhiko, do you have any plans for changes that you would like to include in beta 7?
I'm trying to fix libavcodec MPEG-2 timestamps calculation.
Next month is good. I would call it RC1.

[ipanema]

Haruhiko. I just tried to PM you a debug log as requested, but after sending it the Control Panel still says I have 0 sent messages so I do not know if the message was really sent - there were no error messages.

If you don't receive it please let me know.

[Gleb Egorych]

OK, please send the file to me anyway. MPEG-2 + mono audio is not so common in my collection. I'll try to simulate imaginable mediatypes.
Here you are: http://www.mediafire.com/?tobwmzjokzn

[MatMaul]

an other issue :
ffdshow don't get back to ac3 decoding after a switch to eac3.
so with my dvb software change from an ac3 track to an eac3 track is ok but not the other way around.

[albain]

I propose releasing Beta 7 sometime next month. Of course all regressions since beta 6 should be fixed first.

Improving AutoCrop before releasing beta 7 would be nice. But since it was broken before, not a requirement.

I am working on it (improvements of the algorithm : temporal analysis added, tuning of the parameters)

I will release binary builds links on this forum only for testing purposes.

If some have samples to test with (with vhs quality vs dvd quality, with logo) I am interested.

an other issue :
ffdshow don't get back to ac3 decoding after a switch to eac3.
so with my dvb software change from an ac3 track to an eac3 track is ok but not the other way around.
How does the streams switching occur ? Is it made within FFDShow internal audio splitter or from an external audio splitter ?
Do you have a sample ?

[MatMaul]

How does the streams switching occur ? Is it made within FFDShow internal audio splitter or from an external audio splitter ?
Do you have a sample ?
It is internal to ffdshow.
My dvb software uses the same decoder for ac3 and eac3 decoding (they use the same mediatype anyway if I remember correctly).
When a switch occurs the graph isn't rebuild, there is no connection or reconnection of the filters, and the dvb soft doesn't "warn" in any way the decoder that the codec have changed.
perhaps we need to separate the ac3/eac3 decoding as we do with ac3/mp2 if you think this is not the responsability of ffdshow to detect the change.
The weird thing is it works great when switching from ac3 to eac3 (I can see the codec change in the info tab of ffdshow).

EDIT : I know you are french so if you have access to a dvb-t tuner you can test that with the software I am talking about :
http://pouchintv.baysse.fr/forum/
just switch from an HD channel (eac3) to an SD channel and set the audio track to ac3.

[haruhiko_yamagata]

OK I set the dwSequenceHeader to NULL and filled in some of the fields in the main part of the media type structure. The values should be correct because I'v read them from the SPS and PPS of the file.

It seems to play pretty much the same. Most of the frame is grey blockiness but there are patches of near-perfectly decoded video.

I think ffdshow's access unit parsing is not enabled for your filter because your filter does not set file name on connection. If it is a file, please set file name in FILTER_INFO::achName. But if it is a live stream, it does not have file name and extension.

It was AVI files that I had to add that broken hack.
AVI splitter and "Neuview Source" (http://www.neuviewed.com/neuview6) use MEDIASUBTYPE_H264 and use start code. This is OK, but it doesn't have access unit delimiter at all. One DirectShow sample is one access unit in this case. So if ffdshow enables access unit parsing that depends on access unit delimiters, it won't find any access units.

That's why ffdshow has to look at the extension of the file name. AVI splitter set file name on connection, no problem. However, Neuview Source does not set file name at all.

If the file name is NULL, ffdshow has no hint if it should parse access units depending on access unit delimiter. Now ffdshow parses access units if the file name extension is NULL. I have added workaround to find "Neuview Source" in the upper stream and if it is detected, ffdshow does not parse access units.

I'll commit this local change soon.

Of course the best solution is to parse access units without depending on access unit delimiters. I know what I should do, but it's too hard.

[clsid]

Can't you also add detection of the "AVI splitter" filter so no filename check is needed at all?

[haruhiko_yamagata]

Can't you also add detection of the "AVI splitter" filter so no filename check is needed at all?
There are several AVI splitters that are familiar to us and there may be some other AVI splitters.

[ipanema]

If it is a file, please set file name in FILTER_INFO::achName. But if it is a live stream, it does not have file name and extension.

Now ffdshow parses access units if the file name extension is NULL.


I'm a little confused by your explanation. If ffdshow parses access units if the file extension is NULL, then this seems to be what I need because the stream I deliver does have 00 00 00 01 09 sequences, and they MAY occur in the middle of a directshow sample.

Does that mean I can just hard code FILTER_INFO::achName to be something like

"dummy"

This has no file extension, and it sounds like ffdshow does not need to know the actual filename - it is just looking for the absence or presence of a filename extension.

This would seem to be safer than specifying the full pathname of the file with its extension because I don't know what specific filename extensions ffdshow might be checking for - it might do different things depending on the filename extension.

Also do you mean that "Now ffdshow parses access units if the file name extension is NULL" has already been implemented or that it will be implemented in a future revision?

How are you reading the FILTER_INFO::achName ? Are you just getting the upstream filter and calling QueryFilterInfo() on it? I ask because you said a a while back that you would need the filter GUID, but I think you can get the upstream filter without having to know its GUID in advance via IPin::QueryPinInfo() on the upstream connected pin.

[haruhiko_yamagata]

I'm a little confused by your explanation. If ffdshow parses access units if the file extension is NULL, then this seems to be what I need because the stream I deliver does have 00 00 00 01 09 sequences, and they MAY occur in the middle of a directshow sample.

Does that mean I can just hard code FILTER_INFO::achName to be something like

"dummy"

This has no file extension, and it sounds like ffdshow does not need to know the actual filename - it is just looking for the absence or presence of a filename extension.
If you want to support current and older (say, beta 6) ffdshow, please use "dummy.ts".

This would seem to be safer than specifying the full pathname of the file with its extension because I don't know what specific filename extensions ffdshow might be checking for - it might do different things depending on the filename extension.

Current ffdshow (not my local working copy) compare extension with
ts, m2ts, m2t, mts, mpg, mpeg.
So it's safe to specify Drive:\path\filename\extension.

The new version will compare extension with avi.

Also do you mean that "Now ffdshow parses access units if the file name extension is NULL" has already been implemented or that it will be implemented in a future revision?
In the near future, this week probably.

How are you reading the FILTER_INFO::achName ? Are you just getting the upstream filter and calling QueryFilterInfo() on it?
Yes.

I ask because you said a a while back that you would need the filter GUID, but I think you can get the upstream filter without having to know its GUID in advance.
I don't seem to require your filter's GUID for now, but I required NeuviewSource's GUID for the workaround.

[ipanema]

The upstream filter from the ffdshow filter is the MPEG-2 demux, so I set the FILTER_INFO::achName of the demux and also of my source filter by calling AddFilter specifying a filter name ending in ".ts" - for example:

m_pGraph->AddFilter(pDemux, L"Demux.ts");

Unfortunately this makes no difference. The Source File is still shown as blank and the decoded frames are still corrupted.

I'm setting up the media type WITH the dwSequenceHeader (SPS and PPS) and the main media type structure set as posted earlier.

P.S. I'v just tried to query the achName of the Demux from my source filter (the opposite direction to which you'll be looking from the ffdshow filter) using the following code:

CComPtr <IPin> nDemuxInPin;
pSourceOutPin->ConnectedTo(&nDemuxInPin);
PIN_INFO pininfo;
nDemuxInPin->QueryPinInfo(&pininfo);
CComPtr <IBaseFilter> pDemuxFilter = pininfo.pFilter;
FILTER_INFO filterinfo;
pDemuxFilter->QueryFilterInfo(&filterinfo);

and the filterinfo.achName does indeed read "Demux.ts"

[haruhiko_yamagata]

The upstream filter from the ffdshow filter is the MPEG-2 demux, so I set the FILTER_INFO::achName of the demux and also of my source filter by calling AddFilter specifying a filter name ending in ".ts" - for example:

m_pGraph->AddFilter(pDemux, L"Demux.ts");

Unfortunately this makes no difference. The Source File is still shown as blank and the decoded frames are still corrupted.

I'm setting up the media type WITH the dwSequenceHeader (SPS and PPS) and the main media type structure set as posted earlier.

P.S. I'v just tried to query the achName of the Demux from my source filter (the opposite direction to which you'll be looking from the ffdshow filter) using the following code:

CComPtr <IPin> nDemuxInPin;
pSourceOutPin->ConnectedTo(&nDemuxInPin);
PIN_INFO pininfo;
nDemuxInPin->QueryPinInfo(&pininfo);
CComPtr <IBaseFilter> pDemuxFilter = pininfo.pFilter;
FILTER_INFO filterinfo;
pDemuxFilter->QueryFilterInfo(&filterinfo);

and the filterinfo.achName does indeed read "Demux.ts"
Sorry, I found ffdshow is using IFileSourceFilter::GetCurfile.
ffdshow searches to the upper stream and find the source filter.const char_t* TinputPin::getFileSourceName(void)
{
IFilterGraph *graph=m_pFilter->GetFilterGraph();
if (!graph || wasGetSourceName || !filesourceFlnm.empty())
return filesourceFlnm.c_str();
wasGetSourceName=true;
comptr<IBaseFilter> filter;
if (searchPrevNextFilter(PINDIR_INPUT,this,this,&filter,TpinFileSourceComp()) && filter)
{
comptr<IFileSourceFilter> ifsf;filter->QueryInterface(IID_IFileSourceFilter,(void**)&ifsf);
LPOLESTR aviNameL=NULL;
ifsf->GetCurFile(&aviNameL,NULL);
if (aviNameL)
{
filesourceFlnm = text<char_t>(aviNameL);
CoTaskMemFree(aviNameL);
}
else
filesourceFlnm=_l("");
}
return filesourceFlnm.c_str();
}

struct TpinFileSourceComp
{
bool operator ()(IBaseFilter *bff,IPin*) const
{
comptr<IFileSourceFilter> ifsf=NULL;
return SUCCEEDED(bff->QueryInterface(IID_IFileSourceFilter,(void**)&ifsf)) && ifsf!=NULL;
}
};

Please implement IFileSourceFilter.
Thank you for your consideration of this matter.

[ipanema]

Haruhiko, that works fine now. :)

Thankyou for looking into this.

"Now ffdshow parses access units if the file name extension is NULL"

By NULL do you just mean a filename without an extension - e.g. "myfile" or "myfile." ?

[haruhiko_yamagata]

Haruhiko, that works fine now. :)

Thankyou for looking into this.



By NULL do you just mean a filename without an extension - e.g. "myfile" or "myfile." ?
Yes, ffdshow parses access units if the file name does not have extension or IFileSourceFilter is not implemented.

[ipanema]

Now that this seems to be working, here's an observation.

Most of the time ffdshow does not seem to decode the 2 B frames that preceed the first I frame (in presentation order) in the stream. It does seem to do this most of the time at the start of a file but if you start streaming data from an I-frame located elsewhere in a file then the 2 preceding B frames are not decoded (as if it was an open GOP). But Mainconcept always seems to decode the 2 preceeding B frames no matter which I-frame in the file you start streaming from (as if all GOPs were closed).

Do you know what flags in the stream the ffdshow's H.264 decoder is using to decide not to decode the 2 leading B frames? And is it making a mistake? If Mainconcept can perfectly decode these B frames OK then surely ffdshow should be able to aswell.

[therealjoeblow]

Improving AutoCrop before releasing beta 7 would be nice. But since it was broken before, not a requirement.


I'm not sure if it's even possible so please forgive me if not, but if anyone's going to be working on AutoCrop, what I've been longing for is for AutoCrop to be able to determine the actual image's height (I believe in the same standard way it currently does), and then crop the sides off to as close to a true 16:9 image as it can (a modification to the code would be necessary for this variation).

The reason is that I have a plasma display used exclusively for movies, and to avoid uneven screen wear and potential burn-in, I don't want the black-bars displayed on 2.35:1 or anything wider than 16:9. Right now I use ZoomPlayer to manually zoom the image for extra-widescreen movies, but it's not automatic. Having a variation of AutoCrop that could do it automatically would be great if possible!

[Komano]

Hello Guys, I have been having problems with S/PDIF and if you need me to test the S/PDIF Fixes, i can help. PM me or something?

[Snowknight26]

There appears to be a bug in libavcodec where FRAPS videos have the right-most 16 pixels a different level, or so it appears. Doesn't happen when using decoder that comes with FRAPS but does with libavcodec.

Sample source: http://stfcc.org/misc/epiczip.zip (disregard the other files)
Sample encode from source using libavcodec: http://stfcc.org/misc/fraps.libavcodec.mkv

http://i39.tinypic.com/11qqqkx.png

[Casshern]

Hi,

the Dolby Tru HD decoder works well with the latest releases. Also nice is that when you turn it off, ffdshow extracts the ac3 core and sends that over sp-dif. So far, so good. But some Dolby Tru HD tracks lack an ac3 core. In this case the preferred behaviour would be:
1) If ac3 core of a dolby tru hd track is available send that over sp-dif
2) if ac3 core is not available use the dolby tru had decoder

All that would be required would be an option on the dolby tru hd decoder like "Only use if no ac3 core is detected"

The reason for all this is, that many people got receivers only supporting ac3 and dts. Here it is nicer to have the ac3 core, than to decode and reencode or even than to use the analog outs.

[Gleb Egorych]

There is a problem with ffdshow and DVBViewer. Looks like a bug in DVBViewer source filter that confuses ffdshow (from beta6 to r3004).

On mono channels (MPEG2 video + MP2 audio) with DVBViewer I had no sound, so I made a discovery and supposed that DVBViewer sent wrong channel number to decoder.
I tested all available options for MP2 decoding, here are the results:

libmad: detects input as stereo and gives no output, info page is blank.
mp3lib: detects input as stereo and outputs as stereo.
libavcodec: detects input as mono and outputs as mono.

Looks like libavcodec is the most foolproof, but by default libmad is used for MP2. So I believe libmad has higher quality (?).

It can be fixed within DVBViewer, the option is called "TV/Radio Pre-Format Detection".

[Gleb Egorych]

Question about presets:
I tried to setup ffdshow specially for ReClock, made a preset with autoloading condition "on a DirectShow filter presence"="ReClock Audio Renderer", but ffdshow all the same uses default preset. Also tried "*ReClock*" with no luck. Does the condition work only for filters before ffdshow?

Using rev2940

[Leak]

Question about presets:
I tried to setup ffdshow specially for ReClock, made a preset with autoloading condition "on a DirectShow filter presence"="ReClock Audio Renderer", but ffdshow all the same uses default preset. Also tried "*ReClock*" with no luck. Does the condition work only for filters before ffdshow?
Yes, as autoloading conditions are only checked when ffdshow is inserted into the graph, which usually happens before audio and renderer filters are added.

[leeperry]

Also tried "*ReClock*" with no luck. Does the condition work only for filters before ffdshow?
you'd be better off using one specific player for Reclock and make a ffdshow preset depending on its .exe name

[Gleb Egorych]

Yes, as autoloading conditions are only checked when ffdshow is inserted into the graph, which usually happens before audio and renderer filters are added.
Maybe this behaviour should be changed (if possible). Strange situation: filter exists, name is correct, rule does not work.
you'd be better one using one specific player for Reclock and make a ffdshow preset depending on its .exe name
Already done, it was the only option. Now I have mplayerc_rc.exe with mplayerc_rc.ini for ReClock.

[haruhiko_yamagata]

There appears to be a bug in libavcodec where FRAPS videos have the right-most 16 pixels a different level, or so it appears. Doesn't happen when using decoder that comes with FRAPS but does with libavcodec.

Sample source: http://stfcc.org/misc/epiczip.zip (disregard the other files)
Sample encode from source using libavcodec: http://stfcc.org/misc/fraps.libavcodec.mkv

http://i39.tinypic.com/11qqqkx.png
ffplay has the same issue. Please report it directly to FFmpeg.

[haruhiko_yamagata]

Now that this seems to be working, here's an observation.

Most of the time ffdshow does not seem to decode the 2 B frames that preceed the first I frame (in presentation order) in the stream. It does seem to do this most of the time at the start of a file but if you start streaming data from an I-frame located elsewhere in a file then the 2 preceding B frames are not decoded (as if it was an open GOP). But Mainconcept always seems to decode the 2 preceeding B frames no matter which I-frame in the file you start streaming from (as if all GOPs were closed).

Do you know what flags in the stream the ffdshow's H.264 decoder is using to decide not to decode the 2 leading B frames? And is it making a mistake? If Mainconcept can perfectly decode these B frames OK then surely ffdshow should be able to aswell.
They are simply dropped without any flags. It's a bug then. I'll re-read the spec and think about the fix.

[leeperry]

I was wondering, is there a way to decode MP3 in a DivX/AVI file w/ ffdshow audio at all? I keep getting an "ACM wrapper" in my graph files....even w/ using HMS :confused:

apparently MP3 in an AVI container has to go through ACM? only MPC's Audio Decoder fixes the problem...but I still can't seem to be able to use ffdshow's MP3 decoder w/ AVI files :o

[clsid]

That normally works fine. Perhaps the audio in your file is stored in an unusual way. Otherwise is is just user fail :)

[albain]

Autocrop feature
Description : autocrop is a feature included in the crop filter of FFDShow that scans and removes the black borders (vertical, horizontal, or both).
The scan is performed each x milliseconds (delay parameter), and the scan is stopped once one considers that the analysis is done (stop parameter, after 100 seconds usually it is okay, after the logos, warnings,...).
There is a tolerance parameter (the one that needs to be tuned) that lets tune the scan according to the video quality and the difference of luminance levels between the bars and the real picture.
Also, once the scan has reach a highwater level, the crop won't be increased anymore, only decreased : the reason is simple, if black bars are detected on a clear picture, and after this clear picture there is for example a dark area at the bottom for example, the autocrop won't be increased to include this dark area.
Last detail : the scan works only on colored bars (not necessarily black).

Since last time I have brought some modifications to the algorithm.

Don't forget to change the tolerance level : I used 30 for my tests and it seems to be working fine

Here is the link of the test build
FFDShow with Autocrop (http://damienbt.free.fr/ffdshow_rev3014_20090620_dbt.exe)

I am waiting for your feedbacks, thanks !

[Atak_Snajpera]

after this revision

Revision 2976 - Directory Listing
Modified Fri May 29 14:47:09 2009 UTC (3 weeks, 1 day ago) by h_yamagata

ffmpeg-mt:
DirectShow workaround for delivering the first frame without delay.
This fixes frame drops on seek.

ffmpeg-mt causes crashes during generating previews

http://img38.imageshack.us/img38/4338/new1f.th.png (http://img38.imageshack.us/i/new1f.png/)

Tested 2986 and 3008 on vista sp1

[clsid]

That should be fixed at 3013. Please test.

[Atak_Snajpera]

excellent! Can't wait for new revision :)

[clsid]

Build is online now.

[Atak_Snajpera]

thanks! Problem solved :)

[leeperry]

That normally works fine. Perhaps the audio in your file is stored in an unusual way. Otherwise is is just user fail :)
I'm on XP SP3, Gabest Audio Decoder works fine, but ffdshow doesn't ?! apparently both MPC and KMP use an "ACM wrapper" to decode MP3, they don't seem to be able to connect w/ ffdshow...I've tried several AVI splitters(HMS/Gabest/KMP/windows built-in), problem remains :

http://thumbnails7.imagebam.com/3963/c08cf039627413.gif (http://www.imagebam.com/image/c08cf039627413) http://thumbnails18.imagebam.com/3963/16b64139627414.gif (http://www.imagebam.com/image/16b64139627414)

it happens w/ any divx/mp3 AVI file and seems to be related to "WAVEFORMATEX: wFormatTag: MPEG LAYER 3(0x55)($0055)"

same problem occurs w/ MKV(using HMS)...basically I can't get ffdshow to decode mp3 in a container.

maybe I miss a source filter?! will try to find one...I guess Gabest's Audio Decoder runs its own.

PS: MpaSplitter.ax doesn't help, it won't connect to ffdshow..

[clsid]

It works fine here on XPSP3 and it has worked fine for years. There is nothing wrong with ffdshow.

Your graph contains two audio decoders, which is weird. My guess: you are playing a file with multiple audio streams. And You messed with the "multiple ffdshow instances" setting in ffdhsow audio decoder.
OR
Reclock is fucking things up.

[leeperry]

these are single audio track files, and I'm forcing ffdshow audio to EQ/volume/matrix mixing/winamp2 plugin...same happens w/o Reclock :

http://thumbnails11.imagebam.com/3963/33178a39629973.gif (http://www.imagebam.com/image/33178a39629973)

and it works w/o an ACM wrapper using the external Gabest's Audio Decoder .ax :

http://thumbnails18.imagebam.com/3963/670c8339629974.gif (http://www.imagebam.com/image/670c8339629974)

even if I allow multiple ffdshow audio instances, problem remains...care to share your DS graph please? I lack a proper MP3 source filter obviously, Gabest's Decoder seems to run its own :o

[Komano]

No one needs me to help in S/PDIF?

[clsid]

@leeperry
Upload a sample file.
Why the fuck do you think you need a MP3 source filter? You don't. Period.
Also check if you haven't accidentally disabled MP3 in ffdshow. The ACM wrapper normally only gets used if no suitable DS Filter can be found.

@Komano
Download the latest build (3013) and test it.

[netwolf]

Last detail : the scan works only on colored bars (not necessarily black).
Thank you for the updated algorithm!
One question though: aren't those borders almost always black?
Does this mean that auto-crop doesn't work in all those cases, or did I misunderstand you?

[ipanema]

They are simply dropped without any flags. It's a bug then. I'll re-read the spec and think about the fix.

OK. Incidentally, I've just noticed an ambiguity in the wording of my previous message - here it is clarified:


Most of the time ffdshow's H.264 decoder does not seem to decode the 2 B frames that preceed the first I frame (in presentation order) in the stream. It DOES seem to decode the 2 B frames most of the time at the start of a FILE, but if you start streaming data from an I-frame located elsewhere in a file then the 2 preceding B frames are NOT decoded ...

[clsid]

Thank you for the updated algorithm!
One question though: aren't those borders almost always black?
Does this mean that auto-crop doesn't work in all those cases, or did I misunderstand you?
It works on all bars that are evenly colored. So black, grey, white, blue, etc.

[netwolf]

Thank you for clarification, clsid! Good to hear :)
I was afraid that black didn't qualify as a color, that's why I asked.

[STaRGaZeR]

I found this:

When you export all settings to a .reg file before reinstalling Windows for example, when you restore them, the avisynth script you had previously is not restored. The script itself is in the .reg file, it's just not restored.

[Snowknight26]

ffdshow (r2981 using RGB32 as the output) doesn't decode MPNG properly. Colors are severely messed up (similar to an issue I reported a while back).

http://stfcc.org/misc/epiczip.zip

Steps to reproduce:
ffmpeg.exe -i "hl2 2009-06-18 14-34-56-64.avi" -vcodec png -an test.avi
Play test.avi with ffdshow
ffmpeg.exe -i test.avi -vcodec huffyuv test2.avi
Play test2.avi with ffdshow

Screenshot from test.avi:
http://i42.tinypic.com/k9crxh.png

[Komano]

ffdshow_rev3013_20090620_clsid_x64 seems to be running fine with S/PDIF working properly.
Tested with
Full HD 1080P Files with DTS and DTS-ES Audio

My System Specs are:
Windows 7 Build 7229 X64
DFI LP DK P35 T2RS + Q6600
Nvidia 8800GT
4GB Ram
and i think thats about what you need to know. Though, my PC has known (http://www.cccp-project.net/forums/index.php?topic=3599.new#new) to produce fake results. -__-"

[fastplayer]

@Komano:
6) No warez, cracks, serials or illegally obtained copyrighted content! Links to content of a questionable nature, asking for, offering, or asking for help/helping to process such content in any way or form is not tolerated. (http://forum.doom9.org/forum-rules.htm)

[Blight]

Question:
How does the FFDShow subtitle renderer handle multiple subtitle files?

Case sample:
"my video.avi"
"my video.en.srt"
"my video.jp.srt"

In this sample, you have an AVI video file with two external subtitle streams. One is English and the other Japanese. How do you cycle through the subtitle tracks? Is there such a mechanism. How about supporting the IAMStreamSelect interface and exposing all subs through it? Refreshing the subtitle property page for multiple sub files might also be needed.

[haruhiko_yamagata]

Question:
How does the FFDShow subtitle renderer handle multiple subtitle files?

Case sample:
"my video.avi"
"my video.en.srt"
"my video.jp.srt"

In this sample, you have an AVI video file with two external subtitle streams. One is English and the other Japanese. How do you cycle through the subtitle tracks? Is there such a mechanism. How about supporting the IAMStreamSelect interface and exposing all subs through it? Refreshing the subtitle property page for multiple sub files might also be needed.
As far as I know, ffdshow doesn't have such mechanism.
The idea is good and I accept it as feature request, but please don't expect too much as I already have too many todo.

[Komano]

Does the CCCP Project forum link count as questionable content? O_O

[Eragon4ever]

No, your files do.

[Ger]

There is a problem with ffdshow and DVBViewer. Looks like a bug in DVBViewer source filter that confuses ffdshow (from beta6 to r3004).

On mono channels (MPEG2 video + MP2 audio) with DVBViewer I had no sound, so I made a discovery and supposed that DVBViewer sent wrong channel number to decoder.
I tested all available options for MP2 decoding, here are the results:

libmad: detects input as stereo and gives no output, info page is blank.
mp3lib: detects input as stereo and outputs as stereo.
libavcodec: detects input as mono and outputs as mono.

Looks like libavcodec is the most foolproof, but by default libmad is used for MP2. So I believe libmad has higher quality (?).

With another DVB program mono channels are perfeclty detected and played with all three libraries.

I can confirm this. I have been using libavcodec in ffdshow for mp2 for a while now for this reason. libmad is silent on mono channels.

It can be fixed within DVBViewer, the option is called "TV/Radio Pre-Format Detection".

I can't confirm this though. I have had "TV/Radio Pre-Format Detection" enabled in the DirectX options for a long time and it doesn't solve the libmad/mono issue here. I tried disabling this setting as well in case that was what Gleb meant (and rebuilding the graph several times), but it doesn't matter. libmad is still silent with the live stream. Just tested again with DVBViewer 4.2.0.1beta and ffdshow build 3014.

The recorded files work though, so I can't reproduce the issue with a recording, but Haruhiko mentioned he might be able to simulate the live stream conditions.

I haven't noticed any negative side effects after switching to libavcodec for mp2, but I'm no audiophile and I have no idea if there are sound quality differences.

@Haruhiko
Since you mentioned you had few of these mono samples I made another sample (http://sharebee.com/d3567e2e).

There is also an issue with ffdshow deinterlacing with this sample (this issue can be reproduced easily with this sample in MPC-HC + ffdshow and doesn't need any live stream simulation or DVBViewer). It seems to be badly flagged so unless I enable "force bob" for hardware deinterlacing or "process frames flagged as progressive" for ffdshow software deinterlacing the interlacing is still visible.

The problem is that when I enable either of these forced deinterlacing methods the picture becomes shaky/jumpy. Notice how the animated bottles are shaking early on. Also notice how the channel logo jumps now and then. This is tested with Nvidia hardware deinterlacing and with ffdshow's yadif with double frame rate, and the shaking is there in both cases. However with MPC-HC's internal MPEG2 decoder/deinterlacer it's a steady/stable picture that looks much better than in ffdshow. I don't think it was always like this in ffdshow. I think the shaking started a few months ago.

EDIT:
Setting "Top field first" manually instead of "auto" fixes of the shaky bottles/animation in both Yadif and hardware deint. The channel logo is still flickering a bit (completely stable with MPC-HC decoder), but I guess that may just be a side effect of the Nvidia and Yadif deinterlacing methods, which are both good with scrollers/fast motion. So I guess the field order detection may have changed at some point some months ago?

[clsid]

I found this:

When you export all settings to a .reg file before reinstalling Windows for example, when you restore them, the avisynth script you had previously is not restored. The script itself is in the .reg file, it's just not restored.
I remember that in the past there were problems with exporting the avisynth script to .reg. It got garbled because it wasn't escaped properly. The workaround was to simply not export it. The proper solution was never implemented.

There should be no problems if you manually export the registry keys from Regedit.exe.
HKCU\Software\GNU\ffdshow*

[STaRGaZeR]

"avisynthScript"="LoadPlugin("W:\\\\aa\\\\bb\\\\cc\\\\dd\\\\TIVTC.dll")
tfm(order=1,mode=3)
tdecimate(mode=1)"

That's what it is in the .reg, the \\\\ were obviously not there before. Other than that the script is correctly exported as far as I can see.

[clsid]

Ok, then it is just a bug with escaping.

[Leak]

That's what it is in the .reg, the \\\\ were obviously not there before. Other than that the script is correctly exported as far as I can see.
Except for the line breaks, which are actually the culprit.

Those would need to be escaped, as regedit will read the first line as an unterminated string and the following lines as syntax errors. Escaping the line breaks as '\n' should fix that...

np: Fink - Sort Of Revolution (Sort Of Revolution)

[leeperry]

I personally gave up on exporting settings a long time ago, regedit works perfectly fine :)

I usually had empty avisynth scripts..

[Gleb Egorych]

I can't confirm this though. I have had "TV/Radio Pre-Format Detection" enabled in the DirectX options for a long time and it doesn't solve the libmad/mono issue here. I tried disabling this setting as well in case that was what Gleb meant (and rebuilding the graph several times), but it doesn't matter. libmad is still silent with the live stream. Just tested again with DVBViewer 4.2.0.1beta and ffdshow build 3014.

Actually, I still have some issues with mono channels, let's continue here: http://www.dvbviewer.info/forum/index.php?showtopic=34657

[Blight]

With regards to exporting settings,
You know you can run regedit with command line parameters to export a registry tree section to a file silently.

Basically:
regedit.exe /s /a "c:\ffdshow.reg" "HKEY_CURRENT_USER\Software\GNU\ffdshow"

If I remember correctly, "/A" means ANSI output (for win9x compatibility), so it wont work if ffdshow stores unicode registry string values.

And integrating back by executing the "ffdshow.reg" file in explorer.

[swalker]

I am waiting for your feedbacks, thanks !

Autocropping isn't working here using varying levels of tolerance with the downloaded mp4 of http://video.google.com/videoplay?docid=-580692223856788375

Edit: Actually it does work if one selects the Resize & Aspect -> Aspect ratio -> No aspect ratio correction radio button. The nextFrameMS=msec+50; // gradual application of autocrop feels funny though.

[STaRGaZeR]

I found why autocrop wasn't working for me. Resize & aspect "Process whole image" checkbox should be disabled, otherwise it won't work.

________________________________________________________________________


Forget what I was saying about the subtitle filter. The problem is not that. After more testing, when playing DVDs in MPC-HC, the Subtitles filter is always loaded, even if it's unchecked. Using RGB32 as output and Auto as YCbCr specification, then:



With Resize & aspect disabled:



With Subtitles enabled or disabled, the displayed subtitles have colors like when manually selecting BT.709. Manually changing YCbCr specification makes no difference in the subtitles' color.



With Resize & aspect enabled:



With Subtitles disabled, the displayed subtitles have colors like when manually selecting BT.709. Manually changing YCbCr specification makes no difference in the subtitles' color.


With Subtitles enabled, the displayed subtitles have colors like when manually selecting BT.601. Manually changing YCbCr specification changes subtitles' color.



The source (http://www.megaupload.com/?d=R3JHNSQH) is 720x480, and Resize & aspect is configured to specify horizontal size of 1680 pixels resulting in 1680x1260 (4:3 DVD). With this DVD menu is very easy to see the difference between BT.601 and BT.709 because it has red dots. My suggestions after this testing:


ffdshow Subtitles filter should not be loaded if it's not checked.

What's the subtitles' color specification, BT.601 or BT.709? Whatever it is, make it consistent between using and not using Resize & aspect and between manually using BT.601 or BT.709.

[Nicholi]

What is the difference between the "High quality YV12 to RGB conversion" checkbox on the Output tab with all the colorspace outputs, and the "High quality YV12 to RGB conversion" checkbox on the RGB conversion tab in the Method groupbox?

Or is this just accidental duplication of the same feature in two spots?

[Snowknight26]

It's the same thing. You uncheck one and the other one gets unchecked, and vice versa.

[haruhiko_yamagata]

Actually, I still have some issues with mono channels, let's continue here: http://www.dvbviewer.info/forum/index.php?showtopic=34657
OK, please wait. I'll look into it when I have time.

[haruhiko_yamagata]

"avisynthScript"="LoadPlugin("W:\\\\aa\\\\bb\\\\cc\\\\dd\\\\TIVTC.dll")
tfm(order=1,mode=3)
tdecimate(mode=1)"

That's what it is in the .reg, the \\\\ were obviously not there before. Other than that the script is correctly exported as far as I can see.
avisynthScript exists just for backward compatibility, it is not used in current version of ffdshow. avisynthScriptMULTI_SZ is the current one.
I can't reproduce. Could you send the reg file to me?

[clsid]

The WMV3/VC-1 crashes have been fixed at revision 3016.

[STaRGaZeR]

avisynthScript exists just for backward compatibility, it is not used in current version of ffdshow. avisynthScriptMULTI_SZ is the current one.
I can't reproduce. Could you send the reg file to me?

ffdshow32.reg - 0.34MB (http://www.zshare.net/download/61753639876e1bf5/)

Write an avisynth script and export it, then uninstall and reinstall ffdshow and restore the .reg. The avisynth script will be empty.

[albain]

Autocropping isn't working here using varying levels of tolerance with the downloaded mp4 of http://video.google.com/videoplay?docid=-580692223856788375

Edit: Actually it does work if one selects the Resize & Aspect -> Aspect ratio -> No aspect ratio correction radio button. The nextFrameMS=msec+50; // gradual application of autocrop feels funny though.
Okay thanks I will test your sample

[haruhiko_yamagata]

ffdshow32.reg - 0.34MB (http://www.zshare.net/download/61753639876e1bf5/)

Write an avisynth script and export it, then uninstall and reinstall ffdshow and restore the .reg. The avisynth script will be empty.
After writing an avisynth script, did you move to other page before you exported the settings?

[STaRGaZeR]

After writing an avisynth script, did you move to other page before you exported the settings?

I don't understand what you mean with page. Since I was going to format and reinstall Windows, I did a backup. Then when Windows was installed, I restored the .reg. Everything was in place except the Avisynth script. But the script itself is in the .reg, in avisynthScript, but with those ////. It'll work with them, the only real problem here is that you lose the script when restoring.

[haruhiko_yamagata]

I don't understand what you mean with page. Since I was going to format and reinstall Windows, I did a backup. Then when Windows was installed, I restored the .reg. Everything was in place except the Avisynth script. But the script itself is in the .reg, in avisynthScript, but with those ////. It'll work with them, the only real problem here is that you lose the script when restoring.
If you type script into the editbox and go to for example resize page without pressing apply, the script is lost. Probably this is not the case for you.
I can't reproduce though. It works for me.

[STaRGaZeR]

If you type script into the editbox and go to for example resize page without pressing apply, the script is lost. Probably this is not the case for you.
I can't reproduce though. It works for me.

That happens to me too. No big deal if you ask me.
Regarding my issue, if you say that avisynthScriptMULTI_SZ is used now instead of avisynthScript, avisynthScriptMULTI_SZ only shows hex(7):00,00 in the .reg, but it's fine in the registry (http://i41.tinypic.com/1gn0bs.jpg). This means empty, right? If that's true then the script is not being exported to the .reg file, like clsid said, therefore it can't be restored later on. I'm using the "Export all settings (http://i44.tinypic.com/1zxusfr.jpg)" button, not the regedit.exe /s /a "c:\ffdshow.reg" "HKEY_CURRENT_USER\Software\GNU\ffdshow" command.

Have you been able to reproduce this (http://forum.doom9.org/showthread.php?p=1299161#post1299161)?

[Snowknight26]

http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/ffmpeg-mt/libavcodec/thread.h?r1=3018&r2=3017&pathrev=3018

I find it kind of ironic that equivalent was fixed while separate wasn't.

[albain]

I found why autocrop wasn't working for me. Resize & aspect "Process whole image" checkbox should be disabled, otherwise it won't work.


Okay, there is something that is not working correctly, I don't know why, working on it...

[STaRGaZeR]

That happens to me too. No big deal if you ask me.
Regarding my issue, if you say that avisynthScriptMULTI_SZ is used now instead of avisynthScript, avisynthScriptMULTI_SZ only shows hex(7):00,00 in the .reg, but it's fine in the registry (http://i41.tinypic.com/1gn0bs.jpg). This means empty, right? If that's true then the script is not being exported to the .reg file, like clsid said, therefore it can't be restored later on. I'm using the "Export all settings (http://i44.tinypic.com/1zxusfr.jpg)" button, not the regedit.exe /s /a "c:\ffdshow.reg" "HKEY_CURRENT_USER\Software\GNU\ffdshow" command.

Look at what happens when using the regedit.exe command:

"avisynthScriptMULTI_SZ"=hex(7):23,4c,6f... ...31,29,00,00

So I guess it's confirmed that the problem is in the "export all settings" button.

Okay, there is something that is not working correctly, I don't know why, working on it...

Nice. The progressive cropping effect is really necessary? It can be really annoying when dealing with huge bands. Also an option for setting max cropping values could be useful.

[haruhiko_yamagata]

Look at what happens when using the regedit.exe command:

"avisynthScriptMULTI_SZ"=hex(7):23,4c,6f... ...31,29,00,00

So I guess it's confirmed that the problem is in the "export all settings" button.

Yes, but I still cannot reproduce...

[STaRGaZeR]

Yes, but I still cannot reproduce...

I dunno, but I'm not the only one experiencing it.

[leeperry]

this problem comes and goes, maybe some specific characters make the whole thing flip out..anyway, regedit does the job :)

[Barleyman]

I got a question. Maybe it's a stupid one but bear with me.

There's ffdshow-mt for multicore processors, fine. And there's ffdshow-x64 for 64bit vista/w7, ok. What about 64-bit mt-build? Doesn't exist?

[clsid]

There is no ffdshow-mt. There is ffmpeg-mt, which is included in ffdshow, in both 32 and 64-bit builds.

[Barleyman]

There is no ffdshow-mt. There is ffmpeg-mt, which is included in ffdshow, in both 32 and 64-bit builds.

So this one is just a trick (http://sourceforge.net/project/showfiles.php?group_id=173941&package_id=214245&release_id=523000)? :confused:

[Eragon4ever]

No, it's an old version from before ffdshow and ffdshow-mt where combined.

[Barleyman]

No, it's an old version from before ffdshow and ffdshow-mt where combined.

So in other words I should be using latest clsid or icl10 build instead? clsid builds do not use SSE?

Maybe that ffdshow-mt build should be moved to old builds as it's pretty confusing.

[clsid]

All builds use MMX/SSE/SSE2/SSSE3/3dnow!/etc. Just only when your CPU supports those extensions. The ICL10 SSE build requires SSE. The additional optimizations in the ICL10 build affect only a few filters and don't have an impact on decoding. Decoding performance is equal for all builds. I advice to use the generic builds, as those get updated more often.

[fuzz!]

Have you chaps ever considered some sort of 'room correction' capability? Such like what high-end receivers do.

Almost every PC has a microphone jack, and all processing can be done in software.

I'm not sure exactly what the software in the receiver does in terms of delays, but it would be awesome just to EQ channels to match the corresponding speakers' frequency response capabilities.

[clsid]

If you want a separate EQ for each channel then you should try out AC3Filter 1.61b.

[Blight]

Crash Bug Report, here's how to reproduce (tested with build 3008):
Play a video with subtitles enabled and a sub file rendering (tested with .SRT subs), then switch from "search in" to "file" while the video is playing and ... crash.

[Blight]

Another crash report (build 3008 again), possibly related to my previous post, here's how to replicate:

1. Play a video.
2. Seek within the timeline and within 1-2 seconds switch to a different subtitle file.

I'm not sure if the crash is due to the quick sub file change after seeking, or perhaps because a different language subtitle is currently visible (actual sub line visible). I believe one of the two causes the crash.

I'm switching the subs programmatically using "IffDecoder :: putParamStrW"

To ease in debugging this issue, here's the latest Zoom Player internal build (http://www.inmatrix.com/temp/zplayer_nct27.zip).
This ZIP file contains updated files and not the full build, you should have ZP v6 installed and overwrite with the files from the ZIP. You also need to set ffdshow as the preferred subtitle renderer (adv. options / playback / video / subtitles).

To test, you need this file structure:
MyVideo.avi
MyVideo.sub1.srt
MyVideo.sub2.srt

Press Ctrl+B to cycle through the subtitle tracks, ffdshow crashes quite easily after this (but not always).

[Leak]

Is it just me, or has SourceForge all of a sudden turned into Angry Fruit Salad?

(And the ffdshow project page now offers the beta 3 off the "Older releases" package as the main download - I'd hazard a guess that's probably because it's the newest release with an i386 processor type; I've added that the beta 6 release, but I still can't set it as the primary download - WTF?)

np: Jared Emerson-Johnson - Ice Station Santa Cutscene Suite (Sam & Max Season Two OST (Disc 1))

[MatMaul]

@haruhiko_yamagata :
http://forum.doom9.org/showthread.php?p=1302094#post1302094

[fastplayer]

(And the ffdshow project page now offers the beta 3 off the "Older releases" package as the main download - I'd hazard a guess that's probably because it's the newest release with an i386 processor type; I've added that the beta 6 release, but I still can't set it as the primary download - WTF?)
Welcome to the club! I tried yesterday for like half an hour and gave up...

[issa]

Here is the latest gcc 4.4.1 with "Graphite loop transformation infrastructure" for mingw32
url: http://rapidshare.com/files/251274698/mingw32-gcc-4.4.1.tar.lzma.html

[albain]

Hi,

@clsid,haruhiko or any project admin

I have still the ability to commit SVN modifications but I can no longer manage the download page from sourceforge.

Thanks

Damien

[clsid]

With the recent SourceForge changes, some things are just hard to find. It took me a while to find the file manager as well. This link should work for you:
https://sourceforge.net/project/admin/editpackages.php?group_id=173941

[albain]

With the recent SourceForge changes, some things are just hard to find. It took me a while to find the file manager as well. This link should work for you:
https://sourceforge.net/project/admin/editpackages.php?group_id=173941

Thanks, it works

[hayan]

r3009 or above, play movie on kirikiri2 get a black screen, if the video format is h264.avi(fourcc is h264).

[haruhiko_yamagata]

r3009 or above, play movie on kirikiri2 get a black screen, if the video format is h264.avi(fourcc is h264).
What is kirikiri2?

[hayan]

What is kirikiri2?
http ://en.wikipedia.org/wiki/KiriKiri

[Mercury_22]

When playing a 5.1 channel LPCM audio (inside a m2ts) the BackR channel it's wrong mapped to LFE or it's missing (I'm not sure)!
Tested with MPC-HC and FFD rev3028
MPC-HC's internal LPCM it's playing the same file with no problem !

Sample http://sharebee.com/c778a520

[Blight]

haruhiko:
Was the recent FLV patch integrated into any recent builds?

[haruhiko_yamagata]

haruhiko:
Was the recent FLV patch integrated into any recent builds?The patch is working for me. I'm testing it for more formats.

[haruhiko_yamagata]

MatMaul's patch use VIDEOINFOHEADER::rcSource instead of BITMAPINFOHEADER::biHeight.
These two things are basically the same things. Why FLV splitter does not set BITMAPINFOHEADER::biHeight correctly?
We can workaround this problem using his patch, but it's a bit hard to check the compatibility problems with all other splitters.

[MatMaul]

MatMaul's patch use VIDEOINFOHEADER::rcSource instead of BITMAPINFOHEADER::biHeight.
I think BITMAPINFOHEADER::biHeight represents the real height of the video (which is mod16) and VIDEOINFOHEADER::rcSource the rectangle of the video to display to the user.
however if this is not the case the flv splitter can be fixed easily instead of this patch.

[haruhiko_yamagata]

I think BITMAPINFOHEADER::biHeight represents the real height of the video (which is mod16) and VIDEOINFOHEADER::rcSource the rectangle of the video to display to the user.
however if this is not the case the flv splitter can be fixed easily instead of this patch.
I think it is safer to change flv splitter rather than ffdshow.
Though I don't notice any real inconvenience with your patch.

[clsid]

The FLV splitter is open-source. So changing that is no problem. I have write access to MPC-HC and Guliverkli2 SVN so applying a patch is no problem.

http://mpc-hc.svn.sourceforge.net/viewvc/mpc-hc/trunk/src/filters/parser/flvsplitter/

[MatMaul]

I think it is safer to change flv splitter rather than ffdshow.
Though I don't notice any real inconvenience with your patch.
I disagree with you on this one.
according to this MSDN page :
http://msdn.microsoft.com/en-us/library/dd377618%28VS.85%29.aspx
my previous explanation seems to be good.
so this is not a bug in the splitter which correctly does its job but in ffdshow.
set the rectangle infos (rcSource) in the buffer infos (BITMAPINFOHEADER) in the splitter could break other decoders which rely on the buffer infos to decode the stream.

[STaRGaZeR]

Guys, I'm getting a buttload of errors when compiling ffmpegmt for x64, in pthread.c:

http://i32.tinypic.com/2mp0vhg.png

They don't make any sense. No errors when compiling for x86, like it should be. Any suggestions?

[clsid]

You need to put pthread headers/libs in your MinGW64 installation.

[haruhiko_yamagata]

I disagree with you on this one.
according to this MSDN page :
http://msdn.microsoft.com/en-us/library/dd377618%28VS.85%29.aspx
my previous explanation seems to be good.
so this is not a bug in the splitter which correctly does its job but in ffdshow.
set the rectangle infos (rcSource) in the buffer infos (BITMAPINFOHEADER) in the splitter could break other decoders which rely on the buffer infos to decode the stream.
OK, then please commit the patch.

[MatMaul]

OK, then please commit the patch.
done with a little modification to also handle theora cases.

[STaRGaZeR]

You need to put pthread headers/libs in your MinGW64 installation.

Done, but now I get this:

http://thumbnails9.imagebam.com/4194/fa59f441931332.gif (http://www.imagebam.com/image/fa59f441931332)

[clsid]

I have noticed a few issues with the patch.

1) The cropping does not work properly when using the Overlay Mixer. The previously blurry parts now show whatever is visible in the background at the time the playback starts.
2) The "Output size" value of the OSD reports the original size.

Edit: it also doesn't work with VMR-7.


@STaRGaZeR, maybe you need a different build of pthread.a. Sorry, I don't have time to help you further.

[STaRGaZeR]

The only one I know of is the one present here: http://sourceforge.net/projects/mingw-w64/files/

Which I've already tried with the same results :(

[mark0077]

Guys, been watching alot of old content recently that needs to be de-interlaced. I have found moving to hardware de-interlacing is giving me alot nicer de-interlacing than yadif so I am happy enough for the moment outputting nv12 from ffdshow.

My problem again comes back to DVD's that are badly marked as coming from interlaced sources (most of my DVDs at least just need simple weave to 24fps). My question is (as there isn't a good software fix /detector for this yet), is there a way to change the interlaced flag of each frame with one setting in ffdshow. I need this because for those disks that are badly marked, nvidia de-interlaces so I get 48fps instead of 24, aswell as slight artifacts when the nvidia de-interlacer messes with the frames. I would much rather the 24 original progressive frames that are there.

I have a script that I made months ago that is called from xbmc when I start a movie / tv show, which looks up a text file containing content that 'really' needs to be de-interlaced. Previously using software de-interlacing, I made this script write the ffdshow de-interlacing setting in the registry to on or off, (yadif on or off), but now that I am outputting the interlaced flag to the hardware de-interlacer, this main ffdshow de-interlacing option no longer can work. I am in search of another setting that will let me force each frame to be detected as interlaced or progressive instead of Auto, and effect everthing it should within ffdshow.

I suppose I could change the "Method:" in the "Output" tab in ffdshow, to "Force Weave" or "Force Bob", but would this mean that other processes within ffdshow would be effected?? Mainly what I ask here is, is good yv12 -> nv12 conversion dependent on proper interlaced flag detection.... because I see conversion to YV12 -> YUY2 needs this "Interlace or progressive" option to be set correctly in the "RGB Conversion" area, but I am unsure if it also needs to be set for YV12 -> NV12 only. If not I am hoping I can just find the registry setting for "Output" -> "Method", and change this using my script to have some sort of automation.

EDIT: Actually looking closely at some scenes, yadif actually does a much better job than outputting nv12 to nvidia... I wonder do I need to set something special to use NVidias best de-interlacing method. Control panel options are limited to one setting "Use inverse telicine" which I have enabled. Doing a frame step on some scenes that show it off, and nvidia does an absolutely terrible job at some scenes outputting both yuy2 and nv12... I'm confused now... I am in Windows 7 with latest 186.24 drivers.

[Mercury_22]

E-AC3 (Dolby Digital Plus) channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

= No Side Channels

Test 7.1 E-AC3 (Dolby Digital Plus) (http://sharebee.com/69772daa)

Dolby TrueHD channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - REAR RIGHT wrong mapped to SIDE RIGHT
3 - SIDE LEFT wrong mapped to REAR LEFT
4 - REAR LEFT wrong mapped to SIDE LEFT

Test 7.1 Dolby TrueHD (http://www.megaupload.com/?d=7P9P81CW)

[mark0077]

E-AC3 (Dolby Digital Plus) channels are wrong mapped :
1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT
= No Side Channels
Test file http://sharebee.com/69772daa

Dolby TrueHD channels are wrong mapped :
1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - REAR RIGHT wrong mapped to SIDE RIGHT
3 - SIDE LEFT wrong mapped to REAR LEFT
4 - REAR LEFT wrong mapped to SIDE LEFT

Test file http://www.megaupload.com/?d=7P9P81CW

I'm hoping this is the key to my HDAV Deluxe problems, rear right and rear left sound like they are getting blurred together with ac3 5.1 surround sound tests. BUT it also happens for wma 5.1 files. Also using reclock to output in wasapi exclusive mode this rear blurring effect seems to disappear.

If someone confirms this is a bug in mpc-hc, could it apply to all rear outputs, and only in directsound mode? I get correct ac3 5.1 channels using wasapi out but not directsound.

[Mercury_22]

I'm hoping this is the key to my HDAV Deluxe problems, rear right and rear left sound like they are getting blurred together with ac3 5.1 surround sound tests. BUT it also happens for wma 5.1 files. Also using reclock to output in wasapi exclusive mode this rear blurring effect seems to disappear.

If someone confirms this is a bug in mpc-hc, could it apply to all rear outputs, and only in directsound mode? I get correct ac3 5.1 channels using wasapi out but not directsound.

ac3 5.1 surround sound test (posted by me on MPC-HC's thread) works fine with FFD ! (it's not working with MPC-HC's INTERNAL filters ! )

P.S. I might be wrong but I think you meant to post this on MPC-HC's thread not on FFD's ? ( MPC's INTERNAL filters wrong mapped channels are different from FFD's wrong mapped channels)

[travolter]

In the codec tab I set all decoders to use libavcodec.

Ev3rything OK .. I play WMV/VC-1/MKV/MPG/MOV/MP4 ok... but when I load a XVID/DIVX.. the performance is terribly bad.. I lose frames when I use same filters that are working ok in 1080VC-1 videos..

I have a Quad core CPU and I notice that only one core is really boosted..


There is any way to fix that?

Maybe installing some compatible Xvid/divx dshow codec?


EDIT.- I solved the problem installing Haali splitter.. and enabling the option "decoder priority boost"
(any fix if you dont want to have Haali installed?)

[MatMaul]

I have noticed a few issues with the patch.

1) The cropping does not work properly when using the Overlay Mixer. The previously blurry parts now show whatever is visible in the background at the time the playback starts.
2) The "Output size" value of the OSD reports the original size.

Edit: it also doesn't work with VMR-7.
... microsoft seems to do not follow their own recommendation in their old renderers...
same "problem" appears with mpc internal decoder.
I will not try to fix that because I don't really care about those old renderers (and I have no idea how to fix that without doing a real crop of the output video buffer).

[clsid]

Considering that in absolute numbers, probably more people use the old renderers than the new ones, a fix would be very welcome. If that would require ffdshow to do the cropping instead of the renderer, then so be it. If you don't want to do it, then maybe Haruhiko can once he has some time.

[Mercury_22]

LPCM channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to SIDE LEFT
2 - SIDE LEFT wrong mapped to LFE
3 - LFE wrong mapped to SIDE RIGHT

Test 7.1 LPCM (http://www.megaupload.com/?d=0HM2JCEP)

[Amour]

Today I did the monthly Windows Update. (Vista)
Result ?
I got big sound issues: there is a gap of 2 seconds between sound and video.
I will try to re-install ffdshow.

[edit]
re-installing ffdshow doesn't work!
I think there is a big issue with my computer now, because I can't watch videos normally anymore. :(

[Amour]

I tested different players, and the issue only came with MPC-HC.
Sorry, it might not be an ffdshow issue after all.
I just wonder if it is related to this:
http://blogs.technet.com/msrc/archive/2009/07/08/july-2009-advance-notification.aspx

[edit]
I fixed it!
I manually deleted "HKEY_CURRENT_USER\Software\Gabest" and it did the trick. :)
Sorry for bothering you with something not related to ffdshow.

[liquidskin76]

Hi,

Just after an update regarding DTS-HD and TrueHD/DTS-HD passthrough.

Is there any new news on DTS-HD decoding (not DTS core passthrough)? Is this being developed? I read somewhere that it was pretty close however not sure if true?

Regarding TrueHD/DTS-HD passthrough (bitstreaming), am i right in thinking that as long as the source files are decypted (ie. using AnyDVD HD), PAP does not play a part and the passthrough is theoretically possible? Is this being developed for FFDShow to work alongside AC3/DTS passthrough (please god, say yes!!!). I believe SlySoft are hoping to implement bitstreaming of HD audio over an HDMI port with their new upcoming SlyPlayer.

Cheers

[Snowknight26]

If whatever you're passing TrueHD through to can decode TrueHD, then it can also decode LPCM. So why wouldn't you just have ffdshow decode TrueHD?

[liquidskin76]

If whatever you're passing TrueHD through to can decode TrueHD, then it can also decode LPCM. So why wouldn't you just have ffdshow decode TrueHD?

I know what you mean, it's just that i'd love to see TrueHD and DTS-HD light up on the av receiver!!

In the end i geuss i'd be happy with just DTS-HD decoding (as well as the TrueHD decoding already in place).

I've tried bitstreaming with the ASUS Xonar HDAV 1.3 and Arcsoft TMT3 however it's a mess of a setup!

Cheers

[Mercury_22]

Is it normal to have DTS SIDE RIGHT mapped to REAR RIGHT and SIDE LEFT mapped to REAR LEFT (is it DTS-HD HR or MA working)?

Test 7.1 DTS-HD HR (http://sharebee.com/abf92989)

[mark0077]

Mercury, because I have had trouble with both rear outputs with my soundcard and setup (some form of fuzzy mixing between the two) you have sparked my curiosity. I post here because I notice in ffdshow, in the mixer section, the "5 channels" output configuration uses Back L and Back R, where I would have always assumed the rear channels of a 5.1 track would be Side L and Side R. I have my rear speakers setup as per dolby recommendations, at the side, slightly to the rear.

So confused now as to what each piece of the chain intends, the original track, the software like ffdshow, and finally the drivers.

[Thunderbolt8]

whats that ffmpeg page again where I can also make feature requests?

[Mercury_22]

Mercury, because I have had trouble with both rear outputs with my soundcard and setup (some form of fuzzy mixing between the two) you have sparked my curiosity. I post here because I notice in ffdshow, in the mixer section, the "5 channels" output configuration uses Back L and Back R, where I would have always assumed the rear channels of a 5.1 track would be Side L and Side R. I have my rear speakers setup as per dolby recommendations, at the side, slightly to the rear.

So confused now as to what each piece of the chain intends, the original track, the software like ffdshow, and finally the drivers.

I think FFD'S settings are correct I don't know why SOME drivers are using SIDE instead of REAR that's why I recommend you to use 7.1 in windows sound panel even with 5.1 speakers (not so many tracks with 7.1) and you can use window's audio control panel to remap side to rear & front (just uncheck "side pair" in Vista's or Windows 7's "configure") if you have any 7.1 soundtrack

[albain]

I didn't test the sample yet, do we have a problem here on channels mapping ? It may be a problem with the channel mapping between ffmpeg and FFDShow. FFMpeg has recently integrated a channel remapper. I have fixed this already for some formats but they may have remap channels for other formats since.

Concerning DTS-HD decoding (MA or not), we depend on FFMpeg project.
However, DTS-HD format is much more complicated than DolbyTrueHD to decode, and of course there is no white paper on it.

About HDMI passthrough for DTS-HD/TrueHD/LPCM, I don't have the equipment (nor the OS) to test it : is this working for any of those ?
I guess that the work to do on FFDShow would be to encapsulate the decoded stream into LPCM, especially if this is decoded TrueHD or decoded DTS-HD

I know that Microsoft (on windows 7 only) brought some modifications to the directshow structures (to WAVEFORMATEXTENSIBLE) in order to deal with this.

[Peekstra]

Although ffdshow functions fine with most DVD menu's still some of them don't work as they should.

For example, the msmpeg decoder works fine here as it only shows the title where the mouse pointer hovers above it. But with ffdshow navigation is impossible because you can't see which movie has been selected.

Using Zoom Player, MS DVD navigator & ffdshow 3020.

[Mercury_22]

I didn't test the sample yet, do we have a problem here on channels mapping ? It may be a problem with the channel mapping between ffmpeg and FFDShow. FFMpeg has recently integrated a channel remapper. I have fixed this already for some formats but they may have remap channels for other formats since.

Concerning DTS-HD decoding (MA or not), we depend on FFMpeg project.
However, DTS-HD format is much more complicated than DolbyTrueHD to decode, and of course there is no white paper on it.

About HDMI passthrough for DTS-HD/TrueHD/LPCM, I don't have the equipment (nor the OS) to test it : is this working for any of those ?
I guess that the work to do on FFDShow would be to encapsulate the decoded stream into LPCM, especially if this is decoded TrueHD or decoded DTS-HD

I know that Microsoft (on windows 7 only) brought some modifications to the directshow structures (to WAVEFORMATEXTENSIBLE) in order to deal with this.

Yes we have :

E-AC3 (Dolby Digital Plus) channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

= No Side Channels

Test 7.1 E-AC3 (Dolby Digital Plus) (http://sharebee.com/69772daa)

Dolby TrueHD channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - REAR RIGHT wrong mapped to SIDE RIGHT
3 - SIDE LEFT wrong mapped to REAR LEFT
4 - REAR LEFT wrong mapped to SIDE LEFT

Test 7.1 Dolby TrueHD (http://www.megaupload.com/?d=7P9P81CW)

LPCM channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to SIDE LEFT
2 - SIDE LEFT wrong mapped to LFE
3 - LFE wrong mapped to SIDE RIGHT

Test 7.1 LPCM (http://www.megaupload.com/?d=0HM2JCEP)

Is it normal to have DTS SIDE RIGHT mapped to REAR RIGHT and SIDE LEFT mapped to REAR LEFT (is it DTS-HD HR or MA working)?

Test 7.1 DTS-HD HR (http://sharebee.com/abf92989)

All tested on a 7.1 config, cause the 5.1 config in windows sound scheme has most of the time (depending of the driver) SIDE instead of REAR channels.

I'm sure they are wrong mapped cause the same files with other filters (even MPC-HC's for TrueHD) are playing the correct channel on my system

Although the last one, DTS-HD I think it's "normal" do to no DTS-HD HR or MA decoding, I guess

[tetsuo55]

I posted about correct channel order here:
Please use this information when fixing the problem

http://forum.doom9.org/showthread.php?p=1306597#post1306597

[Octo-puss]

I've got this weird clip with audio being reported as:

ID : 1
Format : ADPCM
Format profile : U-Law
Codec ID : ulaw
Duration : 1mn 43s
Bit rate mode : Constant
Bit rate : 256 Kbps
Channel(s) : 1 channel
Sampling rate : 32.0 KHz
Resolution : 16 bits


I can't get it to play at all. Is it supported in FFDS at all?

[poisondeathray]

I've got this weird clip with audio being reported as:

ID : 1
Format : ADPCM
Format profile : U-Law
Codec ID : ulaw
Duration : 1mn 43s
Bit rate mode : Constant
Bit rate : 256 Kbps
Channel(s) : 1 channel
Sampling rate : 32.0 KHz
Resolution : 16 bits


I can't get it to play at all. Is it supported in FFDS at all?

Did you try enabling different adpcm choices in the ffdshow configuration? (they are disabled by default)

[Thunderbolt8]

whats that ffmpeg page again where I can also make feature requests?
^^ ???

thanks!

[Jeremy Duncan]

Please add this patch to the ffdshow svn: link (http://www.mediafire.com/?tmdwym5nnoo)

I got the patch from SEt: link (http://forum.doom9.org/showthread.php?t=148117)

I need a ffdshow generic build made that has this patch available.
The sooner the better. many thanks! :)

[leeperry]

Please add this patch to the ffdshow svn: link (http://www.mediafire.com/?tmdwym5nnoo)
hummm, that patch makes some of my movies run at half speed.

it'd be wiser to wait for Leak to watch over the problem(if there's any) whenever he has time I think....pushing dirty hacks into the SVN doesn't sound too wise :o

[Octo-puss]

Did you try enabling different adpcm choices in the ffdshow configuration? (they are disabled by default)

Yes I did, there is only libavcodec available anyway.

[lych_necross]

hummm, that patch makes some of my movies run at half speed.

it'd be wiser to wait for Leak to watch over the problem(if there's any) whenever he has time I think....pushing dirty hacks into the SVN doesn't sound too wise :o
I second that. AFAIK, Leak is already looking in to this issue.

[Blight]

Possible API Bug Report:
I believe that "IDFF_subCurrentFlnm" value determines whether ffdshow has found a suitable subtitle stream for rendering.
The problem is, this value is not set when an embedded sub stream is displayed, so there's no way of knowing whether ffdshow had found a valid subtitle stream and is ready to display it.
Is it possible to enable this flag when the embedded subtitle pin is connected?

[albain]

I fixed the channels order for Dolby TrueHD.

For LPCM, I will need Haruhiko's hand : channels order is wrong
Side left is sent to side right
Side left is sent to LFE
LFE is sent to side left.

Problem is located in TaudioUncompressed.cpp, line 100
if (lpcm24) {
if (highDef)
*dst++ = (src[i]<<24) + (src[i+ 1]<<16) + (src[i+ 2]<<8);
*dst++ = (src[i+3]<<24) + (src[i+ 4]<<16) + (src[i+ 5]<<8);
*dst++ = (src[i+6]<<24) + (src[i+ 7]<<16) + (src[i+ 8]<<8);
*dst++ = (src[i+9]<<24) + (src[i+10]<<16) + (src[i+11]<<8);
{

[Mercury_22]

I fixed the channels order for Dolby TrueHD.

For LPCM, I will need Haruhiko's hand : channels order is wrong
Side left is sent to side right
Side left is sent to LFE
LFE is sent to side left.

Problem is located in TaudioUncompressed.cpp, line 100
if (lpcm24) {
if (highDef)
*dst++ = (src[i]<<24) + (src[i+ 1]<<16) + (src[i+ 2]<<8);
*dst++ = (src[i+3]<<24) + (src[i+ 4]<<16) + (src[i+ 5]<<8);
*dst++ = (src[i+6]<<24) + (src[i+ 7]<<16) + (src[i+ 8]<<8);
*dst++ = (src[i+9]<<24) + (src[i+10]<<16) + (src[i+11]<<8);
{

Maybe I'm wrong but for LPCM I think you meant

LPCM channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to SIDE LEFT
2 - SIDE LEFT wrong mapped to LFE
3 - LFE wrong mapped to SIDE RIGHT

Test 7.1 LPCM (http://www.megaupload.com/?d=0HM2JCEP)
Anyway thanks for TrueHD I can't wait to test this :thanks:

[albain]

Yes that's correct

[Mercury_22]

It's my understanding that libdts has been upgrade / replace with libdca, (http://www.videolan.org/developers/libdca.html) am I wrong ? If I'm not shouldn't libdts be update ?

@albain Can you Please compile a x64 for testing TrueHD

[albain]

Possible API Bug Report:
I believe that "IDFF_subCurrentFlnm" value determines whether ffdshow has found a suitable subtitle stream for rendering.
The problem is, this value is not set when an embedded sub stream is displayed, so there's no way of knowing whether ffdshow had found a valid subtitle stream and is ready to display it.
Is it possible to enable this flag when the embedded subtitle pin is connected?

No, this value has been disabled a while ago

To get the current subtitles, instead of calling the sendmessage method with WPRM_GETPARAM and "IDFF_subCurrentFlnm",
call it with MSG_GET_CURRENT_SUBTITLES and 0

[MatMaul]

It's my understanding that libdts has been upgrade / replace with libdca, (http://www.videolan.org/developers/libdca.html) am I wrong ? If I'm not shouldn't libdts be update ?
libdts = libdca
and it has not been updated since 2006 or so.

[tetsuo55]

isn't libdca deader than a doornail?

I unberstood it the other way around, libdca got dropped in favor of libavcodec.

[clsid]

Not all software requires endless development. Certainly not small projects like libdca that have a fixed specification.

[leeperry]

libdts = libdca
and it has not been updated since 2006 or so.
yep, there's no DTS96/24 GPL decoder :(

[tetsuo55]

Not all software requires endless development. Certainly not small projects like libdca that have a fixed specification.True for a lot of applications, DTS has however evolved and libdca never catched up.

[clsid]

Those should be considered new formats, not plain old DTS. And those are also not yet supported by libavcodec or any other open source library.

[Mercury_22]

I fixed the channels order for Dolby TrueHD.

For LPCM, I will need Haruhiko's hand : channels order is wrong
Side left is sent to side right
Side left is sent to LFE
LFE is sent to side left.

Problem is located in TaudioUncompressed.cpp, line 100
if (lpcm24) {
if (highDef)
*dst++ = (src[i]<<24) + (src[i+ 1]<<16) + (src[i+ 2]<<8);
*dst++ = (src[i+3]<<24) + (src[i+ 4]<<16) + (src[i+ 5]<<8);
*dst++ = (src[i+6]<<24) + (src[i+ 7]<<16) + (src[i+ 8]<<8);
*dst++ = (src[i+9]<<24) + (src[i+10]<<16) + (src[i+11]<<8);
{

I've just tested your fix for TrueHD with clsid's ffdshow_rev3040_20090724_clsid and there is NO change :confused:
E-AC3 (Dolby Digital Plus) channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

= No Side Channels

Test 7.1 E-AC3 (Dolby Digital Plus) (http://sharebee.com/69772daa)

Dolby TrueHD channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - REAR RIGHT wrong mapped to SIDE RIGHT
3 - SIDE LEFT wrong mapped to REAR LEFT
4 - REAR LEFT wrong mapped to SIDE LEFT

Test 7.1 Dolby TrueHD (http://www.megaupload.com/?d=7P9P81CW)

[tetsuo55]

Those should be considered new formats, not plain old DTS. And those are also not yet supported by libavcodec or any other open source library.What i read about DTS states that all formats are plain DTS with extra data. This would mean that it's not a new format but an extention of a existing one. DTS even states that any version of DTS is fully backwards compatible with older players, the extra data of the newer format is simply discarded.

[clsid]

Logic dictates that it should decode DTS, not some extensions that were made up long after it was developed.

But it all does not matter. It does not support those extensions, neither do any other open source decoders. I can't change that. You can't change that. So we are all wasting our time here.

[tetsuo55]

LBut it all does not matter. It does not support those extensions, neither do any other open source decoders. I can't change that. You can't change that. So we are all wasting our time here.It's the sad truth :(

PS i rewrote my post...

EDIT:

I just checked the history, all the extensions already existed while it was being written, even DTS-HD was released before the last SVN commit. I guess no-one cared or people are afraid of DTS laywers.
It should be pretty easy to reverse engineer all these formats from a hardware decoder chip (easy for people who know how to read a chip)

PS if that last sentence is against the rules let me know and i will remove it.

[clsid]

Two quotes from the VideoLAN website that could explain why the project has not progressed further:
None of the extensions (extra channels, 96/192kHz sample rates) have been implemented (and unfortunately the public standard provides only minimal information on them).
Provisional Warning: DTS Inc. claims that use of libdca software, to decode DTS compressed sound data on a DVD could violate DTS's patent rights. If you are unsure about the legality of using and distributing this code in your country, in particular in the USA, please consult your lawyer before downloading it

It is also possible (and perhaps easier) to reverse engineer a commercial decoder. That is what the FFmpeg devs usually do.

[lych_necross]

I guess no-one cared or people are afraid of DTS laywers.
It should be pretty easy to reverse engineer all these formats from a hardware decoder chip (easy for people who know how to read a chip)

Lets not forget the law of supply and demand. I believe the main reason the DTS stuff hasn't been updated is because there is not a lot of demand for it (it requires special equipment...).

[albain]

I've just tested your fix for TrueHD with clsid's ffdshow_rev3040_20090724_clsid and there is NO change :confused:

The fixes requires mplayer and ffdshow to be recompiled.

Maybe mplayer was not recompiled in this build ?

However I didn't fixed EAC3 (yet), only TrueHD

[albain]

Concerning DTS, patents are dated from 1991, and AFAIK patents are limited in duration (20 years) so there should be any problems to use and extend the decoding of this format next year.

Not the same thing for DTS-HD, more recent.
But can't you post some reversed engineer code in source form only ? (binary form would be prohibited for sure)

[travolter]

"Postprocessing tab"

I have found a bug with "SPP deblocking" when I check the "soft threshold" checkbox.

The bug only happend when (IMPORTANT):
you play an AVI file (XVID codec or similar) + you place postprocessing after "resize" in the effect chain

http://img129.imagevenue.com/loc500/th_25930_2_122_500lo.jpg (http://img129.imagevenue.com/img.php?image=25930_2_122_500lo.jpg)
notice the square at top left. Processing not affect that screen area "592x320".. thats the original screen resolution of my movie.. and I have rescaled to 1280x720

(I know that many people will save me saying.. "place postprocessing before resize".. but sometimes is better for CPU saving use postprocessing after resize.. like when using a 1080 original rescaled to 720)

[Mercury_22]

The fixes requires mplayer and ffdshow to be recompiled.

Maybe mplayer was not recompiled in this build ?

However I didn't fixed EAC3 (yet), only TrueHD

Can you post one ( x64 ) compiled by you then ?

I was refering only to TrueHD too but I quote the entire post because it was quicker :)

EDIT :
I've tested with ffdshow_rev3040_20090725_clsid_icl10 too and still no change. I don't know if this helps but MPC-HC had the correct mapping for TrueHD until rev 1180 !

[leeperry]

BTW, ages ago I asked Harihuko if he could add 1.78/1.85/2.35/2.40 presets in the resize filter...well it's actually 2.39 :
http://www.arrimedia.com/2-Perforation.php
http://www.arrimedia.com/upload_img/arrimedia.com/image_153.jpg

http://en.wikipedia.org/wiki/Aspect_ratio_(image)#Why_16:9.3F
2.39:1 (the ratio of anamorphic widescreen films).

I think you can leave it as "CinemaScope" as this is what is :
Anamorphic widescreen was not used again for cinematography until Twentieth Century-Fox bought the rights to the technique in 1952 to create its CinemaScope widescreen technique.

but 1.85 should be called "US widescreen", 1.67 "EU widescreen" and 2.35 "anamorphic widescreen"...as it's not really CinemaScope :
http://en.wikipedia.org/wiki/Anamorphic_format#2.35.2C_2.39.2C_or_2.40.3F
Anamorphic prints are still often called Scope or 2.35 by projectionists, cinematographers, and others working in the field, if only by force of habit. 2.39 is in fact what they generally are referring to

[Mr VacBob]

notice the square at top left. Processing not affect that screen area "592x320".. thats the original screen resolution of my movie.. and I have rescaled to 1280x720

(I know that many people will save me saying.. "place postprocessing before resize".. but sometimes is better for CPU saving use postprocessing after resize.. like when using a 1080 original rescaled to 720)

Yes, the reason they say that is because it won't work. postproc and spp depend on the compression qscales of the video and they aren't resized with it (it wouldn't work very well if you tried to, either). Instead you have to use some generic avisynth filter or a faster postproc.

[albain]

Can you post one ( x64 ) compiled by you then ?

I was refering only to TrueHD too but I quote the entire post because it was quicker :)

EDIT :
I've tested with ffdshow_rev3040_20090725_clsid_icl10 too and still no change. I don't know if this helps but MPC-HC had the correct mapping for TrueHD until rev 1180 !

I don't have any 64 bits compiler (anymore).

Here is a 32 bits link (http://www.mediafire.com/?sharekey=d2549da476034d8c08f8df73f2072ed6e04e75f6e8ebb871) if you can test it

[Mercury_22]

I don't have any 64 bits compiler (anymore).

Here is a 32 bits link (http://www.mediafire.com/?sharekey=d2549da476034d8c08f8df73f2072ed6e04e75f6e8ebb871) if you can test it

Tested but NO change.
Dolby TrueHD channels are wrong mapped :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - REAR RIGHT wrong mapped to SIDE RIGHT
3 - SIDE LEFT wrong mapped to REAR LEFT
4 - REAR LEFT wrong mapped to SIDE LEFT

Test 7.1 Dolby TrueHD (http://www.megaupload.com/?d=7P9P81CW)

P.S. maybe you can talk with Casimir cause MPC-HC now has the channels for all format correct mapped but you need to set MPC-HC's internal filters to "3 Front + 2 Rear" even if you have 7.1 speakers

[albain]

You're right, either I wrongly tested or I messed up with the patch
I am on it

*EDIT* : this is fixed now. I tested it on TrueHD but it should also work on EAC3. Done in revision 3043
For uncompressed formats (LPCM) this is harder to figure out what's wrong. I'll try but we may have to wait for Haruhiko's return

[Jeremy Duncan]

Please add this patch to the ffdshow svn: link (http://www.mediafire.com/?tmdwym5nnoo)

I got the patch from SEt: link (http://forum.doom9.org/showthread.php?t=148117)

I need a ffdshow generic build made that has this patch available.
The sooner the better. many thanks! :)

To the two fellows who said this patch doesn't work for them. They are using SEt's avisynth and mt dll's.
I suggest they try the patch with the ffdshow version SEt made the patch for, and using my July 23rd avisynth and mt dlls.

Then if it still doesn't work for them then we wait for leak to post his patch, but his two week vacation ends on friday.

I will say the SEt patch works fine for me if I use the ffdshow version SEt made the patch for.

Please add SEt's patch to the ffdshow svn. This is very rough to wait for Leak and how do we know he will add a patch?
Nothing is certain in this life and we have a working patch ready to go. Add the SEt patch and when leak adds his patch remove SEt's version. :)

dlls (http://forum.doom9.org/showthread.php?t=144852)

[clsid]

Leak is our avisynth guru. He will decide if and when to apply the patch. He already said he will look at it.

[Mercury_22]

You're right, either I wrongly tested or I messed up with the patch
I am on it

*EDIT* : this is fixed now. I tested it on TrueHD but it should also work on EAC3. Done in revision 3043
For uncompressed formats (LPCM) this is harder to figure out what's wrong. I'll try but we may have to wait for Haruhiko's return

Can you please, compile ?

[XhmikosR]

Can you please, compile ?

Here (http://www.mediafire.com/?uzi104ynzdj) is r3044. I built it yesterday for personal use.

[Mercury_22]

Here (http://www.mediafire.com/?uzi104ynzdj) is r3044. I built it yesterday for personal use.
:thanks:

Can you please add x64 too ?

You're right, either I wrongly tested or I messed up with the patch
I am on it

*EDIT* : this is fixed now. I tested it on TrueHD but it should also work on EAC3. Done in revision 3043
For uncompressed formats (LPCM) this is harder to figure out what's wrong. I'll try but we may have to wait for Haruhiko's return

The mapping for the TrueHD it's now correct !:)
The mapping for E-AC3 still incorrect :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

Same as MPC-HC

[XhmikosR]

:thanks:

Can you please add x64 too ?



The mapping for the TrueHD it's now correct !:)
The mapping for E-AC3 still incorrect :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

Same as MPC-HC
Unfortunately no, at least for now.

[leeperry]

I will say the SEt patch works fine for me if I use the ffdshow version SEt made the patch for.

Please add SEt's patch to the ffdshow svn.
I'm using the "official" 2.57 MT files, as all my plugins were compiled for this version(that works fine BTW)...even SEt said that this patch was a dirty hack, hence he didn't want to push it into the SVN.

this patch made some of my movies run at half speed..

any "fix"(if required) should work w/ whatever 2.57 or 2.58....Leak is the judge here, anyway you got SEt's fffdshow.ax for the time being....what's wrong w/ that?

[albain]

:thanks:

Can you please add x64 too ?



The mapping for the TrueHD it's now correct !:)
The mapping for E-AC3 still incorrect :

1 - SIDE RIGHT wrong mapped to REAR RIGHT
2 - SIDE LEFT wrong mapped to REAR LEFT

Same as MPC-HC

Unfortunately, the problem does not come from FFDShow but ffmpeg : the EAC3 decoder does not support 7.1 channels yet (6 channels max).
As a result, back and side channels are the same.

Please report the problem to ffmpeg team.

*EDIT* : this fixed for LPCM highdef too now (revision 3050) . Don't have the time to release a build

[Mercury_22]

Unfortunately, the problem does not come from FFDShow but ffmpeg : the EAC3 decoder does not support 7.1 channels yet (6 channels max).
As a result, back and side channels are the same.

Please report the problem to ffmpeg team.

*EDIT* : this fixed for LPCM highdef too now (revision 3050) . Don't have the time to release a build

Nice can't wait to test it !
Maybe someone can find the time to release a build (x64 too please)

[XhmikosR]

Download (http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c) ffdshow_rev3050_20090801 and ffdshow_rev3050_20090801_icl11
(the link will always point to a folder)

Note: I include theora in my builds, although most people won't use it and it might be an old version of theora. Maybe it will get updated sometime in the future, but some people might want to use it.

Unfortunately I haven't managed to build the x64 version.

If someone knows how to parse some arguments to the Inno Setup compiler via cmd, that would help me a lot. E.g.I'd like to change a define from true to false in order to build first the normal build and after that the ICL build of the installer. Now I have to do it manually.

[Mercury_22]

Download (http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c) ffdshow_rev3050_20090801 and ffdshow_rev3050_20090801_icl11
(the link will always point to a folder)

Note: I include theora in my builds, although most people won't use it and it might be an old version of theora. Maybe it will get updated sometime in the future, but some people might want to use it.

Unfortunately I haven't managed to build the x64 version.

If someone knows how to parse some arguments to the Inno Setup compiler via cmd, that would help me a lot. E.g.I'd like to change a define from true to false in order to build first the normal build and after that the ICL build of the installer. Now I have to do it manually.

:thanks:

P.S. I'm hoping you'll manage to build a x64 too

Unfortunately, the problem does not come from FFDShow but ffmpeg : the EAC3 decoder does not support 7.1 channels yet (6 channels max).
As a result, back and side channels are the same.

Please report the problem to ffmpeg team.

*EDIT* : this fixed for LPCM highdef too now (revision 3050) . Don't have the time to release a build

I can confirm LPCM now has the channels correct mapped

[Gleb Egorych]

I can confirm LPCM now has the channels correct mapped
Does it mean that channel mapping problems in ffdshow are now fixed? Regardless of the problems causes by decoders themselves (libdts for DTS and libavcodec for E-AC3 which are limited to 6 channels).

[Mercury_22]

Does it mean that channel mapping problems in ffdshow are now fixed? Regardless of the problems causes by decoders themselves (libdts for DTS and libavcodec for E-AC3 which are limited to 6 channels).
Yes with those exceptions

[albain]

What is the problem with libdts ?

This problem also occurs with FFMPeg DCA (=dts) decoder ?

[avivahl]

clsid, could you please compile and release rev 3051 for both generic and x64 builds on SourceForge?

[ipanema]

In June I mentioned this:


Most of the time ffdshow's H.264 decoder does not seem to decode the 2 B frames that preceed the first I frame (in presentation order) in the stream. It DOES seem to decode the 2 B frames most of the time at the start of a FILE, but if you start streaming data from an I-frame located elsewhere in a file then the 2 preceding B frames are NOT decoded (as if it was an open GOP). But Mainconcept always seems to decode the 2 preceeding B frames no matter which I-frame in the file you start streaming from (as if all GOPs were closed).

Do you know what flags in the stream the ffdshow's H.264 decoder is using to decide not to decode the 2 leading B frames? And is it making a mistake? If Mainconcept can perfectly decode these B frames OK then surely ffdshow should be able to aswell.


and Haruhiko replied:


They are simply dropped without any flags. It's a bug then. I'll re-read the spec and think about the fix.


I was just wondering if there has been any progress with this. The latest version 3048 from Sourceforge seems to be the same.

[Gleb Egorych]

What is the problem with libdts ?
I read about that in MPC-HC thread: http://forum.doom9.org/showthread.php?p=1308404&highlight=dts#post1308404

[Ginsonic]

Maybe a bug ?
Since Version 3040 when resizing SD material to e.g. 720p, picture seems to be zoomed (only a part of the picture is resized to full screen resolution). When unchecking and checking resize checkbox, everything works like a charm. When opening a new video, trouble starts again. Version 3029 works flawlessly.
OSD always shows correct in- and output resolutions.

BTW: Thanks for Your great work !

[Gleb Egorych]

Since Version 3040 when resizing SD material to e.g. 720p, picture seems to be zoomed (only a part of the picture is resized to full screen resolution). When unchecking and checking resize checkbox, everything works like a charm. When opening a new video, trouble starts again. Version 3029 works flawlessly.

Also checking/unchecking resize breaks recent autocropping feature in FLV. Using ffdshow 3054 and FLV splitter 1.2.908. Same situation when using MPC-HC's r1207 internal FLV splitter.

BTW the feature does work only in EVR and VMR7_without_overlay for me. Overlay mixer and VMR7_with_overlay - no cropping at all. VMR9 - black screen, need to check/uncheck resize to get picture (without cropping). Tested in MPC-HC and ZP7.
WinXP, 8800GT, FW 186.18.

[albain]

Maybe related with those revisions :
in rev 3039,
"Added workaround for Overlay Mixer that prevent an issue with 'transparent' cropped areas. "

or

in rev 3031,
"copy VIDEOINFOHEADER.rcSource from the input pin to the output pin.
fix a crop problem with FLV videos.
https://sourceforge.net/tracker/index.php?func=detail&aid=2298876&group_id=173941&atid=867360"

[clsid]

Rev 3039 restores the 'old' behavior when using Overlay. So that one isn't the cause. Rev 3031 may be the cause because it possibly overrides any cropping settings of ffdshow.

[leeperry]

isn't there something wrong w/ 320x240 MPEG1 files in ffdshow :confused:

here's a sample : http://www.nanoed.org/courses/carbon_nanotube/ballv02.mpg

I upscale them to 1024x768(keep AR), and I get zoomed video :

http://thumbnails14.imagebam.com/4460/dab8b244593448.gif (http://www.imagebam.com/image/dab8b244593448)

if I uncheck then recheck the resize filter, then it's fine :

http://thumbnails7.imagebam.com/4460/c727bf44593450.gif (http://www.imagebam.com/image/c727bf44593450)

both verified w/ the default ffdshow(latest ICL10 build) settings, and in both KMP/MPC.

also these 320x240 MPEG1 videos refuse to play in HR(closes the player instantly), but work fine in EVR(just zoomed up to death)?! maybe because HR doesn't like the bogus data ffdshow sends out?

I've tried every possible MPEG1 decoder/splitter/ffdshow setting.. :thanks:

[Ginsonic]

I upscale them to 1024x768(keep AR), and I get zoomed video

It seems, You get the same troubles like me as described above, but it also happens with MPEG2 material (DVD in my case).

[leeperry]

ok, so something's fishy...I really don't think that it comes from my setup.

should one of us create a bug report on sourceforge? last time I did, they were ignored :o

Haruhiko's on vacation anyway, he well deserves it!

[dvd_]

using ffdshow tryouts revision 2857 with latest mpc.
when playing 720p and 1080p mkv files i witness a strange artifact, like a soft "snow" effect in some areas of the picture in some parts of the movie.
also, the video playback is not "smooth" in some parts and seems like the fps is slow.

when playing full bluray movie in powerdvd i dont witness those problems.

cpu is i7, and maximum load is 20% with mpc and 3% with powerdvd. so no trouble there.

any help or suggestions on how to improve the video playback quality using ffdshow and mpc would be great.
thanks.

[clsid]

@leeperry
Post a detailed bugreport on sourceforge. Otherwise it will certainly be forgotten.

@dvd_
Try a more recent build first.

[leeperry]

@leeperry
Post a detailed bugreport on sourceforge. Otherwise it will certainly be forgotten.
ok, just did : https://sourceforge.net/tracker/?func=detail&aid=2834209&group_id=173941&atid=867360

I was afraid my KMP config was wrong, but it's indeed ffdshow...as when using HR I'm getting an error in the windows XP event viewer, stating that ffdshow crashed KMP.

Type of event: DrWatson
Event ID : 4097
Description :
KMPlayer.exe, has generated an application error. The generated exception
was c0000005 at the adress 023F3815 (ffdshow!DllGetClassObject)

[travolter]

Hey guys.. anyone know how to solve this bug that I have?

I lose frames using current settings in ffdshow..
** that only appear into avi files with divx/xvid... other files like wmv are working ok.

http://usuarios.lycos.es/kakhen/kakhen/loseframes.PNG

Im decoding with ffdshow and haali spliter

[Leak]

Hey guys.. anyone know how to solve this bug that I have?
Ummm... for SPP deblocking the slider just sets the number of deblocking passes done - the further you move it to the right, the more CPU power you'll need for this to be done in realtime; each step roughly doubles the amount of CPU use.

If you get dropped frames, your CPU simply isn't fast enough to do that amount of SPP deblocking - except for getting a faster CPU there's nothing that can be done.

np: Iwasaki Taku - Love Is Destiny, Destiny Is Paper (R.O.D The TV Original Soundtrack 2)

[travolter]

If you get dropped frames, your CPU simply isn't fast enough to do that amount of SPP deblocking - except for getting a faster CPU there's nothing that can be done.

Leak.. but.. why it only happend with AVI files that contain (divx/xvid)?

Im using same settings with wmv (wmv3/VC-1), mkv (H264).. and decoding is perfect using these settings at max.


Maybe a bug into the ffdshow divx decoder? or really deblocking divx requires more cpu than movies with other codecs?

EDIT.- Big question!!!! Postprocessing is really working with other kind of videos but AVI? Im testing with WMV and I dont see difference

[netwolf]

Isn't 'Automatic quality control' (which I see is enabled) supposed to prevent Postprocessing from taking too much CPU time?
(by moving the slider automatically to the left if CPU usage approaches max.)

[JarrettH]

I propose releasing Beta 7 sometime next month. Of course all regressions since beta 6 should be fixed first.

Known regressions:
- WMV3 decoding crash. This is a bug in FFmpeg code. Will hopefully be fixed soon.
- SPDIF related issues? Can anyone who suffers from this problem summarize the problem, including last known good revision and first known bad revision?

If there is anything I forgot, please let me know.

Improving AutoCrop before releasing beta 7 would be nice. But since it was broken before, not a requirement.

@Haruhiko, do you have any plans for changes that you would like to include in beta 7?

soon soon soon? :) :D

[tetsuo55]

Microsoft made a page on MSDN explaining the bitstreaming of all supported audio formats over HDMI, Displayport and Spdif(on a different page).

Every bitstreamable format is explained on that this page except for 1, DTS-HD.

http://msdn.microsoft.com/en-us/library/dd316761(VS.85).aspx

Maybe FFdshow could be update to bitstream all those formats instead of only ac3 and dts

[EpsilonX]

Leak.. but.. why it only happend with AVI files that contain (divx/xvid)?

Im using same settings with wmv (wmv3/VC-1), mkv (H264).. and decoding is perfect using these settings at max.


Maybe a bug into the ffdshow divx decoder? or really deblocking divx requires more cpu than movies with other codecs?

EDIT.- Big question!!!! Postprocessing is really working with other kind of videos but AVI? Im testing with WMV and I dont see difference

In my personal experience...
It seems that H.264 uses its own deblocking routine...
It utilize lower amount of CPU power, way lower than decoding DivX/XviD...
Can't say about WMV...
But in my E5200@4Ghz, I can only use SPP level 2 slider...
That's with resize and sharpening...
CMIIW...

[ipanema]

Should I raise a bug report for

http://forum.doom9.org/showpost.php?p=1310827&postcount=7780

on SourceForge? In case it gets forgotten.

[clsid]

Yes. Please do.

[jruggle]

Unfortunately, the problem does not come from FFDShow but ffmpeg : the EAC3 decoder does not support 7.1 channels yet (6 channels max).
As a result, back and side channels are the same.

Please report the problem to ffmpeg team.

If someone has a 7.1 E-AC-3 file other than the one I have, which is just a channel test stream, please send it my way. I'm not going to use my time trying to implement something that has no practical benefit, so until I get a report of a real-world sample, FFmpeg will only support 6-channel E-AC-3...unless someone other than myself decides to do it.

[albain]

If someone has a 7.1 E-AC-3 file other than the one I have, which is just a channel test stream, please send it my way. I'm not going to use my time trying to implement something that has no practical benefit, so until I get a report of a real-world sample, FFmpeg will only support 6-channel E-AC-3...unless someone other than myself decides to do it. I agree, I have never seen 7.1 DD Plus movies, TrueHD is the standard for 8 channels for Dolby

Microsoft made a page on MSDN explaining the bitstreaming of all supported audio formats over HDMI, Displayport and Spdif(on a different page).

Every bitstreamable format is explained on that this page except for 1, DTS-HD.

http://msdn.microsoft.com/en-us/library/dd316761(VS.85).aspx

Maybe FFdshow could be update to bitstream all those formats instead of only ac3 and dts

I don't understand the interest of this structure : in vista some players (powerdvd, totalmedia theater) already support bitstreaming of TrueHD/DTS-HD. So what does this bring ?

As far as I understand, the gap to fill is the following : add 2 options in the output section "Dolby TrueHD passthrough" and "DTS-HD passthrough" next to AC3/DTS passthrough.
Next step is to bitstream those 2 HD *compressed* streams as for SPDIF.
The good news is that we have an audio parser that knows everything : sample size, number of channels, bitrate and of course source format.
So correct me if I am wrong, but the thing to do is to parse the HD compressed stream and send samples directly to the renderer as for SPDIF. However maybe the samples need to be reordered into a certain way.

I have taken a look at the current code and for DD and DTS the passthrough is done respectively into liba52 and libdts because the input stream needs to be parsed (for sync frames) but after that it is sent to output.

The other problem is testing : I don't have any radeon HD4xxxx or asus xonar

Any thoughts ?

[tetsuo55]

Those commercial players only allow untouched bitstreaming over protected path(which requires a special soundcard and software).

Adding this to ffdshow might enable bitstreaming for everyone.

Also i think the best feature would be bitstreaming of formats not fully supported by ffdshow, i'm interested in all of these:
DTS, DTS-ES, DTS 96/24, DTS-HD (MA)
Dolby Digital, Dolby Digital Surround EX, Dolby Digital Plus, Dolby TrueHD

[albain]

Indeed, that's what I've read : radeon (even 4xxx) and nvidia cards hold protected video path but not audio, so commercial players downgrade the LPCM signals to DVD quality and won't bitstream HD compressed signals.

Only the xonar with the TMT holds this protected media path.

The solution is anydvd which removes the AACS protection so the protected path, but commercial players also downgrade in that case.

What to do next : implementing bitstream of TrueHD, DTS-HD, and EAC3 to the numeric output.

*EDIT* : the issue here is also how to send the compressed stream directly to the renderer ?
For AC3, there are 2 mediatypes, AC3 and AC3_SPDIF to use when we want to do passthrough
There is only one for TrueHD, DTSHD...

[Fizick]

It seems, I found a bug in ffdshow porting of Yadif to SSE2
http://sourceforge.net/tracker/?func=detail&aid=2836414&group_id=173941&atid=867362
Output of SSE2(SSE3) is not the same as C(and SSEMMX)
Please check - I do not have SSE2 CPU now :)

[STaRGaZeR]

It seems, I found a bug in ffdshow porting of Yadif to SSE2
http://sourceforge.net/tracker/?func=detail&aid=2836414&group_id=173941&atid=867362
Output of SSE2(SSE3) is not the same as C(and SSEMMX)
Please check - I do not have SSE2 CPU now :)

This would produce any visible errors?

[albain]

I have made a development build for TrueHD/DTSHD/EAC3 passthrough.

I have no idea if it will work because I don't have the equipment.

Requirements :

Windows 7 (I don't think it will work on Vista, unless DirectShow is updated with upcoming WMP12)
A HDMI 1.3 output : Asus xonar, azuentech XFI Home theater HD, Radeon HD4xxxx
A HT system which supports HDMI 1.3 and the decoding of EAC3/TrueHD/DTSHD
M2TS samples with one of the following formats (it won't work on commercial players I guess).
MPC HC
FFDShow configured as followed : in the output section, there are new checkboxes for passthrough. Tick the formats that fit to your equipment


How it works : a new structure is provided in output so that the stream should be passed directly to the audio renderer.
FFDShow does nothing on the stream, it simply adds the correct information in output. MPC-HC will have no idea of this new structure (and does not recognize the new mediatypes yet), and it will pass the audio stream to the renderer with no moditification too.

I think that EAC3 passthrough should work on a SPDIF output (but I am not sure), but for the 2 others HDMI 1.3 is required to transport the audio.

Here is the URL :
http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871

To do next :
1/Make it work
2/Implement PCM/LPCM passthrough too (to avoid OS interfer too)

[tetsuo55]

I changed everything to "libavcodec"
I checked all the passthrough formats

I have 4770 and onkyo 607

First results are in:

-Dolby Digital works fine (Display on reciever changes to Dolby D)
-Dolby Digital EX works fine (Display on reciever changes to Dolby D EX)
-Dobly Digital Plus is not being sent, only silence. (Display on reciever changes to Dolby D)
-TrueHD is not being sent, only silence. (DDisplay on reciever changes to Dolby D)
-DTS works fine (Display changes to DTS)
-DTS-HD gets it's core sent as DTS (Display changes to DTS)

please let me know which details you need.

[albain]

I changed everything to "libavcodec"
I checked all the passthrough formats

I have 4770 and onkyo 607

First results are in:

-Dolby Digital works fine (Display on reciever changes to Dolby D)
-Dolby Digital EX works fine (Display on reciever changes to Dolby D EX)
-Dobly Digital Plus is not being sent, only silence. (Display on reciever changes to Dolby D)
-TrueHD is not being sent, only silence. (DDisplay on reciever changes to Dolby D)
-DTS works fine (Display changes to DTS)
-DTS-HD gets it's core sent as DTS (Display changes to DTS)

please let me know which details you need.

Have you tried to disable internal audio switcher of MPC-HC ?

Otherwise, could you tell me the input PIN information received by the audio renderer (right click/Filters/<sound device>) ?

Normally you should get the following for EAC3 :
subtype: Unknown GUID Name {0000000A-0CEA-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}

Otherwise, the problem may come from MPC itself that won't propagate the new structure (or change it by a legacy one) to the audio renderer.

The alternate test it to make it work from WMP12, but I don't know any way to play TrueHD/DTSHD/EAC3 stream through WMP

Thanks for your tests

[tetsuo55]

With stream switcher disabled, EAC3 and TrueHD now show PCM stereo on the reciever.
With stream switcher disabled DTS-HD no longer works, still showing DTS on the reciever

eac3
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}
Filter: ffdshow Audio Decoder
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0



EDIT:
With stream switcher disabled all bitstream formats are broken

[tetsuo55]

DTS 24/96 doesnt work, the reciever sees PCM stereo

[albain]

Weird, I also get PCM in the pin input of the audio render but not in the pin output of FFDShow

FFDShow audio (EAC3)------->(PCM) Audio renderer

I thought this was due to my sound adapter that won't support EAC3 bitstream

Here is the result I get from graphedit :

http://damienbt.free.fr/EAC3.jpg

As I said maybe the problem comes from MPC that won't propagate the structure

*EDIT* : can you test with the following build also ? (with and without the internal switcher) ?
FFDShow audio bitstream beta 2 (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)
Thanks

PS : the real question for directshow experts and that is not answered by MSDN doc (http://msdn.microsoft.com/en-us/library/dd316761%28VS.85%29.aspx) is : how to fill CMediaType.subtype ? with MEDIASUBTYPE_PCM or with the same value as
WAVEFORMATEXTENSIBLE_IEC61937.FormatExt.SubFormat=WAVEFORMATEXTENSIBLE.SubFormat= KSDATAFORMAT_SUBTYPE_IEC61937_DOLBY_DIGITAL_PLUS for EAC3, KSDATAFORMAT_SUBTYPE_IEC61937_DTS_HD for DTS HD, KSDATAFORMAT_SUBTYPE_IEC61937_DOLBY_MLP for TrueHD ...

Maybe clsid or Haruhiko have clues on this subject.
Also if someone know how to play HD audio streams within WMP12....

[albain]

Okay, this is my fault, I mixed media subtype and format tags which are 2 different notions

It should be better with the following :
FFDShow bitstream beta 3 (http://www.mediafire.com/file/0nmttl5jkye/ffdshow_rev3054_20090803_dbt_beta3.exe)

[tetsuo55]

still not working

eac3
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0



TrueHD
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0



DTS-HD
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0



Edit:
and this is without the audio switcher, disabling it breaks all streams including regular dts and ac3
DTS-HD
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}
Filter: ffdshow Audio Decoder
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0

[albain]

Some more work needs to be done then

Lastly, could you try the following sample with WMP12 : its a MKV so it should play fine in it
MKV Sample with EAC3 (http://www.mediafire.com/file/g0jztyyhioy/Battle.Of.The.Bulge.1965.HD-DVD.Re.x264.720P.EAC3-C@SiLU-Sample.mkv)


@developers
Here is the patch to work on. I followed MS guidelines from the MSDN link above, but maybe something is wrong with the stream itself (does it need to be swabbed or reformated ?)
Patch of source code of bitstream (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)

[tetsuo55]

Some more work needs to be done then

Lastly, could you try the following sample with WMP12 : its a MKV so it should play fine in it
MKV Sample with EAC3 (http://www.mediafire.com/file/g0jztyyhioy/Battle.Of.The.Bulge.1965.HD-DVD.Re.x264.720P.EAC3-C@SiLU-Sample.mkv)


@developers
Here is the patch to work on. I followed MS guidelines from the MSDN link above, but maybe something is wrong with the stream itself (does it need to be swabbed or reformated ?)
Patch of source code of bitstream (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)WMP only wants to use its own codecs on win7

[nm]

WMP only wants to use its own codecs on win7
Not with MKV.

[tetsuo55]

the sample refuses to play in wmp, hangs at 0 seconds, first frame is visable.

Does play in MPC-HC, but audio is silent

------
EDIT:

Maybe it requires a wasapi audio renderer?

[albain]

Indeed I suspect the audio renderer to be failing.

In WMP the correct renderer should be used, I don't understand. Maybe a problem with the audio drivers

In the doc I have it is said that the KSDATARANGE_AUDIO structure should be supported by the driver with the new driver model (the one from vista or 7 I don't know).

If I don't find the answer I will try to post this issue in MSDN forums, but we are in advance or the only one interested by this implementation : commercial applications have their own audio renderer I guess (to work with vista).

[tetsuo55]

Indeed I suspect the audio renderer to be failing.

In WMP the correct renderer should be used, I don't understand. Maybe a problem with the audio drivers

In the doc I have it is said that the KSDATARANGE_AUDIO structure should be supported by the driver with the new driver model (the one from vista or 7 I don't know).

If I don't find the answer I will try to post this issue in MSDN forums, but we are in advance or the only one interested by this implementation : commercial applications have their own audio renderer I guess (to work with vista).Yeah, i will look into hacking wmp with clsid's tool.

The commercial players have licences and documentation required for implementing these codecs.
Also there is only 1 player that does true bitstreaming.

[tetsuo55]

I tested with graphedit, still no sound.

The problem is not related to MPC-HC

[albain]

Yes, this is "normal" because WMP has the same behaviour.

Actually, the problem could come from FFDShow, because I forgot an important step : reconnect the output pin with the new media type

Here is the new build :
FFDShow dev build - bitstream beta 4 (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)

Also, another issue that I am not sure about, about EAC3 : normally EAC3 can be bitstreamed over AC3 (it is called AC3+).
So, I see 2 possibilities :
1/ If you have SPDIF output, EAC3 must be encoded into AC3 (it will still be EAC3 but embedded into an AC3 structure)
2/ If you have HDMI, EAC3 should be sent with no modification

[tetsuo55]

allright,

Its still not working, TrueHD and eac3 is silent, DTS-HD still gets stripped to core.

FFdshow in
Filter : ffdshow Audio Decoder - CLSID : {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}

- Connected to:

CLSID: {1365BE7A-C86A-473C-9A41-C0A6E82C9FA3}
Filter: C:\Users\HTPC\Documents\Downloads\TEST_7.1_DDPLUS.m2ts
Pin: Audio

- Connection media type:

Audio: Dolby AC3 48000Hz 6ch 640Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_DOLBY_AC3 {E06D802C-DB46-11CF-B4D1-00805F6CBBEA}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 18

WAVEFORMATEX:
wFormatTag: 0x2000
nChannels: 6
nSamplesPerSec: 48000
nAvgBytesPerSec: 80000
nBlockAlign: 2560
wBitsPerSample: 0
cbSize: 0 (extra bytes)

pbFormat:
0000: 00 20 06 00 80 bb 00 00 80 38 01 00 00 0a 00 00 . ..€»..€8......
0010: 00 00 ..


Ffdshow out:
Filter : ffdshow Audio Decoder - CLSID : {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: TEST_7.1_DDPLUS.m2ts / Audio

- Connection media type:

Audio: WAVE_FORMAT_EXTENSIBLE 48000Hz 4ch 3072Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {A7FB87AF-2D02-42FB-A4D4-05CD93843BDD}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 92

WAVEFORMATEX:
wFormatTag: 0xfffe
nChannels: 4
nSamplesPerSec: 48000
nAvgBytesPerSec: 384000
nBlockAlign: 8
wBitsPerSample: 16
cbSize: 34 (extra bytes)

pbFormat:
0000: fe ff 04 00 80 bb 00 00 00 dc 05 00 08 00 10 00 þÿ..€»...Ü......
0010: 22 00|10 00 33 00 00 00 0a 00 00 00 ea 0c 10 00 "...3.......ê...
0020: 80 00 00 aa 00 38 9b 71 80 bb 00 00 04 00 00 00 €..ª.8›q€»......
0030: 00 00 00 00 61 75 64 73 00 00 10 00 80 00 00 aa ....auds....€..ª
0040: 00 38 9b 71 af 87 fb a7 02 2d fb 42 a4 d4 05 cd .8›q¯‡û§.-ûB¤Ô.Í
0050: 93 84 3b dd 01 00 00 00 00 00 00 00 “„;Ý........

- Enumerated media type 0:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 1:

Unknown

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Stream {E436EB83-524F-11CE-9F53-0020AF0BA770}
subtype: MEDIASUBTYPE_None {E436EB8E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_None {0F6417D6-C318-11D0-A43F-00A0C9223196}
bFixedSizeSamples: 0
bTemporalCompression: 0
lSampleSize: 307200
cbFormat: 0


audio switcher in:
Filter : Audio Switcher - CLSID : {18C16B08-6497-420E-AD14-22D21C2CEAB7}

- Connected to:

CLSID: {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}
Filter: ffdshow Audio Decoder
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..


audio switcher out:
Filter : Audio Switcher - CLSID : {18C16B08-6497-420E-AD14-22D21C2CEAB7}

- Connected to:

CLSID: {E30629D1-27E5-11CE-875D-00608CB78066}
Filter: Realtek HDMI Output (ATI HDMI A
Pin: Audio Input pin (rendered)

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Set as the current media type

- Enumerated media type 1:

Unknown

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Stream {E436EB83-524F-11CE-9F53-0020AF0BA770}
subtype: MEDIASUBTYPE_None {E436EB8E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_None {0F6417D6-C318-11D0-A43F-00A0C9223196}
bFixedSizeSamples: 0
bTemporalCompression: 0
lSampleSize: 307200
cbFormat: 0


aduio renderer in:
Filter : Realtek HDMI Output (ATI HDMI A - CLSID : {E30629D1-27E5-11CE-875D-00608CB78066}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: Out

- Connection media type:

Audio: 0x0092 48000Hz stereo 1536Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0092
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 192000
nBlockAlign: 4
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 92 00 02 00 80 bb 00 00 00 ee 02 00 04 00 10 00 ’...€»...î......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 80 bb 00 00 02 00 00 00 ........€»......
0030: 00 00 00 00 61 75 64 73 00 00 ....auds..

- Enumerated media type 0:

Audio

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
formattype: TIME_FORMAT_NONE {00000000-0000-0000-0000-000000000000}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 0


EDIT:

with this new version battle of the buldge samples gets played twice as fast with loud static

[albain]

Okay, do you have the debug messages trace ? I added a trace which says "TffdshowDecAudioInputPin::Receive : the output sample format (compressed stream to bitstream) %d is not accepted by the renderer" which is raised when the connection is refused.

I guess that this is the case. I don't know how yet but if the connection is refused, something should be done (remove ffdshow out of the graph, for example).

I will submit the problem to MSDN.

Could you also try with WMP and the EAC3 mkv ?

Thanks

[tetsuo55]

Okay, do you have the debug messages trace ? I added a trace which says "TffdshowDecAudioInputPin::Receive : the output sample format (compressed stream to bitstream) %d is not accepted by the renderer" which is raised when the connection is refused.

I guess that this is the case. I don't know how yet but if the connection is refused, something should be done (remove ffdshow out of the graph, for example).

I will submit the problem to MSDN.

Could you also try with WMP and the EAC3 mkv ?

ThanksSame result as MPC, it plays 2x as fast with loud static.

My system is:
Windows 7 RC1 x64
ATI 4770
Onkyo 607

I do not see that debug error. where can i find it?

[albain]

Ok thank your for your tests.

I have submitted the problem to MSDN forums : here is the link (http://social.msdn.microsoft.com/Forums/en/windowsdirectshowdevelopment/thread/e6603225-4f1a-49a9-9ce9-68919e9264e1)

I think that the problem comes from the audio renderer (wrong renderer used or incompatible audio driver)

[clsid]

You can view the debug trace with a tool called "DebugView".

[albain]

You can view the debug trace with a tool called "DebugView".

You have to launch this tool while you play the video, and prior you have to enable debug traces by editing the following registry key :
HKEY_CURRENT_USER\Software\GNU\ffdshow\allowDPRINTF => 1

[tetsuo55]

I don't think this works for windows 7 X64.

[albain]

With windows 64 bits the registry key is slightly different but my builds are for 32 bits only so this should be this one

You made the tests with the 32 bits versions of MPC & WMP ?

*EDIT* : I have made a new build, which is a little different (it disconnects then reconnects to the audio renderer)
Also, if it detects that the audio renderer won't accept the media type it will switch to decode mode (EAC3/TrueHD... will be decoded instead of being sent compressed), of course the corresponding decoders must be enabled.

FFDShow bitstream beta 5 (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)

Patch of the source code : FFDShow bitstream source patch (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)

[Revgen]

If someone has a 7.1 E-AC-3 file other than the one I have, which is just a channel test stream, please send it my way. I'm not going to use my time trying to implement something that has no practical benefit, so until I get a report of a real-world sample, FFmpeg will only support 6-channel E-AC-3...unless someone other than myself decides to do it.

You don't need to do DD+ 7.1.

But DTS-HD would be nice. :D

[tetsuo55]

With windows 64 bits the registry key is slightly different but my builds are for 32 bits only so this should be this one

You made the tests with the 32 bits versions of MPC & WMP ?

*EDIT* : I have made a new build, which is a little different (it disconnects then reconnects to the audio renderer)
Also, if it detects that the audio renderer won't accept the media type it will switch to decode mode (EAC3/TrueHD... will be decoded instead of being sent compressed), of course the corresponding decoders must be enabled.

FFDShow bitstream beta 5 (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)

Patch of the source code : FFDShow bitstream source patch (http://www.mediafire.com/?sharekey=05fb19b4b2e4495ee7c82ed4b8f0c380e04e75f6e8ebb871)I only use 32bit player/filters.

With the new build:
-EAC3 gets decoded by ffdshow
-TrueHD gets decoder by ffdshow, the video/audio is played back ar half speed
-DTS-HD still gets stripped to DTS-Core and sent over spdif
-AC3/DTS still get send over spdif

So i guess that means we need a new audio renderer??

[albain]

Yes as long as the audio renderer will refuse the new structures, the stream will be decoded by ffdshow (the new security I added to avoid no sound).

But I expect MS to give a valuable answer on this : they wouldn't have added this for nothing

[clsid]

Yes as long as the audio renderer will refuse the new structures, the stream will be decoded by ffdshow (the new security I added to avoid no sound).I assume that this also useful for the 'old' code in the current SVN trunk? If so, you could already commit that part.

[tetsuo55]

Nothing appears to be broken on the surface, everything works like it did before the patch.

I never tested TrueHD with a pre-patch build.

[Thunderbolt8]

is it possible that ffdshow looks quite a bit inferior compared to progs like powerdvd or arcsoft tmt when it comes to 1080i 29.97fps stuff (blu-ray content)? the picture has a more "noisy" look, which is not so much present with those 2 other programs. apart from that, it also feels a bit like that powerdvd & tmt present a little more smooth picture, though this could be hardware related.

[nm]

is it possible that ffdshow looks quite a bit inferior compared to progs like powerdvd or arcsoft tmt when it comes to 1080i 29.97fps stuff (blu-ray content)? the picture has a more "noisy" look, which is not so much present with those 2 other programs.
Most likely those programs are using hardware deinterlacing and denoising. What do you use in ffdshow?

[Thunderbolt8]

(codec is AVC btw.)

either ffmpeg or yadif deinterlacer. denoising could be possible, since in some scenes theres a tiny amount of sharpness lost, but thats not really as bad as that little noise is anoying. especially since it always 'shakes' a little from left to right like a minimal shaking of the camera although its not moving, just the noise shifts minimally 'back and forth' at the same place.

[nm]

Upload a sample clip to MediaFire so we can take a look at it. Have you tried using hardware deinterlacing with ffdshow?

[leeperry]

BTW, I see some coders are working on ffdshow audio...would that be possible that the "number of channels" autoload condition works w/ AC3Filter please? it works w/ madflac :o

I made a bug report here: http://sourceforge.net/tracker/index.php?func=detail&aid=2725991&group_id=173941&atid=867360

maybe AC3Filter is at fault?! I know it likes to give BSOD's w/ the C-Media drivers(Asus Xonar mainly)....so maybe all the problems related to this decoder are not everyone's else fault after all :confused:

ffdshow uses the same AC3/DTS routines as AC3Filter? I couldn't find what librairies this thing uses...and it seems that ffdshow decodes AC3 in 32float, but not DTS :o

[STaRGaZeR]

(codec is AVC btw.)

All H.264 decoders must have bit-exact output. If two decoded streams are different it implies postprocessing (different deinterlacers or other filters) or a faulty decoder.

[Thunderbolt8]

Upload a sample clip to MediaFire so we can take a look at it. Have you tried using hardware deinterlacing with ffdshow?
hm how can I do that, using hardware deinterlacing with ffdshow?
and how can I cut a sample and be sure it will still play with those commercial progs?

[Thunderbolt8]

Upload a sample clip to MediaFire so we can take a look at it.heres a sample: http://www.sendspace.com/file/u83a49

[albain]

I assume that this also useful for the 'old' code in the current SVN trunk? If so, you could already commit that part.

I am currently making a cleaner code and removing the non working part.

Then I will commit it

[nm]

hm how can I do that, using hardware deinterlacing with ffdshow?
Go to "Output (http://ffdshow-tryout.sourceforge.net/wiki/video:output)" page in ffdshow video decoder configuration,
Select "Set interlace flag in output media type" and Method: "Bob"
Select NV12 colorspace and unselect all other colorspaces.

Remove all filters (deinterlacers, resizing, ...) that you were using.
If hardware deinterlacing and denoising doesn't work after these settings, try different renderers.

heres a sample: http://www.sendspace.com/file/u83a49
Yes, the problem of jumping texture/noise patterns is in the video and the only way to reduce it during realtime playback seems to be heavy denoising that destroys fine details. Not all videos need such filtering, so I'd rather re-encode this problematic movie with some AviSynth filtering or set up a playback filtering chain separate from the one that I normally use.

[Thunderbolt8]

hardware deinterlacing works now that way, thanks. playback is quite a bit faster now and also had noise problem got somewhat better due to that. not as good as with powerdvd, but still better than all those software renderers.
only remaining problem seems to be that when I activate sharpening and the camera moves then the sharpened layer also seems to jump as well :S

btw. is it still possible to use (.ass) subs with hardware deinterlacing?

[clsid]

If you use the internal subtitle renderer of MPC then it should work.

[Thunderbolt8]

whats actually the purpose of directvobsub/vsfilter then if the internal filter already should work?

[tetsuo55]

whats actually the purpose of directvobsub/vsfilter then if the internal filter already should work?All filters are available for "other" directshow video players like windows media player.

Vsfilter is also interesting for mpc-hc use, some bugs that are in the internal splitter are not in vsfilter and vice versa (this has to do with the way the subs are drawn to the screen)

[nm]

only remaining problem seems to be that when I activate sharpening and the camera moves then the sharpened layer also seems to jump as well :S
What sharpening? If you sharpen with ffdshow before hardware deinterlacing, serious artifacting is certainly expected.

[Thunderbolt8]

I sharpen with ffdshow, yes (needed for downscaling in order not to have a too blury picture). but how 'before hardware deinterlacing', can I influence the order which of these comes first?

[DarkZell666]

I sharpen with ffdshow, yes (needed for downscaling in order not to have a too blury picture). but how 'before hardware deinterlacing', can I influence the order which of these comes first?

Hardware de-interlacing comes last no matter what, so ... you can't =) Either use a sharpener that supports interlaced input, or separate fields + sharpen + weave.

[nm]

I sharpen with ffdshow, yes (needed for downscaling in order not to have a too blury picture). but how 'before hardware deinterlacing', can I influence the order which of these comes first?
You can't. AFAIK, hardware deinterlacing (and other hardware postprocessing) is done at the rendering stage, so ffdshow can't get the frames back for further processing.

[leeperry]

sharpen with ffdshow [..] hardware deinterlacing
you only option is the PS sharpening scripts in MPC/KMP I think..try the default one at 0.65, it looks good, not as good as LSF though.

or through your graphic card drivers maybe?

[Thunderbolt8]

hm I have a radeon 4770, at least I haven't seen such an option in catalyst control center

[nm]

hm I have a radeon 4770, at least I haven't seen such an option in catalyst control center
Looks like edge enhancement and denoising should be adjustable (or at least used to be): http://www.avsforum.com/avs-vb/showthread.php?t=1060000

[Thunderbolt8]

wtf, denoising of 64 is the default setting 0o

in order to be as close to be to the original intended presentation as possible, I guess such options as denoising & edge enhancement & dynamic constrast have to be disabled then?

[nm]

in order to be as close to be to the original intended presentation as possible, I guess such options as denoising & edge enhancement & dynamic constrast have to be disabled then?
Yes, if the video is intended to look the way it does. Your sample most likely isn't. ;)

One of the reasons for denoising by default is that review sites test the video playback quality of graphics cards and other devices by playing benchmark videos such as HD HQV. The tested device gets better benchmark scores if it denoises noisy scenes properly.

[leeperry]

hehe, you opened Pandora's box...graphic drivers doing funky stuff on VMR/EVR(especially ATi). there's a good reason why some ppl prefer HR/mVR :D

[ikarad]

What means "Added new audio structure for windows 7" in rev 3056?

"Revision 3056 - Directory Listing
Modified Mon Aug 17 12:37:38 2009 UTC (30 hours, 59 minutes ago) by albain

Added new audio structure for windows 7
Added a verification of supported output sample format when the codec changes"

[clsid]

It means better S/PDIF support.

[Mercury_22]

Some interesting news Dolby Digital Plus Chosen as an Audio Format for Windows 7 (http://investor.dolby.com/ReleaseDetail.cfm?ReleaseID=403647) !
Maybe we will need the full 7.1 channels DD+ working in the end, cause this may rise back the almost dead DD+ format ?!

[Revgen]

Seems like they are saying that Dolby Digital Decoding will be present in all Windows Operating systems. The reference decoder will ship with Windows 7.

Therefor Windows 7 users may not even need FFDShow for EAC3 decoding.

[Mercury_22]

Seems like they are saying that Dolby Digital Decoding will be present in all Windows Operating systems. The reference decoder will ship with Windows 7.

Therefor Windows 7 users may not even need FFDShow for EAC3 decoding.
Yes if your player it's using Media Foundation

[albain]

So far I have no answer from Microsoft on bitstream
http://social.msdn.microsoft.com/Forums/fr-FR/windowsdirectshowdevelopment/thread/6e5985fb-3ca2-41b3-9a99-7a9b97ae7357

About revision 3056, a few details :

About the new audio structure : this new audio structure supported only on windows 7 (maybe on vista in the future) is supposed to bring bitstream of compressed format to the HDMI/SPDIF output, supposing that your HT system will support the decoding of these formats. This will be useful especially for DTS-HD decoding (no open source decoder exists yet), and for other formats where HT systems sometimes decode better.
Better SPDIF support : currently there are several source filters that are not up to date and for example tell FFDShow the stream is DolbyDigital when this TrueHD or EAC3... So the codec change and the output format may change too. In this case the output filter (audio renderer) may not be compatible. So a verification has been added before changing the output format, and if not compatible, decode internally


Today some commercial filters are able even on xp/vista to bitstream compressed formats to the HDMI output : totalmedia theatre sends TrueHD/DTSHD directly to the HDMI 1.3 output, requiring a asus xonar, but this requirement is only to ensure the protected audio path, otherwise technically it is possible to do it on any HDMI output.
So either this feature is brought by the OS and can be implemented into ffdshow, or else a new audio renderer will be required I guess and/or a new audio driver

[Octo-puss]

I came along somewhat interesting discussion. A dude said "why would I use FFDShow when MPC-HC plays everything for me?". Well, what would you tell him?

[DarkZell666]

I came along somewhat interesting discussion. A dude said "why would I use FFDShow when MPC-HC plays everything for me?". Well, what would you tell him?

If he's happy with MPC-HC, leave him alone, unless you know ffdshow-tryouts offers extra functionnalities he might want, without breaking MPC-HC's.

[Mercury_22]

I came along somewhat interesting discussion. A dude said "why would I use FFDShow when MPC-HC plays everything for me?". Well, what would you tell him?

On a windows 7 system if MPC-HC will replace ffmpeg with ffmpeg-mt (for the non-DXVA files) and will add independent volume for each channel I'll say the same thing, cause for me Winows 7's (DXVA) filters are better then FFDShow's or MPC-HC'S

[albain]

On a windows 7 system if MPC-HC will replace ffmpeg with ffmpeg-mt (for the non-DXVA files) and will add independent volume for each channel I'll say the same thing, cause for me Winows 7's (DXVA) filters are better then FFDShow's or MPC-HC'S

I disagree, I find DXVA decoding (sometimes) less good than software decoding (for example coreavc without cuda or some software mpeg2 decoders).

[Snowknight26]

Would it be possible to have the option 'Show current volume levels' actually show them when Volume is not checked?

[ohmy]

Hi!

I'm a longtime ffdshow and MPC user, and still use ffdshow, although MPC-HC can play almost anything now just like ffdshow. But the subtitle feature is much better in ffdshow with its heuristic search and additional options.

Lately I noticed something. Some movies need volume normalising because of their loud explosions and almost silent voices. I used ffdshow for this, but it didn't help much. It is changing the volume, but only slightly. Voices still low, Explosions high. Regain volume was checked/unchecked, max. amplification changed.

Than I tried MPC-HC normalise option and Wow, it did exactly what I wanted. Everything was normalised at the instant I checked the normalise option. Voices were audible, and explosions went down to the same level. The movie became watchable without constantly changing the volume on the remote.

Is it possible to make an MPC-HC like normalisation? Maybe as a new option at the Volume tab.

[clsid]

Would it be possible to have the option 'Show current volume levels' actually show them when Volume is not checked?
That would waste CPU cycle for everyone that never looks there.
At default settings, Volume should not alter the sound, so enable it when you want to see the volume levels.

[clsid]

@ohmy
You probably have wrong speaker settings in ffdshow. Enable the 'Mixer' and set the correct amount of speakers.

[ohmy]

@ohmy
You probably have wrong speaker settings in ffdshow. Enable the 'Mixer' and set the correct amount of speakers.

I have 2 speakers. Mixer is enabled, and set to stereo.
MPC-HC and ffdshow normalisation still don't behave the same.

[Snowknight26]

That would waste CPU cycle for everyone that never looks there.
At default settings, Volume should not alter the sound, so enable it when you want to see the volume levels.

In the end though, it's just an extra step to do that as it would be the user's choice to have it checked.

[ohmy]

Here's a sample file for my volume normalization problem. It's a fake trailer to an upcoming movie grabbed from youtube:

http://rapidshare.com/files/269547376/safa.mp4.html

I modified the first half of it, so the sound is very quiet (20%), and in the second half it changes back to normal. (100%)
The sound change is at 0:54


While MPC-HC makes the volume level equal and sounds good, ffdshow seems to do almost nothing:(

[Octo-puss]

On a windows 7 system if MPC-HC will replace ffmpeg with ffmpeg-mt (for the non-DXVA files) and will add independent volume for each channel I'll say the same thing, cause for me Winows 7's (DXVA) filters are better then FFDShow's or MPC-HC'S

That's interesting. But how do I force MPC-HC to use Windows' filters rather the builtin ones?

[clsid]

@Octo-puss
Simply disable the internal ones in MPC-HC. Options -> Internal Filters.

@ohmy
For me volume normalization works pretty good when using ffdshow. I don't recall any others who have troubles with it. Are you sure ffdshow is being used and not any of the internal decoder of MPC-HC?

[ohmy]

@clsid
I'm sure, it's not MPC-HC-s internal decoder. All the transform filters are unchecked, and the audio is also unchecked. If it is checked, the normalization is working:)

The ffdshow audio icon is present while playing the file. The volume option is checked, and looking at the volume levels, it shows, that the first half is quieter, than the second half.
In ffdshow audio the following are checked: normalize, regain volume. Master volume is set to 100%. Max. amplification set to 400%.

And normalization is working, but it's not as strong as MPC-HC's. With only ffdshow there is an obvious volume change at 0:54. With MPC-HC, it can't be heard.


OK, I tried with GOM player using FFdshow. It is doing the same thing. The sound gets much better, when I click GOM player's normalize option.

ffdshow is at rev. 3054

[Leak]

@ohmy
For me volume normalization works pretty good when using ffdshow. I don't recall any others who have troubles with it. Are you sure ffdshow is being used and not any of the internal decoder of MPC-HC?
Errr... (http://forum.doom9.org/showthread.php?p=1197804#post1197804) (I guess that was the last time normalization came up here...)

np: Billy Bragg - M For Me (Mr. Love & Justice (Solo Version))

[ikarad]

I have a problem with ffdshow 64 and mpc-hc 64. If I use deband filter, I have desynchronisation of video and sound.
With 32 bits version, I have no problem.

I use ffdshow 64 3054

[Leak]

I have a problem with ffdshow 64 and mpc-hc 64. If I use deband filter, I have desynchronisation of video and sound.
With 32 bits version, I have no problem.
Check the CPU usage. I'm pretty sure the 64 bit version of ffdshow uses pure C code for deband, so it's simply running out of CPU power...

np: Moonbeam - You Can Hear Them (100)

[ikarad]

Check the CPU usage. I'm pretty sure the 64 bit version of ffdshow uses pure C code for deband, so it's simply running out of CPU power...

np: Moonbeam - You Can Hear Them (100)

I check and I have no problem with cpu load. my cpu load is 30% even with deband. Q6600@3ghz.

The problem exist with all videos: a simple divx or mkv with light cpu load until bluray movie.

with ffshow 32bits, there is no problem.

It's a bug with deband under ffdshow 64 bits.

[nm]

I check and I have no problem with cpu load. my cpu load is 30% even with deband. Q6600@3ghz.
One core or overall load? 25 % overall can mean that one core of a quad is fully saturated.

[ikarad]

One core or overall load? 25 % overall can mean that one core of a quad is fully saturated.

each core is 30% load, no core saturated.

but even with video with 5 % core load, deband doesn't work with ffdshow 64.

with ffdshow 32 it works very well with each movie.

It's just a bug of deband with ffdshow 64.

[Leak]

each core is 30% load, no core saturated.
Forget that. That's one process being bounced around all 4 cores to keep the CPU temperature down by distributing the produced heat over all cores.

Just look at the "Process" page in Task Manager. If the player process is using 25% CPU one core IS saturated, and ffdshow is going to be dropping frames. And yes, ffdshow 64 has no optimized assembly code for deband. That's why it's a lot slower.

Why use ffdshow 64 anyway?

np: Whitetree - Koepenik (Cloudland)

[ikarad]

Forget that. That's one process being bounced around all 4 cores to keep the CPU temperature down by distributing the produced heat over all cores.

Just look at the "Process" page in Task Manager. If the player process is using 25% CPU one core IS saturated, and ffdshow is going to be dropping frames. And yes, ffdshow 64 has no optimized assembly code for deband. That's why it's a lot slower.

Why use ffdshow 64 anyway?

np: Whitetree - Koepenik (Cloudland)

i already said, no core is saturated. It's the third time that I said. I am not stupid.

Even with simple mkv where each core is 5% load, deband with ffdshow 64 doesn't work and video is desynchronized. With ffsdhow 32 it works very well. although, core load for each cores is the same between ffdshow 32 and ffdshow 64.

I confirm there is a problem between deband and ffdshow 64 3054.

I want to try ffdshow 64 because I use a 64 bits OS. If ffdshow 64 exists it's to be used.

I only just report a bug for that this bug will be corrected in a new version.

[Leak]

I want to try ffdshow 64 because I use a 64 bits OS. If ffdshow 64 exists it's to be used.
I run a 64 bit OS (Windows 7) myself and I wouldn't even think of installing ffdshow 64 - unless your media playback needs more than 2 gigs of RAM AND every codec, filter and player you use is available as a 64 bit version you'll be missing something, and the playback experience can't get any smoother if it's already smooth with the 32 bit version.

Granted, AFAIK there's no 32 bit version of Windows Media Center installed with the 64 bit version, but except for that I'd steer far away from 64 bit media playback...

np: Fink - If I Had A Million (Sort Of Revolution)

[ikarad]

I run a 64 bit OS (Windows 7) myself and I wouldn't even think of installing ffdshow 64 - unless your media playback needs more than 2 gigs of RAM AND every codec, filter and player you use is available as a 64 bit version you'll be missing something, and the playback experience can't get any smoother if it's already smooth with the 32 bit version.


listen, if ffdshow 64 doesn't interesting for you, it's your problem. Your opinion doesn't interest me. I only report a bug, not a debate..

Like I said, I just tried ffdshow 64 to see the differences (I have the right to test, no?) and I report a bug. If ffdshow 64 is more faster or less faster than ffdshow 32 it's not my problem. I just report a bug.

Do you know what means a bug?

thank you to understand that a debate on the uselfull of ffdshow 64 or not isn't the problem. I don't care if ffdshow 64 is more faster than ffdshow 32 or not. The problem is that deband filter doesn't work with ffdshow 64 3054rev.

[clsid]

It is your choice to use it. We can only advice not to use it. The chance that the deband bug will be fixed is very small, unless you can find a new developer that want to spend time on figuring out why that particular code does not work for 64-bit builds.

[clsid]

@albain

There seems to some bugs in revision 3056. I have seen several reports of people with audio problems.
1) AVI files with MP3 audio give mono output (right channel) instead of stereo (tested on XP and Vista).
2) Input channels on Volume page are wrong. For example stereo input shows BackL and LFE as active. Mono input shows LFE as active input.
3) ffdshow connects to MONOGRAM AMR Encoder or AC3Filter when enabling SPDIF.

[ikarad]

It is your choice to use it. We can only advice not to use it. The chance that the deband bug will be fixed is very small, unless you can find a new developer that want to spend time on figuring out why that particular code does not work for 64-bit builds.

why make a 64 bits version of ffdshow if this version is inadvisable? It would be interresting to disable filters on ffdshow 64 that don't work.

[clsid]

The is a 64-bit build because some people do have a valid reason for using it. Of course broken functionality could be disabled. There just needs to be somebody that wants to test everything and adjust the code. This project is run by volunteers.

[albain]

@albain

There seems to some bugs in revision 3056. I have seen several reports of people with audio problems.
1) AVI files with MP3 audio give mono output (right channel) instead of stereo (tested on XP and Vista).
2) Input channels on Volume page are wrong. For example stereo input shows BackL and LFE as active. Mono input shows LFE as active input.
3) ffdshow connects to MONOGRAM AMR Encoder or AC3Filter when enabling SPDIF.

I reproduce the problem. If I can't fix it quickly I will revert the changes

[clsid]

Ok. Thanks.

[albain]

The is a 64-bit build because some people do have a valid reason for using it. Of course broken functionality could be disabled. There just needs to be somebody that wants to test everything and adjust the code. This project is run by volunteers.

Also FFDShow 64 is for many formats the only alternative for using with Media Center which uses directshow64

Otherwise, I think I have fixed the problem for stereo giving mono. I disabled the new audio structure temporarirely. If I can't find the solution I will remove it.

[mark0077]

Guys, not sure if this could be classed as a bug or not, but I am using ffdshow with mpc-hc to convert some interlaced content to progressive.

I am using the de-interlace, resize and finally grab feature of ffdshow to get the effect I want in the output file. I am able to grab perfectly de-interlaced frames at the correct resolution. Unfortunately, even though I have asked ffdshow grab to grab every frame, it seems like every second frame is missed.

Is grab in this case incorrectly doing its thing based on frame numbers from the input file, or from the output from the yadif de-interlacer. I think it should grab frames coming from the de-interlacer, ie give me 50 grabs per second of video instead of 25.

Thanks

[diizzy]

As usual, a big thanks to the developers and contributors for keeping this project alive.

I just have a small question in case I'm missing something, this "issue" is on both Vista and Win7 if you have your computer hooked up to a receiver with HDMI. If you set your speaker set to 5.1 Windows always outputs it as 5.1 (multichannel) but with only L,R active if you have a stereo source. The workaround would be to enable some surround effect in either ffdshow or in Windows (I only think its available in Windows 7) if you want sound in all speakers. Is it possible to force regular stereo output for 2 ch sound like when you set 2 speakers in Windows so the stream is truly 2ch?
That way you can use upmixing features on your reciever.

Thanks in advance
//Danne

[clsid]

I don't think there is such an option in Windows itself. You would have to check the settings of your sound driver. If you have a fancy sound card, then there may be such an option.

[diizzy]

Realtek doesn't seem to have such an option...
I think users would've noticed this issue earlier though...
//Danne

[yesgrey]

Is it possible to force regular stereo output for 2 ch sound like when you set 2 speakers in Windows so the stream is truly 2ch?
Have you tryed with reclock? I'm not sure, but maybe reclock could do that for you...

[leeperry]

I am using ffdshow with mpc-hc to convert some interlaced content to progressive.
BTW, my first filter in fffdshow is YADIF("process frames flagged as progressive" disabled), so ideally it should only turn itself on for interlaced content...but it's been processing progressive content as well lately :o

I think it started to happen after I upgraded to a newer version...how can I check if ffdshow detected an interlaced source? I loved when it was fully automatic, I'll try to change the MPEG decoders maybe....I think it was working fine w/ the MPC-HC decoder, which seemed to set the interlaced bit properly.

[leeperry]

oh, and I dunno what the defaults are...but I think it'd be good that you set liba52 and libdts as default decoders as they both decode in 32float(libavcodec only decodes in 16int, which is much lossier if you use post-processing).

[mark0077]

BTW, my first filter in fffdshow is YADIF("process frames flagged as progressive" disabled), so ideally it should only turn itself on for interlaced content...but it's been processing progressive content as well lately :o

I think it started to happen after I upgraded to a newer version...how can I check if ffdshow detected an interlaced source? I loved when it was fully automatic, I'll try to change the MPEG decoders maybe....I think it was working fine w/ the MPC-HC decoder, which seemed to set the interlaced bit properly.

Are you by any chance having the problem with DVD's? which you might already know don't have reliable flags. You will know if yadif is in use if you have "Double framerate" on, you can then see in mpc-hc's status, the frame rate will be 50.00 instead of 25.00 for example.

Anyone know if grab should catch these second frames that are output from yadif, seems like they arn't being grabbed in the current versions.

[leeperry]

Are you by any chance having the problem with DVD's? which you might already know don't have reliable flags. You will know if yadif is in use if you have "Double framerate" on, you can then see in mpc-hc's status, the frame rate will be 50.00 instead of 25.00 for example.
yes, I had the problem on DVD's...but also on progressive AVI files?! apparently something's messed up w/ the flag, or in the latest ffdshow revs/my user settings :o

if you enable YADIF, it doesn't turn on for AVI files for you? and I've got some interlaced WMV files where it doesn't.

I guess automatic deinterlacing is not quite doable..

[diizzy]

Have you tryed with reclock? I'm not sure, but maybe reclock could do that for you...

It does indeed (checkbox), although it seems to have some bugs also. First video I tried made it hang, probably due to SPDIF. Would be lovely if that feature could be adapted to ffdshow although I don't see any source code available for reclock.
//Danne

[yesgrey]

First video I tried made it hang, probably due to SPDIF.
If you read the option it says "for PCM sound", so probably you cannot use it with spdif... Have you tryed it with PCM sound?

[albain]

As usual, a big thanks to the developers and contributors for keeping this project alive.

I just have a small question in case I'm missing something, this "issue" is on both Vista and Win7 if you have your computer hooked up to a receiver with HDMI. If you set your speaker set to 5.1 Windows always outputs it as 5.1 (multichannel) but with only L,R active if you have a stereo source. The workaround would be to enable some surround effect in either ffdshow or in Windows (I only think its available in Windows 7) if you want sound in all speakers. Is it possible to force regular stereo output for 2 ch sound like when you set 2 speakers in Windows so the stream is truly 2ch?
That way you can use upmixing features on your reciever.

Thanks in advance
//Danne

This is possible if you enable autoload condition on ffdshow audio :
if nbchannels = 2, define a preset with mixer set to 2/0

[mark0077]

Hi all,

I am trying to use the ffdshow encoder to make lossless clips from adobe premiere. What could I be doing wrong.

Using ffv1 in avi, the exported clip plays the first 5 seconds and then crashes.
Using lossless jpeg in avi, the exported clip initially comes up as pure grey image, then shows a streaking effect and crashes.
Using H264 losslessin avi, the exported clip plays fine the whole way through, but its suspiciously small, and shows horrible video corruption the whole way through.

I am exporting a 60fps 1024 x 768 clip which exports fine completely uncompressed (not using ffdshow encoder).

My aim is to get my clip approx 30% smaller than the uncompressed lossless version so that my hdd can keep up with the clip, 4ghz core i7 should be enough I hope to decode this on the fly.

I am using ffdshow and mpc versions released today, 3060 and 1245. h264 lossless is the only one that I can actually see some output, although its badly garbled.

[mrcorbo]

As usual, a big thanks to the developers and contributors for keeping this project alive.

I just have a small question in case I'm missing something, this "issue" is on both Vista and Win7 if you have your computer hooked up to a receiver with HDMI. If you set your speaker set to 5.1 Windows always outputs it as 5.1 (multichannel) but with only L,R active if you have a stereo source. The workaround would be to enable some surround effect in either ffdshow or in Windows (I only think its available in Windows 7) if you want sound in all speakers. Is it possible to force regular stereo output for 2 ch sound like when you set 2 speakers in Windows so the stream is truly 2ch?
That way you can use upmixing features on your reciever.

Thanks in advance
//Danne

This would seem to be the relevant MSDN doc (http://msdn.microsoft.com/en-us/library/ms790524.aspx) if a developer had the time/interest to add a "match speaker config to output" option to ffdshow.

ReClock works well for me, though, for this purpose and in fact the function was added to ReClock specifically to address your situation.

[diizzy]

This is possible if you enable autoload condition on ffdshow audio :
if nbchannels = 2, define a preset with mixer set to 2/0

Unfortunately this doesnt seem to work, even if I set 2/0/0 I get multichannel PCM. The link provided by mrcorbo appears to be that I'm looking for. Unfortunately Im not a developer myself so I cant add it :-/

Best regards
Danne

[STaRGaZeR]

Unfortunately this doesnt seem to work, even if I set 2/0/0 I get multichannel PCM. The link provided by mrcorbo appears to be that I'm looking for. Unfortunately Im not a developer myself so I cant add it :-/

Best regards
Danne

Albain's trick is for creating the surround effect inside ffdshow, then output the multichannel PCM to the receiver as always but only for stereo sources and leaving 5.1 streams untouched. You can even create your personal matrix for the mix, so the possibilities are endless.

[JarrettH]

Do you think there'll be a new beta soon? :confused:

[_xxl]

Maybe a bug ?
Since Version 3040 when resizing SD material to e.g. 720p, picture seems to be zoomed (only a part of the picture is resized to full screen resolution). When unchecking and checking resize checkbox, everything works like a charm. When opening a new video, trouble starts again. Version 3029 works flawlessly.
OSD always shows correct in- and output resolutions.
Is 3031 the cause for that regression? If nobody is going to fix I will revert it.

[albain]

Is 3031 the cause for that regression? If nobody is going to fix I will revert it.

yes but clsid added a workaround since, I don't know if that fixed all the problems though

[clsid]

My workaround was only for a limitation of the Overlay Mixer. It does not fix the reported cropping and resizing issues.

I think it may be best to revert 3031, until a better patch is made that also deals properly with the filters of ffdshow.

[Mercury_22]

Audio it's broken in 3063 for the x64 version
x64 version Video and x86 version (video and audio) are working fine !

EDIT: For example I'm getting this (with MPC-HC)
ffdshow Audio Decoder::Out

Media Type 0:
--------------------------
Audio: PCM 8000Hz mono 128Kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_PCM {00000001-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 58

WAVEFORMATEX:
wFormatTag: 0x0001
nChannels: 1
nSamplesPerSec: 8000
nAvgBytesPerSec: 16000
nBlockAlign: 2
wBitsPerSample: 16
cbSize: 0 (extra bytes)

pbFormat:
0000: 01 00 01 00 40 1f 00 00 80 3e 00 00 02 00 10 00 ....@...€>......
0010: 00 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 40 1f 00 00 01 00 00 00 ........@.......
0030: 00 00 00 00 00 00 00 00 e8 49 ........èI


Media Type 1:
--------------------------
Unknown

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Stream {E436EB83-524F-11CE-9F53-0020AF0BA770}
subtype: MEDIASUBTYPE_None {E436EB8E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_None {0F6417D6-C318-11D0-A43F-00A0C9223196}
bFixedSizeSamples: 0
bTemporalCompression: 0
lSampleSize: 307200
cbFormat: 0

when it gets connected, but usually even that FFD's merit it's "Very High" MPC-HC or graphstudio64 (or WMP) are connecting to other audio decoder with lower merit

[clsid]

That is the output. What is the input?

[Mercury_22]

That is the output. What is the input?
Are you referring to the codec ? This is the file for which I manage to connect the ffd Impreza.3gp (http://sharebee.com/a931666c)
For any other type of file as i've said I can't mange to connect to FFD even that it has the highest merit on my system
Media info for that file General
Complete name : D:\Recorded TV\Impreza.3gp
Format : MPEG-4
Format profile : 3GPP Media Release 4
Codec ID : 3gp4
File size : 67.9 KiB
Duration : 7s 0ms
Overall bit rate : 79.5 Kbps
Encoded date : UTC 2003-11-17 18:26:16
Tagged date : UTC 2003-11-17 18:26:16

Video
ID : 1
Format : H.263
Format profile : BaseLine@1.0
Codec ID : s263
Duration : 6s 913ms
Bit rate mode : Variable
Bit rate : 64.5 Kbps
Width : 128 pixels
Height : 96 pixels
Display aspect ratio : 4:3
Frame rate mode : Variable
Frame rate : 15.044 fps
Minimum frame rate : 12.500 fps
Maximum frame rate : 29.969 fps
Bits/(Pixel*Frame) : 0.349
Stream size : 54.4 KiB (80%)
Writing library : noki
Encoded date : UTC 2003-11-17 18:26:16
Tagged date : UTC 2003-11-17 18:26:16

Audio
ID : 2
Format : AMR
Format/Info : Adaptive Multi-Rate
Format profile : Narrow band
Codec ID : samr
Duration : 7s 0ms
Bit rate mode : Constant
Bit rate : 12.8 Kbps
Channel(s) : 1 channel
Sampling rate : 8 000 Hz
Resolution : 16 bits
Stream size : 10.9 KiB (16%)
Writing library : Nokia
Encoded date : UTC 2003-11-17 18:26:16
Tagged date : UTC 2003-11-17 18:26:16

EDIT: Last (tested) working version is 3054, first (tested) broken 3058

[albain]

It may be my modifs then (rev 3056) but I don't understand why it would work on x86 and not on x64 ??
The modifications I brought is to check if the output filter (audio renderer) will accept the media type in case of format change. This does not apply to your case but the other svn modifications concern x264 and ffmpeg updates.

Could you report the debugview traces when launching this file from mpc-hc 64 ?
You will have to set the allowDPRINTF registry key to 1 in FFDShow *64* registry key.

Test number 2 : could you install ffdshow 3054 64 bits, make a copy of libavcodec.dll and then install a non working version of ffdshow (3058 for example) and overwrite libavcodec.dll by the rev3054 one ?

Thanks

[Mercury_22]

It may be my modifs then (rev 3056) but I don't understand why it would work on x86 and not on x64 ??
The modifications I brought is to check if the output filter (audio renderer) will accept the media type in case of format change. This does not apply to your case but the other svn modifications concern x264 and ffmpeg updates.

Could you report the debugview traces when launching this file from mpc-hc 64 ?
You will have to set the allowDPRINTF registry key to 1 in FFDShow *64* registry key.

Test number 2 : could you install ffdshow 3054 64 bits, make a copy of libavcodec.dll and then install a non working version of ffdshow (3058 for example) and overwrite libavcodec.dll by the rev3054 one ?

Thanks

The libavcodec.dll (3063 replaced with 3054) trick has no result = no changes and I can't find the allowDPRINTF key I even did a full registry search for allowDPRINTF It should be the same path for the x86 version on a x86 windows so if you can be more precisely (HKLM\Software\GNU ? HKCU\Software\GNU ? or ?)

EDIT: Replacing the ffdshow.ax with the old one (3054) does the trick = ffd it's connectable (working)

[alexins]

Test number 2 : could you install ffdshow 3054 64 bits, make a copy of libavcodec.dll and then install a non working version of ffdshow (3058 for example) and overwrite libavcodec.dll by the rev3054 one ?


Replacing the file does not change anything. The same problem exists with the ffdshow x86 icl10 3056-3063.

[_xxl]

Replacing the file does not change anything. The same problem exists with the ffdshow x86 icl10 3056-3063.
Only ICL10 version? Is 3055 working?

[Mercury_22]

Ok I've solved the allowDPRINTF key "mystery" first it has to be created by checking the "Output strings to debugger" but I have no log file (maybe because FFD audio it's not connecting )where should it be? C:\Program Files\ffdshow ? or C:\ ? or...?

EDIT: ICL11 rev 3063 it's working ! http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c

[Daedalus01]

hi, i was wondering if you guys could help me out on a very irritating issue?

i love ffdshow's audio decoder and its ability to normalise / normalize ! it makes the dialogue easier to hear at a volume such that loud sounds dont deafen me!

however, i am having trouble with it when re-encoding a 5.1 source to ac3 5.1 and passing via spdif optical. The normalisation will work but after a while (5-10min, or i think when a scene changes) it will stop working. if i seek to a different part of the video file it works again.

i am using the latest ffdshow.

does anyone know why?

thanks!

[clsid]

@Mercury_22
You need this tool:
http://technet.microsoft.com/en-us/sysinternals/bb896647.aspx

[Mercury_22]

@Mercury_22
You need this tool:
http://technet.microsoft.com/en-us/sysinternals/bb896647.aspx
Thanks !
Here it is http://sharebee.com/e1c4b8ba

Can someone compile revs 3055, 3056 and 3057 x64 version so I can pinpoint the exact rev where it got broken ?

[ale5000]

ffdshow 3054 is able to load this subtitle: http://waarokkufansub.googlepages.com/WaarokkuSetonoHanayome-01.ass
but it is unable to load this: http://waarokkufansub.googlepages.com/WaarokkuSetonoHanayome-02.ass

[mark0077]

Anyone know how to make the "grab" feature grab all frames that are output from the yadif de-interlacer in double framerate mode

ie grab double the amount of frames as would normally be grabbed.

At the moment it currently skips every second frame output from yadif. Its like the grab is being triggered from something in the chain that hasn't got the double frame rate, perhaps as the frames are decoded grab gets triggered?

Hopefully someone can update this to make grab be triggered from whatever was before it in the chain, either yadif, avisynth, whatever... as they can possibly change frame rate.

[yesgrey]

Anyone know how to make the "grab" feature grab all frames that are output from the yadif de-interlacer in double framerate mode
Have you tryed with an avisynth script using ImageWriter function?

[mark0077]

Nope, I could but just wanted to point this out to ffdshow developers as a potential issue with grab. Also grabbing to BMP doesn't give a clear indication as to what type of YV12 -> RGB conversion is done for example. I am hoping its using my RGB Conversion settings, ie high quality, dithering on, assume full range output, limited range input etc.

[rack04]

Will the H.264/AVC codecs need to be updated to decode x264 encodings using multi-slices?

[clsid]

Slices are already supported afaik.

[leeperry]

@leeperry
Post a detailed bugreport on sourceforge. Otherwise it will certainly be forgotten.
ok, just did : https://sourceforge.net/tracker/?func=detail&aid=2834209&group_id=173941&atid=867360
this bug also happens on DVD MPEG2, as GinSonic said: http://forum.doom9.org/showpost.php?p=1312403&postcount=7785

each time I wanna open a DVD, I have to disable resizing in ffdshow, then reenable it once the movie is playing...otherwise HR crashes MPC, and EVR/VMR9 zoom the picture to death :o

http://thumbnails8.imagebam.com/4749/22212247489685.gif (http://www.imagebam.com/image/22212247489685)

[her34]

would an option be added to audio decoder, "delay" filter, so that all channels are controlled by single entry box? something like an extra entry row with checkbox that will grey out individual channels

[Mark_A_W]

Is is possible to create a shortcut to the offline configuration properties for the Audio Processor filter?

Out of the box ffdshow has shortcuts for the Video Decoder and Audio Decoder.


I figured out the Raw Video Filter shortcut (configureRaw).

But I can't figure out one for the Audio Processor (tried configureAudioProcessor, configureProcessor, configureAudioProc, configureProc, etc)


Help please!

TIA

Mark

[jmartinr]

this bug also happens on DVD MPEG2, as GinSonic said: http://forum.doom9.org/showpost.php?p=1312403&postcount=7785

each time I wanna open a DVD, I have to disable resizing in ffdshow, then reenable it once the movie is playing...otherwise HR crashes MPC, and EVR/VMR9 zoom the picture to death :o

http://thumbnails8.imagebam.com/4749/22212247489685.gif (http://www.imagebam.com/image/22212247489685)

Rev 3064 seems to be OK. In 3062 some changes were reverted. I had some WMV's that had the same problem but play fine now. Did you check rev 3064?

[leeperry]

ahhh, lemme try :)

xvidvideo.ru doesn't make ICL10 builds anymore, and those on sourceforge are old as well..

[Mercury_22]

ahhh, lemme try :)

xvidvideo.ru doesn't make ICL10 builds anymore, and those on sourceforge are old as well..
Try XhmikosR's http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c

[leeperry]

ah great, it seems to be fixed indeed! it wasn't clear from the SVN updates commentaries, and getting the newest builds in ICL10 is not an easy task...but I can play DVD again! :thanks:

PS: I will close my bug report after some extensive testing.

[leeperry]

http://www.nanoed.org/courses/carbon_nanotube/ballv02.mpg

http://thumbnails19.imagebam.com/4753/9e4e7347524114.gif (http://www.imagebam.com/image/9e4e7347524114)

ah well, it works...who cares :o

[FoLLgoTT]

Can anyone confirm that when outputting LPCM directly from the gabest splitter (or MPC-HC) the right rear channel gets a wrong/defect signal? With TrueHD everything is fine.

It seems to be no problem when the decoder of MPC-HC decodes anything to LPCM and FFDSHOW outputs to the renderer. In this case the signal of the right rear channel is correct. A good Blu-ray to test is "300", because it has LPCM, TrueHD and AC3 audio tracks.

[anyone23]

Somehow the auto deinterlacing of ffdshow doesn't work here for a few TV channels.

Example: http://www.file-upload.eu/download-1866342/ffdshow_auto_deinterlacing.m2v.html

Open the file in DGIndex and analyze it. After about 14 seconds, the frames are flagged as progressive but they are in fact interlaced.

The Cyberlink MPEG-2 Codec plays the file flawless and on the televisor the deinterlacing works too.

My ffdshow settings: http://www.pic-upload.de/view-3000528/ffdshow_deinterl.png.html

Catalyst settings: Vector adaptive deinterlacing without pulldown detection, ATI HD 4870 PCIE, Catalyst 9.8.

This happens with all the interlaced content of this TV channel.

Is this a bug of ffdshow?

[thewebchat]

I would like to point out a bug about mod4 width video being sent from ffdshow to intermediary filters before the renderer:

http://sourceforge.net/tracker/?func=detail&aid=2849620&group_id=173941&atid=867360

[clsid]

@anyone23
Enable the "Process frames flagged as progressive" option.

[SledgeHammer_999]

I want to report a bug of ffdshow with subs. Just create a mkv file with this track layout:
1.subtitle
2.video
3.audio

The subtitle track isn't picked up by ffdshow.

[anyone23]

@anyone23
Enable the "Process frames flagged as progressive" option.
You mean "Methods: Bob"?

Yeah, but with that option everything gets deinterlaced even content correctly flagged as progressive.

Isn't there a way to make hardware deinterlacing like in Cyberlink (PowerDVD) Decoder because here the auto option works?

[STaRGaZeR]

You mean "Methods: Bob"?

Yeah, but with that option everything gets deinterlaced even content correctly flagged as progressive.

Isn't there a way to make hardware deinterlacing like in Cyberlink (PowerDVD) Decoder because here the auto option works?

No, because Haruhiko (developer) specifically designed it that way. It's the channel responsability to flag the content as they should. I was against this decision precisely because of DVB issues, but I understand his position. You can create a profile with "Process frames flagged as progressive" option checked and set autoload conditions for this kind of streams.

[clsid]

I want to report a bug of ffdshow with subs. Just create a mkv file with this track layout:
1.subtitle
2.video
3.audio

The subtitle track isn't picked up by ffdshow.
You should put the subtitles last. Older versions of MPC also do not like subs as the first track.

[SledgeHammer_999]

You should put the subtitles last. Older versions of MPC also do not like subs as the first track.

Well it is still a bug of ffdshow. I don't think there's an explicit guideline for subtitle tracks being last. It's just a convention. Furthermore if I enable the internal subtitle renderer in MPC the subs are detected.

MPC = 6.4.9.1(rev 104)

[clsid]

Yes, it is a bug that should be fixed. I just wanted to say that triggering it can be avoided.

Please add it to the BugTracker (http://sourceforge.net/tracker/?group_id=173941&atid=867360).

[khagaroth]

Is is possible to create a shortcut to the offline configuration properties for the Audio Processor filter?

Out of the box ffdshow has shortcuts for the Video Decoder and Audio Decoder.


I figured out the Raw Video Filter shortcut (configureRaw).

But I can't figure out one for the Audio Processor (tried configureAudioProcessor, configureProcessor, configureAudioProc, configureProc, etc)


Help please!

TIA

Mark

You almost got it, it's configureAudioRaw.

[albain]

Can anyone confirm that when outputting LPCM directly from the gabest splitter (or MPC-HC) the right rear channel gets a wrong/defect signal? With TrueHD everything is fine.

It seems to be no problem when the decoder of MPC-HC decodes anything to LPCM and FFDSHOW outputs to the renderer. In this case the signal of the right rear channel is correct. A good Blu-ray to test is "300", because it has LPCM, TrueHD and AC3 audio tracks.

I don't know about mpc-hc but there was a recent fix on LPCM, TrueHD, EAC3 channels mapping on FFDShow

[FoLLgoTT]

I don't know about mpc-hc but there was a recent fix on LPCM, TrueHD, EAC3 channels mapping on FFDShow

I use revision 3063 which already includes this fix. It must be a different problem. The mapping is correct (I checked this). Only the right rear outputs a garbled/broken signal with LPCM directly from the splitter.

[iron2000]

Not sure what happened but MPC-HC doesn't use ffdshow for audio after installing 3065.
Going back to 3064 to see if problem persists.

EDIT: ffdshow is back on 3063.

[betaking]

Not sure what happened but MPC-HC doesn't use ffdshow for audio after installing 3065.
Going back to 3064 to see if problem persists.

EDIT: ffdshow is back on 3063.

last mpc-hc 1271 svn i can use ffdshow 3065 for audio!
For those who have problems loading ffdshow, you need to add "mpc-hc.exe" to the whitelist in ffdshow settings (DirectShow control page).

[XhmikosR]

I think that this is fixed in 3067, if I understand the changes correctly.

Download (http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c).

[betaking]

I think that this is fixed in 3067, if I understand the changes correctly.

Download (http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c).

installing you build no problem!:thanks:

[iron2000]

Thanks XhmikosR.
3067 is back to normal behavior.

EDIT: Ok found something off:
http://shup.com/Shup/211837/huh.jpg

[_xxl]

EDIT: Ok found something off:
I have this issue with ICL11 and forced loop unrolling. ex:/Qunroll10

[XhmikosR]

Alright I'll use that in my future ICL 11 builds.

[_xxl]

http://www.afterdawn.com/software/video_software/codecs_and_filters/ffdshow.cfm
http://i30.tinypic.com/2uh4dph.jpg
Since when is rev 610 latest stable release of ffdshow tryouts?

[_xxl]

Alright I'll use that in my future ICL 11 builds.
ex:/Qunroll10
No, MSVC builds are fine, no need for ICL.

[XhmikosR]

Since I have already set up a batch file for this, there is no problem for me to also compile the ICL11 build along with the MSVC build. Unless there is something wrong with the ICL builds.

[ale5000]

No, because Haruhiko (developer) specifically designed it that way. It's the channel responsability to flag the content as they should. I was against this decision precisely because of DVB issues, but I understand his position. You can create a profile with "Process frames flagged as progressive" option checked and set autoload conditions for this kind of streams.
Can it be added but disabled by default?

[Snowknight26]

Does ffdshow not unload the system tray icon properly? After doing some batch encoding, my tray looked like this:

http://i27.tinypic.com/2r4jrjn.jpg

[Mercury_22]

Does ffdshow not unload the system tray icon properly? After doing some batch encoding, my tray looked like this:

http://i27.tinypic.com/2r4jrjn.jpg
I don't think it's FFD's fault see http://forum.doom9.org/showthread.php?t=149134

[Snowknight26]

Separate issue as it has nothing to do with Explorer's thumbnails/integraiton. Maybe I should have clarified that after doing some batch encoding with x264, with a graph of Haali Media Splitter -> ffdshow, even after x264 is no longer in the process list, the icons still remain.

Similarly, if I restart explorer.exe, the ffdshow system tray icons disappear if the filters are currently loaded by whatever program.

[Mark_A_W]

You almost got it, it's configureAudioRaw.



Thank you!!

:thanks:

[ryrynz]

Hi Guys,

When I click export all settings and then add the registry file on a fresh install, only a few of my avisynth scripts copy over.
I have 14 right now and only 3 copied across, the rest of the settings are transferred over fine. The avisynth scripts are in the registry file created so I just view the file and manually copy them across. I have used hundreds of builds and the problem has always existed for me.

Regards,
Ryan.

[Mark_A_W]

Mercury, because I have had trouble with both rear outputs with my soundcard and setup (some form of fuzzy mixing between the two) you have sparked my curiosity. I post here because I notice in ffdshow, in the mixer section, the "5 channels" output configuration uses Back L and Back R, where I would have always assumed the rear channels of a 5.1 track would be Side L and Side R. I have my rear speakers setup as per dolby recommendations, at the side, slightly to the rear.

So confused now as to what each piece of the chain intends, the original track, the software like ffdshow, and finally the drivers.


I've just moved to 7.1 and am having the same issue as Mark077.

When I play 5.1 over 7.1 speakers, it comes out the back surrounds. Not the sides, as I expected.

I've been reading through these channel mapping posts, and I will check my Windows speaker settings, but I don't think there is anything "wrong".

With 7.1 the backs are channels 4 and 5 and the sides are 6 and 7.

When you play a 5.1 soundtrack you get channels 4 and 5 going to the Backs, as that's what channel 4 and 5 map to.


It would be nice if ffdshow could have an option to override this, swapping the sides and backs when 5.1 channels are detected when the mixer is set to 7.1. (It detects 2 channel for the DPL2 processing.)


Actually, it would be REALLY NICE if the upmix from 2 channel to 5.1 channel performed by the DPL2 setting could be extended to upmix 5.1 to 7.1!!!


(Also, the channel selections in the Mixer are confusing - specifically the LFE button. I think it would be more user friendly if the LFE button disappeared and the channel descriptions used the more standard "5.1" or "7.1", rather than 5 channels + LFE ticked, or 7 channels + LFE ticked. It's very easy to get wrong.)


Sorry for meandering on...

Mark

[STaRGaZeR]

Can it be added but disabled by default?

I suggested an option for "agressive deinterlacing" or something like that, for flagging all content suspected of being interlaced, but it was not accepted for the same reason as above. Your only option with ffdshow is using autoloaded profiles or change them manually.

[Mercury_22]

I've just moved to 7.1 and am having the same issue as Mark077.

When I play 5.1 over 7.1 speakers, it comes out the back surrounds. Not the sides, as I expected.

I've been reading through these channel mapping posts, and I will check my Windows speaker settings, but I don't think there is anything "wrong".

With 7.1 the backs are channels 4 and 5 and the sides are 6 and 7.

When you play a 5.1 soundtrack you get channels 4 and 5 going to the Backs, as that's what channel 4 and 5 map to.


It would be nice if ffdshow could have an option to override this, swapping the sides and backs when 5.1 channels are detected when the mixer is set to 7.1. (It detects 2 channel for the DPL2 processing.)


Actually, it would be REALLY NICE if the upmix from 2 channel to 5.1 channel performed by the DPL2 setting could be extended to upmix 5.1 to 7.1!!!


(Also, the channel selections in the Mixer are confusing - specifically the LFE button. I think it would be more user friendly if the LFE button disappeared and the channel descriptions used the more standard "5.1" or "7.1", rather than 5 channels + LFE ticked, or 7 channels + LFE ticked. It's very easy to get wrong.)


Sorry for meandering on...

Mark

There is a bit of confusion between the physical positioning of the speakers and the "audio" positioning
AFAIK 5.1 setup has REAR NO SIDE speakers although for this speaker setup some drivers are exposing the side speakers (usually those with dolby)
The most important thing is to check the labels on your sound card's speakers connectors / "holes" to match the labels of the speakers and windows's speaker setup
Your setup sound fine to me have you use any of my test sample files ?

Separate issue as it has nothing to do with Explorer's thumbnails/integraiton. Maybe I should have clarified that after doing some batch encoding with x264, with a graph of Haali Media Splitter -> ffdshow, even after x264 is no longer in the process list, the icons still remain.

Similarly, if I restart explorer.exe, the ffdshow system tray icons disappear if the filters are currently loaded by whatever program.
When they appear there are ONLY ffdshow's system tray icons or Haali's / mkv too ?

[tetsuo55]

Here is a quote from my last post on the subject, this is how it should work:
There is only 1 correct way to set up speakers.

Regardless of how many speakers you have, you should always set them up in this order, having less speakers simply means that a certain speaker is not used.

NOTE: for 6.1 speakers LB is used.(in this configuration the speaker would be placed in the back-center)
NOTE2: Dolby suggests moving the LS and LR speaks slightly more to the back in 5.1 setup

http://img442.imageshack.us/img442/9582/map.th.png (http://img442.imageshack.us/i/map.png/)

Based on the diagram(ignoring the sub):

1 = L
2 = L,R
3 = L,R,C
4 = L,R,LS,RS
5 = L,R,C,LS,RS
6 = L,R,C,LS,RS,LB
7 = L,R,C,LS,RS,LB,RB

L = Left (front)
R = Right (front)
C = Center (front)
LS = Left Surround Side (often named "back" in 5.1 setup)
RS = Right Surround Side (often named "back" in 5.1 setup)
LB = Left Surround back (This is the "back" or "center back" or "surround center" 6.1 setup)
RB = Right Surround back
S = Subwoofer (often named LFE, this word has several meanings, and is best ignored for speaker setup)

Edit:
Fixed naming, now all the names are industry standard.

If windows does not follow these scheme, we should
A file a bug report
B add a workaround to ffdshow/mpc-hc

[Mark_A_W]

After much discussion with my CRT projector friends, we came to same conclusion as Tetsuo.


When you set a 7.1 channel receiver to 5.1 you get the side speakers.

When I set my Xonar Deluxe soundcard to 5.1, I get the side speakers.


But when playing 5.1 audio over 7.1 speakers, on a PC we get the back channels.

Windows does not seem to behave properly.

An option in the Ffdshow Mixer to work around this would be much appreciated. Or even better, proper 5.1 -> 7.1 upconversion ;)


Mercury, I've been testing using the Microsoft 6 and 8 channel test files. I haven't seen a link to your test files as I browsed though the posts...gotta link?


Mark

[tetsuo55]

After much discussion with my CRT projector friends, we came to same conclusion as Tetsuo.


When you set a 7.1 channel receiver to 5.1 you get the side speakers.

When I set my Xonar Deluxe soundcard to 5.1, I get the side speakers.


But when playing 5.1 audio over 7.1 speakers, on a PC we get the back channels.

Windows does not seem to behave properly.

An option in the Ffdshow Mixer to work around this would be much appreciated. Or even better, proper 5.1 -> 7.1 upconversion ;)


Mercury, I've been testing using the Microsoft 6 and 8 channel test files. I haven't seen a link to your test files as I browsed though the posts...gotta link?


MarkCorrect, the post i made is industry standard, all certified standalone units follow these specs.
It's probably a typo by the windows guys

[Mercury_22]

Here is a quote from my last post on the subject, this is how it should work:


If windows does not follow these scheme, we should
A file a bug report
B add a workaround to ffdshow/mpc-hc

It's not the windows the drivers are exposing the side or the back speakers http://www.freelists.org/post/wdmaudiodev/Vista-51-speaker-mode-question,7 If a
>> system vendor has set the pin config values of the integrated HD
>> Audio device to indicate that the 5.1 device uses side speakers
>> then we expose that to the OS when it asks the driver through the
>> channel config property request but if the system vendor uses the
>> pin config sequence values in the multi-channel device association
>> to indicate that the 5.1 system has rear speakers the HD Audio
>> class driver will expose that to the OS as the channel mask for
>> that device instead.

http://www.microsoft.com/whdc/device/audio/PinConfig.mspx
Similarly, it uses the label 5.1 surround sound speakers to identify pairs of configurations that differ only in whether BL and BR speakers are used instead of SL and SR speakers. The reason for not distinguishing between the back-speaker and side-speaker configurations in either case is that home users tend not to distinguish between these speaker positions—the placement of furniture in the room might be the primary factor in determining whether a pair of speakers ends up beside or behind the listener. Forcing users to recognize the subtle differences between these alternate configurations would complicate the user interface for little benefit.

[tetsuo55]

:scared:

That means we have to adjust ffdshow/mpc-hc to work around that problem
what a pita

they are making stupid assumptions where industry standard cannot be any clearer.....

[Mercury_22]

After much discussion with my CRT projector friends, we came to same conclusion as Tetsuo.


When you set a 7.1 channel receiver to 5.1 you get the side speakers.

When I set my Xonar Deluxe soundcard to 5.1, I get the side speakers.


But when playing 5.1 audio over 7.1 speakers, on a PC we get the back channels.

Windows does not seem to behave properly.

An option in the Ffdshow Mixer to work around this would be much appreciated. Or even better, proper 5.1 -> 7.1 upconversion ;)


Mercury, I've been testing using the Microsoft 6 and 8 channel test files. I haven't seen a link to your test files as I browsed though the posts...gotta link?


Mark

Test 5.1 AC3 (Dolby Digital) (http://www.tfm.ro/win32-projects/test-avi-ac3/)
Test 7.1 E-AC3 (Dolby Digital Plus) (http://sharebee.com/69772daa)
Test 7.1 Dolby TrueHD (http://www.megaupload.com/?d=7P9P81CW)
Test 7.1 DTS-HD HR (http://sharebee.com/abf92989)
Test 7.1 LPCM (http://www.megaupload.com/?d=0HM2JCEP)

EDIT:
So, about the back or side speakers setup,the way you physically setup your speaker it's as you please, but the purpose being for you to hear the sound exactly the way it was intended by the "author", my recommendation to all those using 5.1 speakers is to use the 7.1 speaker setup in windows and to connect the surround speakers to the back speakers connector on your sound card to avoid any problems created by your drivers config cause :

Similarly, it uses the label 5.1 surround sound speakers to identify pairs of configurations that differ only in whether BL and BR speakers are used instead of SL and SR speakers. The reason for not distinguishing between the back-speaker and side-speaker configurations in either case is that home users tend not to distinguish between these speaker positions—the placement of furniture in the room might be the primary factor in determining whether a pair of speakers ends up beside or behind the listener. Forcing users to recognize the subtle differences between these alternate configurations would complicate the user interface for little benefit. (http://download.microsoft.com/download/9/c/5/9c5b2167-8017-4bae-9fde-d599bac8184a/PinConfig.doc)

Assuming your sound card support up to 8 channels :)

[Mark_A_W]

Thanks for the links to the test files.


We do seem to be talking about slightly different things.


My issue occurs when you have 7.1 speakers connected, and 7.1 speakers active in the soundcard driver's control panel....then play a 5.1 channel soundtrack. That's when the surround channels come out the backs instead of the sides.

I think you are talking about setting up 5.1 speakers using 7.1 windows/soundcard settings, which is not my setup.


If I manually intervene and set the soundcard to 5.1, I then get the surround channels coming out the sides. But I'd prefer not to do that.


From the reading I've done, it really does seem like an option in ffdshow to workaround this issue would be nice (a button that appears when 7 channels is selected in mixer (and only then), which says "Swap Surrounds for 5.1 signal Side/Back" or somesuch. Or better, 5.1 to 7.1 Prologic II.

[Mercury_22]

Thanks for the links to the test files.


We do seem to be talking about slightly different things.


My issue occurs when you have 7.1 speakers connected, and 7.1 speakers active in the soundcard driver's control panel....then play a 5.1 channel soundtrack. That's when the surround channels come out the backs instead of the sides.

I think you are talking about setting up 5.1 speakers using 7.1 windows/soundcard settings, which is not my setup.


If I manually intervene and set the soundcard to 5.1, I then get the surround channels coming out the sides. But I'd prefer not to do that.


From the reading I've done, it really does seem like an option in ffdshow to workaround this issue would be nice (a button that appears when 7 channels is selected in mixer (and only then), which says "Swap Surrounds for 5.1 signal Side/Back" or somesuch. Or better, 5.1 to 7.1 Prologic II.

As I've said above ...
Your setup sound fine to me have you use any of my test sample files ?...
Using the 5.1 test file your channels don't seem to be correct mapped ?
Again AFAIK 5.1 sound track has back channels not side Dolby recommendation it's referring to physical positioning

[ryrynz]

Hi guys,

When you select a profile/preset settings window pane it doesn't update the selected profile when you click on the notification icon.

Regards,
Ryan.

[Mark_A_W]

As I've said above
Using the 5.1 test file your channels don't seem to be correct mapped ?
Again AFAIK 5.1 sound track has back channels not side Dolby recommendation it's referring to physical positioning


These test files give the same result as the Windows ones (of course!).

As expected, when playing the 5.1 test track over 7.1 speakers, the BACK channels are active.

It should be the SIDE channels.

[Mercury_22]

These test files give the same result as the Windows ones (of course!).

As expected, when playing the 5.1 test track over 7.1 speakers, the BACK channels are active.

It should be the SIDE channels.

Well let's agree to disagree, :) as the video shows I say
It should be the BACK channels

EDIT: If you use any other codec what channels / speakers are used ?

[mark0077]

Mark_A_W, I think you are also using a Xonar HDAV 1.3 Deluxe right?

Did you ever manage to sort the upmixing of channels on it? I find it hard to get a combination of setting that stops upmixing of channels, or what seems like bleeding of side channels into rear channels on surround tests.

[Mark_A_W]

Mark_A_W, I think you are also using a Xonar HDAV 1.3 Deluxe right?

Did you ever manage to sort the upmixing of channels on it? I find it hard to get a combination of setting that stops upmixing of channels, or what seems like bleeding of side channels into rear channels on surround tests.


Yes, a Xonar HDAV 1.3 Deluxe, on Vista32, analogue out to poweramps, using the 1764 driver (maybe the 1771 W7 driver...I forget). All settings OFF except Flexbass, Splendid crap not installed, just the base driver.

I've never had your channel bleeding issue. All channels are discrete, always, when playing 7.1 channel test files.

For the upmixing issue, I simply send it 7.1 channels using ffdshow audio, for all material. It thinks it's getting 7.1, so never upmixes, even for 2 channel material.

If I want to turn 2 channel into 5.1, I use ffdshow audio's Dolby Prologic II setting (which would be nice if it would upmix 2 channel to 7.1, and 5.1 to 7.1!!!! HINT HINT!!).

[mark0077]

hah excellent, yeah ffdshows mixer was a lifesaver for me. Would be good to get all of this rear versus side channels stuff cleared up.

[ryrynz]

XhmikosR I did a quick test with one of my scripts of your icl11 build vs clsid's builds and didn't see any real difference in performance. Just thought I'd pass that on, cheers.

[XhmikosR]

And that's the expected behavior. In general the performance improvement is negligible. Anyway, I only have ICL11 that's why I build this version.

Although I'm using ICL v11.0.072 for my builds, I thought I'd try to use the latest ICL v11.1.038 but something is not working, I don't see ffdshow audio to be used on Win 7.

If someone has the latest version of ICL10, I'd like to give it a go to see if this (http://forum.doom9.org/showthread.php?p=1322542#post1322542) problem is also solved, because with ICL11 even if I use /Qunroll10 I get the same results as without using it.

[Mercury_22]

And that's the expected behavior. In general the performance improvement is negligible. Anyway, I only have ICL11 that's why I build this version.

Although I'm using ICL v11.0.072 for my builds, I thought I'd try to use the latest ICL v11.1.038 but something is not working, I don't see ffdshow audio to be used on Win 7.

If someone has the latest version of ICL10, I'd like to give it a go to see if this (http://forum.doom9.org/showthread.php?p=1322542#post1322542) problem is also solved, because with ICL11 even if I use /Qunroll10 I get the same results as without using it.

The audio has been broken on x64 versions somewhere between 3054 and 3058

...

EDIT: Last (tested) working version is 3054, first (tested) broken 3058

[XhmikosR]

You mean only with my builds? Because I haven't changed anything.

[Mercury_22]

You mean only with my builds? Because I haven't changed anything.
NO ! All builds !
Alexins also reported the same problem but in ICL10
Replacing the file does not change anything. The same problem exists with the ffdshow x86 icl10 3056-3063. and I can confirm that, tested with ffdshow_rev3065_20090902_clsid_icl10

The bug has been introduced most likely in 3056 and it's in ffdshow.ax !

The libavcodec.dll (3063 replaced with 3054) trick has no result = no changes and I can't find the allowDPRINTF key I even did a full registry search for allowDPRINTF It should be the same path for the x86 version on a x86 windows so if you can be more precisely (HKLM\Software\GNU ? HKCU\Software\GNU ? or ?)

EDIT: Replacing the ffdshow.ax with the old one (3054) does the trick = ffd it's connectable (working)

[XhmikosR]

There is nothing I can do since as you say there is something wrong in the code. I was talking about compiler specific things.

[clsid]

Hopefully Albain can track down that bug.

[Mercury_22]

Hopefully Albain can track down that bug.
ICL10 and x64 audio are not working only on Win7 ?

[clsid]

It also doesn't work with Vista x64.

[albain]

It also doesn't work with Vista x64.

OK but according to that post it works again in recent revisions :
Ok I've solved the allowDPRINTF key "mystery" first it has to be created by checking the "Output strings to debugger" but I have no log file (maybe because FFD audio it's not connecting )where should it be? C:\Program Files\ffdshow ? or C:\ ? or...?

EDIT: ICL11 rev 3063 it's working ! http://www.mediafire.com/?sharekey=3f33c77c2cf9ce251686155677bb26855284166a6fc12f6c

I am not sure to understand the problem, I am working on it

@Mercury_22 : when you posted the DPRINTF traces, with which revision you did your tests ? if revision was < 3059, could you do the test again with more recent revision and post the traces ?
From what I can see the loaded decoder was liba52, do you send the stream to spdif or make it decode internally by liba52 ?

Thanks

[Mercury_22]

OK but according to that post it works again in recent revisions :


I am not sure to understand the problem, I am working on it

@Mercury_22 : when you posted the DPRINTF traces, with which revision you did your tests ? if revision was < 3059, could you do the test again with more recent revision and post the traces ?
From what I can see the loaded decoder was liba52, do you send the stream to spdif or make it decode internally by liba52 ?

Thanks

DPRINTF traces for ffdshow_rev3069_20090907_x64 (http://sharebee.com/5d3a554c)

The audio in FFDShow x64 and ICL10 (I've tested ffdshow_rev3065_20090902_clsid_icl10) versions it's not working !!! = even that the FFDShow's Audio merit it's at default = the highest on my system other filters get connected
This is happening to any audio type LPCM AC3 DTS TRUEHD...
Replacing ffdshow.ax with the one from rev 3054 x64 (for the x64 version of course ) it's solving the problem !

P.S. ffdshow_rev3069_20090907r_icl11 and ffdshow_rev3069_20090907 (x86) are working normally

[albain]

OK thanks, I am anlyzing it

Meanwhile I committed some fixes to definitely remove any traces of the windows 7 structure as I have so far no answer from microsoft. Maybe this will solve the problem

Revision 3071

I will try to post you ffdshow.ax only for 64 bits for that revision

*EDIT* Here is ffdshow.ax x64 revision 3071 compiled under MSVC 2008. Don't take in account the UI changes in the output section of FFDSHow audio
ffdshow.ax - 4.49MB (http://www.zshare.net/download/65312328200c900b/)

Copy it in program files and with an admin dos window just type :
regsvr32 /s /u ffdshow.ax
regsvr32 /s ffdshow.ax

[Mercury_22]

OK thanks, I am anlyzing it

Meanwhile I committed some fixes to definitely remove any traces of the windows 7 structure as I have so far no answer from microsoft. Maybe this will solve the problem

Revision 3071

I will try to post you ffdshow.ax only for 64 bits for that revision

*EDIT* Here is ffdshow.ax x64 revision 3071 compiled under MSVC 2008. Don't take in account the UI changes in the output section of FFDSHow audio
ffdshow.ax - 4.49MB (http://www.zshare.net/download/65312328200c900b/)

Copy it in program files and with an admin dos window just type :

It's working ! :)

[albain]

It's working ! :)



:) Glad to "hear" !

[alexins]

albain
:thanks: