ffdshow tryouts project: Discussion & Development

[fastplayer]

98 support was dropped towards the end of 2007. :search:
I don't need to search. Haruhiko brought back 98 support a few builds after he dropped it by using unicows (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout?view=rev&revision=1668).

Anyway, let's see when the first Win98 users start to complain about missing 98 support :D

Edit: 3000th post in this thread :eek:

[Inventive Software]

Oh bugger, I shoulda known about that, cause I suggested it to him! :D How silly of me! :o

[Thunderbolt8]

is there any way to get to know to which changes/updates of libavcodec the number in brackets e.g. (11539) refer?

[fastplayer]

is there any way to get to know to which changes/updates of libavcodec the number in brackets e.g. (11539) refer?
The comments are pretty short and technical:
http://lists.mplayerhq.hu/pipermail/ffmpeg-cvslog/2008-January/thread.html

[clsid]

Not without comparing the actual changes. It's too much work to specify all applied ffmpeg revisions. So I stopped doing that.

[Kado]

Audio filter bug: winamp2 filter
"click to select winamp 2 directoty..." button is not working, it doesnt pop up a windows and let me choose directory, when configure it offline.
it does work for me when im playing video/audio, but still, it doesnt show up the plugin(Signal processing studio DSP, yes i select winamp5 as dir, not the plugin dir)

In my case if I try to configure it through mpc external filter box I cannot change it, but if I go via windows start menu I can change it.
That plugin is probably blacklisted because I have others like enhancer and dfx and they show up in the list.
Even by changing the dsp plugin name won't work.

[haruhiko_yamagata]

Audio filter bug: winamp2 filter
"click to select winamp 2 directoty..." button is not working, it doesnt pop up a windows and let me choose directory, when configure it offline.
it does work for me when im playing video/audio, but still, it doesnt show up the plugin(Signal processing studio DSP, yes i select winamp5 as dir, not the plugin dir)
It works for me. Which build do you use?

[Yong]

It works for me. Which build do you use?
clsid r1787,
sorry, the bug i encountered is same as kado one. :p

Yup, SPS DSP is not show up, but other DSPs are working.

[haruhiko_yamagata]

Excluded WinAmp plugins are
// DSP stacker, Adapt-X and Vst host are not compatible with ffdshow currently. Maybe ffdshow's bug, but I can't help...
if ( _strnicmp(_l("dsp_stacker.dll"),filename,16)==0
|| _strnicmp(_l("dsp_adaptx.dll"),filename,15)==0
|| _strnicmp(_l("dsp_vst.dll"),filename,12)==0
)
return;

[cc979]

i just tried compiling svn 1790 (because of 1788) with 'make UNICODE=yes' but it fails on makeAVIS

[haruhiko_yamagata]

i just tried compiling svn 1790 (because of 1788) with 'make UNICODE=yes' but it fails on makeAVISmakeAVIS is an ANSI application. Unicode build is not supported.

[Inventive Software]

Nice joke leeperry! :D

I think I found a bug in ffdshow's MPEG-2 decoder.

*breathe*

MPEG-2 video in Matroska with the Aspect Ratio set doesn't display correctly in Media Player Classic 6.4.9.1 or WMP11 (the only 2 media players I have). I tried both libmpeg2 and libavcodec, both wouldn't give me a correct aspect ratio (though ffdshow's Info panel gives the correct SAR and DAR), and I have no idea whether it's the splitter (Haali Media Splitter) or the decoder because using MPC's internal MPEG-2 decoder it displays correctly.

*snip*

Update: I've found the same bug with the same elementary streams. So it's not parser related.

@SMD: Regarding your workaround, that does work, but only because MPC falls back on it's internal MPEG-2 decoder (even though it's specifically disabled).

[Yong]

Excluded WinAmp plugins are
// DSP stacker, Adapt-X and Vst host are not compatible with ffdshow currently. Maybe ffdshow's bug, but I can't help...
if ( _strnicmp(_l("dsp_stacker.dll"),filename,16)==0
|| _strnicmp(_l("dsp_adaptx.dll"),filename,15)==0
|| _strnicmp(_l("dsp_vst.dll"),filename,12)==0
)
return;

hmm ok, i dont the SPS DSP, and the winamp2 audio filter, just reporting bug :)
should be ok if put at low priority for bug fixing.

[haruhiko_yamagata]

MPEG-2 video in Matroska with the Aspect Ratio set doesn't display correctly in Media Player Classic 6.4.9.1 or WMP11 (the only 2 media players I have). I tried both libmpeg2 and libavcodec, both wouldn't give me a correct aspect ratio (though ffdshow's Info panel gives the correct SAR and DAR), and I have no idea whether it's the splitter (Haali Media Splitter) or the decoder because using MPC's internal MPEG-2 decoder it displays correctly.
In 352 x 576, 4:3 movie, libmpeg2 doesn't seem to support aspect ratio handling. I have added a special workaround at rev 1791.
Please let me know if you find a similar case. 352 x 480 may be the case. If you can encode such file, could you send me a sample?

As for libavcodec mpeg2, I'm not willing to fix as usual.

[Inventive Software]

Thank you haruhiko. I'm just waiting for a build. I'll see about making a test 352x480 video just to see if that works or not.

[SeeMoreDigital]

SMD: Regarding your workaround, that does work, but only because MPC falls back on it's internal MPEG-2 decoder (even though it's specifically disabled).Hi mate,

I've just checked my filters. I can confirm that FFdshow is being used: -

http://i17.tinypic.com/8fmc1ef.jpg

EDIT: And the aspect ratio is correct also: -

http://i6.tinypic.com/8fb6715.jpg


Cheers

[cc979]

makeAVIS is an ANSI application. Unicode build is not supported.

maybe i misread the change logs

[ImAhNoBoDy]

When I encode a video using x264 or ffmpeg to 29.97fps, ffdshow shows that the fps is 25 even though I made the fps 29.97 during the encode.

I even checked it out in gspot and it does say 29.97fps. Is this a bug? I have other videos that weren't encoded by me and it shows the correct fps on them.

[haruhiko_yamagata]

When I encode a video using x264 or ffmpeg to 29.97fps, ffdshow shows that the fps is 25 even though I made the fps 29.97 during the encode.

I even checked it out in gspot and it does say 29.97fps. Is this a bug? I have other videos that weren't encoded by me and it shows the correct fps on them.I can't reproduce. Could you send us a sample?

[Wilbert]

bug report:

ffdshow spews to the debug output a LOT. It got bad enough that I hacked code into 1.7.X that allows OutputDebugString() to be intercepted and filtered by calling DLL. I submitted a bug on this and it looks like it already got assigned.
http://forums.virtualdub.org/index.php?act=ST&f=7&t=15093&

I guess this bug is also present in the tryouts?

[Leak]

I guess this bug is also present in the tryouts?
*cough* (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout?view=rev&revision=1770) :D

np: Uusitalo - Konevitsa (Karhunainen)

[SeeMoreDigital]

SMD: Regarding your workaround, that does work, but only because MPC falls back on it's internal MPEG-2 decoder (even though it's specifically disabled).I've done a bit more testing using different MKV muxing applications.

It would seem .MKV muxes generated by (G)DSmux behave quite differently (when played using MediaPlayer Classic/FFdshow), compared to .MKV muxes generated by AVI-Mux GUI or VirtualDubMod.

Does anybody know why there would be such a difference? I can provide short samples.


Cheers

[fastplayer]

ANSI versions of ffdshow.ax and ff_wmv9.dll are included in build 1798. I guess Win98 support is back :)

[Thunderbolt8]

I've done a bit more testing using different MKV muxing applications.

It would seem .MKV muxes generated by (G)DSmux behave quite differently (when played using MediaPlayer Classic/FFdshow), compared to .MKV muxes generated by AVI-Mux GUI or VirtualDubMod.

Does anybody know why there would be such a difference? I can provide short samples.


Cheers
there is ineed a difference between .mkv files created with gdsmux or with mkvmerge for example. for example at least with h264 files I know its somehow not possible to seek into the file, when they are created with gdsmux. for that to work you'd have to an additional mux with mkvmerge and rewrite the timestamps.
best would be to ask madshi about that difference, he knows a bit more about it.

[Jay Bee]

Any news on the VC-1 in TS AV-synch bug?

[haruhiko_yamagata]

Any news on the VC-1 in TS AV-synch bug?It has been fixed. If you still have problems, please send me a sample.

[ImAhNoBoDy]

I can't reproduce. Could you send us a sample?

I use 3GPConverter just for the GUI, but I put in the ffmpeg codes myself. I am using ffmpeg v9133

Command0=""<%AppPath%>\cores\ffmpeg" -y -i "<%InputFile%>" -title "<%Title%>" -timestamp "<%TimeStamp%>" -level 30 -vcodec h264 -coder 1 -trell 2 -partitions +parti4x4+partp8x8+partb8x8 -bufsize 128 -g 250 -s 720x480 -r 30000/1001 -b 1000k -pass 1 "<%outputfile%>.mp4""
Command1=""<%AppPath%>\cores\ffmpeg" -y -i "<%InputFile%>" -title "<%Title%>" -timestamp "<%TimeStamp%>" -level 30 -vcodec h264 -coder 1 -trell 2 -partitions +parti4x4+partp8x8+partb8x8 -bufsize 128 -g 250 -s 720x480 -r 30000/1001 -b 1000k -pass 2 -acodec aac -ac 2 -ar 48000 -ab 128k "<%Outputfile%>.MP4""

I am using Gspot 2.70a to see the fps. I am using ffdshow-rev1763_20080108_xxl currently. For the current sample I had to use an avs script to encode it cause I needed to just use the first 1000 frames. Was using directshowsource(). Either way, whenever I put the file directly into 3GPConverter, ffdshow still output the fps as 25.00 than 29.97. Strange thing is if you put "Changefps(29.97)" in the avs script, ffdshow will show the correct fps. Hope that helps.

Sample: http://www.sendspace.com/file/8b1601

[haruhiko_yamagata]

@ImAhNoBoDy
Thanks, I reproduced using MPC's internal splitter.
If I use Haali's splitter, ffdshow shows 29.97Hz.
It's just OSD problem, low priority, but should be fixed.

[ImAhNoBoDy]

oh cool, thanks.

[SBeaver]

clsid: does your latest x64 build include the h264 multicore patch? (or does any x64 build include it?)

[ShadowVlican]

i've never encountered any instability or problems playing MJPEG avi made by my camera (Canon SD700IS) so why is MJPEG disabled by default in ffdshow video decoder configuration?

[clsid]

clsid: does your latest x64 build include the h264 multicore patch? (or does any x64 build include it?)The patch was committed to SVN some time ago. My x64 build is just a regular SVN build, so yes the patch is included.

i've never encountered any instability or problems playing MJPEG avi made by my camera (Canon SD700IS) so why is MJPEG disabled by default in ffdshow video decoder configuration?Afaik MJPEG is enabled by default in the installer.

[ShadowVlican]

Afaik MJPEG is enabled by default in the installer.
ok i stand corrected

apparently you need to uninstall the previous version before updating to the latest one, then you'll see the option (during setup) to enable it

if you try to install the latest version on top (without uninstalling), then it'll be off by default :)

[rickardk]

Done alot of remuxing (HD DVDs and Blu-rays) into mkv container. Some titles have glitches and rainbow frames when remuxed into mkv.

One example was Ratatouille Blu-ray (AVC). At 40min and 30s there are some strange frames. Looks scambled. When watching the joined m2ts there is no problem.
Same problem reported by other user who also tried to remux into mkv.

Remuxed Million Dollar Baby HD DVD (VC-1) yesterday. It did show the same problem. But when watching the evo no problems.

Is this a mux, codec or splitter problem?



Anyone who know?

[Inventive Software]

Combination of muxing and splitter problem. Does the problem occur with the original? :)

[rickardk]

Combination of muxing and splitter problem. Does the problem occur with the original? :)

Original evo and m2ts plays fine.
A more detailed description can be found here:

http://forum.doom9.org/showthread.php?t=133974

[red5goahead]

is it possibile to specify device monitor in use? My htpc is based in Media Portal and I use the secondary monitor but ffdshow seems from OSD to use primary monitor for resize. I set 1223x688 my custom resolution on Panasonic plasma 37pv60. Always OSD report me 1280x720 (the lcd main montor resolution). "to screen resolution size" is usefully option to use. 1223x688 seems not accepted by ffdshow, it report a red field for the horizontal value.

[Jay Bee]

It has been fixed. If you still have problems, please send me a sample.

Cool. Is it supposed to be fixed in rev1798? Because if it is I'll PM you some samples that don't work.

[SBeaver]

The patch was committed to SVN some time ago. My x64 build is just a regular SVN build, so yes the patch is included.

Thanks for confirming.
It is, however, terribly slow compared to 32bit still.

[clsid]

Thanks for confirming.
It is, however, terribly slow compared to 32bit still.
It will continue to be slow until we are able to compile a 64-bit version of libavcodec using GCC. Currently the 64-bit mingw environment is still in early development.

[fastplayer]

What does the "fast rendering" option for subtitles actually do?
All I notice is a difference in character spacing.

[Zarxrax]

I was wondering if its possible for the makeavis tool to stand on its own? That is, to be able to use it on a computer that does not have ffdshow installed? I found a page saying that makeavis was previously included with Avisynth version 2.5.2, so I'm assuming that its possible to stand on its own somehow?

[haruhiko_yamagata]

Cool. Is it supposed to be fixed in rev1798? Because if it is I'll PM you some samples that don't work.Thank you, I have received your sample.
It works for me. Is your problem same as this (http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=631)?

[haruhiko_yamagata]

What does the "fast rendering" option for subtitles actually do?
All I notice is a difference in character spacing.Well, fast rendering is faster.fast rendering: User: 5s, kernel: 0s, total: 6s, real: 6s, fps: 908.0, dfps: 903.5
normal rendering: User: 16s, kernel: 9s, total: 26s, real: 26s, fps: 218.8, dfps: 218.3
I filled all the screen with OSD for this test.

Fast rendering cashes characters. To render "abc cba", render "a" "b" "c" to cash and connect characters. It is fast, but spaces between characters are not precise.

Normal rendering simply render "abc cba" using GDI.

[Jay Bee]

Thank you, I have received your sample.
It works for me. Is your problem same as this (http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?t=631)?

Thx for testing. No, the problem you linked to seems to be a different one. Can you tell me which player and splitter you used to test the file? I've tried several combinations and AV is always out of synch until I do a seek forward into the file.

Here's the sample in case other people want to test (note: problem is only with libavcodec, not with wmv9): http://rapidshare.com/files/85756920/clip.ts.html

[fastplayer]

Well, fast rendering is faster.
No kidding! :D
I filled all the screen with OSD for this test.
So this setting applies to OSD rendering as well?
Maybe you could put that info into a tooltip. Something like:
Caches characters for improved rendering performance. Applies to OSD as well.
If disabled, GDI will be used instead.
or something similar...
It is fast, but spaces between characters are not precise.
Fortunately, the spacing can be compensated with the spacing option. The difference in performance is impressive though.

Edit: Just found a bug when using italics. Take a look at the screenshot:

http://img3.freeimagehosting.net/uploads/b203500ac2.png

And here's the SRT snippet:
745
01:06:52,265 --> 01:06:54,790
a <i>soupçon</i> of intimidation, <i>et voilà!</i>
As you can see, empty space before an opening <i>-element is rendered incorrectly with fast rendering set to on.
By the way, changing the spacing has no effect on this empty space when using italics. It just affects the characters.

[haruhiko_yamagata]

So this setting applies to OSD rendering as well?No, OSD has its own font configuration page.
Maybe you could put that info into a tooltip. Something like:
Caches characters for improved rendering performance. Applies to OSD as well.
If disabled, GDI will be used instead.
or something similar...My bad explanation. GDI is used either way. Fast rendering send one character per one call, while normal rendering send one sentence per one call.

It is fast, but spaces between characters are not precise.This is good.

Fortunately, the spacing can be compensated with the spacing option. The difference in performance is impressive though.

Edit: Just found a bug when using italics. Take a look at the screenshot:

http://img3.freeimagehosting.net/uploads/b203500ac2.png

And here's the SRT snippet:
745
01:06:52,265 --> 01:06:54,790
a <i>soupçon</i> of intimidation, <i>et voilà!</i>
As you can see, empty space before an opening <i>-element is rendered incorrectly with fast rendering set to on.
By the way, changing the spacing has no effect on this empty space when using italics. It just affects the characters.Thank you, I'll look into it.

[haruhiko_yamagata]

Thx for testing. No, the problem you linked to seems to be a different one. Can you tell me which player and splitter you used to test the file? I've tried several combinations and AV is always out of synch until I do a seek forward into the file.

Here's the sample in case other people want to test (note: problem is only with libavcodec, not with wmv9): http://rapidshare.com/files/85756920/clip.ts.htmlI'm using MPC homecinema and latest Haali's splitter. Core2duo 6700 (2.66GHz).
Is the sync completely lost or just a few frames off?

[Jay Bee]

Problem still there for me with MPC HC + all versions of Haali with AMD X2@2.8GHz.

The synch error is a constant 700ms offset (audio too late). If I set Zoom Player to playback with -700ms offset all is fine. That is until I use the seekbar because the that's when the synch error goes away. :confused:

[haruhiko_yamagata]

Problem still there for me with MPC HC + all versions of Haali with AMD X2@2.8GHz.

The synch error is a constant 700ms offset (audio too late). If I set Zoom Player to playback with -700ms offset all is fine. That is until I use the seekbar because the that's when the synch error goes away. :confused:What is your audio decoder? Do you use S/PDIF?

[Wilbert]

I was wondering if its possible for the makeavis tool to stand on its own? That is, to be able to use it on a computer that does not have ffdshow installed? I found a page saying that makeavis was previously included with Avisynth version 2.5.2, so I'm assuming that its possible to stand on its own somehow?
Sure: http://forum.doom9.org/showthread.php?p=1065297#post1065297

[Jay Bee]

What is your audio decoder? Do you use S/PDIF?

No, it says liba52. Anyone else tested the clip?

[Leak]

Sure.
Well, I guess a standalone MakeAVIS is able to write a fake AVI for an AVS script alright, but what's going to decode it? Unless I'm very wrong that part is built into ffdshow.ax, and the file you linked to doesn't contain anything that would actually decode the produced file...

np: Aesop Rock - Flashflood (Labor Days)

[haruhiko_yamagata]

No, it says liba52. Anyone else tested the clip?Oops, sorry, I can reproduce.
I couldn't reproduce just because I had already fixed the bug on my computer.
Then all I have to do is to commit my local changes. Please wait.

[Jay Bee]

Hehe, coding so fast you forgot what you coded! Looking forward to the fixed version, goodbye PDVD.

[Thunderbolt8]

madshi implemented an own way to remux VC-1 streams now to .mkv in his eac3to tool. unfortunately those files are unplayable with ffdshow libavcodec any more atm. so would it be possible for you to implement compability for that as well. or maybe talk otherwise with him if he has to change something (cant tell whats easier), it would just be nice if I could play those files with libavcodec again.

[haruhiko_yamagata]

madshi implemented an own way to remux VC-1 streams now to .mkv in his eac3to tool. unfortunately those files are unplayable with ffdshow libavcodec any more atm. so would it be possible for you to implement compability for that as well. or maybe talk otherwise with him if he has to change something (cant tell whats easier), it would just be nice if I could play those files with libavcodec again.Please ask madshi to support ffdshow.

[madshi]

Please ask madshi to support ffdshow.
Ha ha! I knew that was coming... :)

Anyway, I debugged the situation and as far as I can say the problem is in ffdshow and you guys need to fix it... ;) Let me explain what happens:

If you remux a VC-1 EVO file to MKV with gdsmux this is how the media type is reported:

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 3b 5d 06 00 00 00 00 00 ........;]......
0030: 00 00 00 00 00 00 00 00 80 07 00 00 38 04 00 00 ........€...8...
0040: 00 00 00 00 00 00 00 00 4a 00 00 00 80 07 00 00 ........J...€...
0050: 38 04 00 00 00 00 00 00 57 56 43 31 00 00 00 00 8.......WVC1....
0060: 01 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 ................
0070:|21 00 00 01 0f db fe 3b f2 1b ca 3b f8 86 f1 80 !....Ûþ;ò.Ê;ø†ñ€
0080: c9 09 a5 f5 df 07 fc 00 00 01 0e 5a df fc ef c8 É.¥õß.ü....ZßüïÈ
0090: 6c 40 l@
For 99% of all VC-1 movies this works just fine. And this is what ffdshow seems to be expecting. However, I found out that the MS VC-1 decoder goes crazy (totally garbled image) with about 1% of all VC-1 movies when reporting the media type this way. Examples are Underworld Evolution (Polish HD DVD), Phantom of the Opera (USA HD DVD) and some VC-1 HD DVD adult titles. I furthermore found out that all the problems go away if I change the media type reporting to this:

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 3b 5d 06 00 00 00 00 00 ........;]......
0030: 00 00 00 00 00 00 00 00 80 07 00 00 38 04 00 00 ........€...8...
0040: 00 00 00 00 00 00 00 00 4a 00 00 00 80 07 00 00 ........J...€...
0050: 38 04 00 00 00 00 00 00 57 56 43 31 00 00 00 00 8.......WVC1....
0060: 01 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 ................
0070:|00 00 00 01 0f db fe 3b f2 1b ca 3b f8 86 f1 80 .....Ûþ;ò.Ê;ø†ñ€
0080: c9 09 a5 f5 df 07 fc 00 00 01 0e 5a df fc ef c8 É.¥õß.ü....ZßüïÈ
0090: 6c 40 l@
It doesn't really make much sense to me that this difference makes or breaks the MS VC-1 decoder - especially cause it only affects a very small number of titles. But that's just the way it is. I got the idea to set that one byte to zero by checking what the "vc12avi" tool does. It also sets that byte to 0.

So please, modify ffdshow to accept this slightly modified media type reporting.

Thanks!! :)

[nautilus7]

and some VC-1 HD DVD adult titles.
Good argument!!! :D:D:D

[rickardk]

How does ffdshow calculate/receive the frame rate (called Movie FPS in OSD)?

I know that if the frame duration is about 41.708ms then it is 23.976fps. But ffdshow reports something else...

When remuxing HD DVDs I came up with the following:

VC-1 from HD DVD without having the timestamps rewritten:
= 25fps (BUT frame duration is 41.708ms)

VC-1 from HD DVD having the timestamps rewritten (24/1.001):
= 23.976fps (Frame duration is 41.708ms)

AVC from HD DVD without having the timestamps rewritten:
= 25fps (Frame duration is 41.708ms)

AVC from HD DVD having the timestamps rewritten (24/1.001):
= still 25fps (Frame duration is 41.708ms)

AVC stream from HD DVD demuxed and muxed with mkvmerge setting fps to 24/1.001:
= 29.97fps (Frame duration is still 41.708ms)

Based on the fact that AVC streams demuxed from Blu-ray and muxed with same settings with mkvmerge gave:
= 23.976fps (Frame duration 41.708ms)
I wrote that I thought the reported 29.97fps from HD DVD AVC had something to do with pulldown.


I know that all the above plays at 23.976fps BUT ffdshow tells me diffrent fps depending on muxing method. And it would be good if ffdshow could see them all as 23.976fps (when frame duration is 41.708ms) if you have features or macros that reads the frame rate from ffdshow. For example Media Control can now read the fps out from ffdshow and change the refresh rate of your graphic card to match the frame rate.

Don't know if decoders affect but I'm using CoreAVC and Microsofts VC-1 decoder.

[haruhiko_yamagata]

Ha ha! I knew that was coming... :)

Anyway, I debugged the situation and as far as I can say the problem is in ffdshow and you guys need to fix it... ;) Let me explain what happens:

If you remux a VC-1 EVO file to MKV with gdsmux this is how the media type is reported:Thank you for the detailed information. Could you send us the sample?

[Thunderbolt8]

Thank you for the detailed information. Could you send us the sample?
try these, these are 2 VC-1 samples which I created with the newer eac3to versions:
http://www.megaupload.com/de/?d=BPUQ477M

[madshi]

Thank you for the detailed information. Could you send us the sample?
Sure:

http://madshi.net/ffdshow.evo
http://madshi.net/ffdshow-bad.mkv
http://madshi.net/ffdshow-good.mkv

The EVO is the original (of course). The "bad" and "good" are 100% identical - apart from that one single byte I mentioned earlier. ffdshow connects to "bad" just fine, but not to "good". The MS VC-1 decoder (and also the "wmv9" option in ffdshow) totally screws up on "bad" but not on "good". The libav VC-1 decoder plays "bad" just fine, though. Might be a bug in the MS VC-1 decoder? But changing that one byte in the media type seems to work around the problem.

[haruhiko_yamagata]

Sure:

http://madshi.net/ffdshow.evo
http://madshi.net/ffdshow-bad.mkv
http://madshi.net/ffdshow-good.mkv

The EVO is the original (of course). The "bad" and "good" are 100% identical - apart from that one single byte I mentioned earlier. ffdshow connects to "bad" just fine, but not to "good". The MS VC-1 decoder (and also the "wmv9" option in ffdshow) totally screws up on "bad" but not on "good". The libav VC-1 decoder plays "bad" just fine, though. Might be a bug in the MS VC-1 decoder? But changing that one byte in the media type seems to work around the problem.Thank you. I have added support "good" files at rev 1812.
libavcodec now plays both.
wmv9 doesn't play "bad" file. I think the DMO is designed for wmv files.

[madshi]

Thank you. I have added support "good" files at rev 1812.
libavcodec now plays both.
wmv9 doesn't play "bad" file. I think the DMO is designed for wmv files.
Thank you!!

[fastplayer]

Revision 1809 - Directory Listing
Modified Sat Jan 26 00:20:13 2008 UTC (21 hours, 34 minutes ago) by h_yamagata

Fast rendering bug fix : http://forum.doom9.org/showthread.php?p=1090990#post1090990
Fast rendering : add tool tip hint.
:thanks:

[ImAhNoBoDy]

I was wondering do I need to turn on the mixer in ffdshow to hear an ac3 file with 6 channels? I know you don't with a wav file with 6 channels. I'm asking cause when I play an ac3 file with 6 channels, ffdshow says the input file is stereo even though there's 6 channels. I turn on the mixer and the ac3 file shows the input is 6 channels. With a 6 channel wav file, I don't need to turn on mixer and ffdshow shows the input as 6 channels.

[Inventive Software]

See if the same problem occurs with AC3Filter: http://ac3filter.net

[rickardk]

Seems like the mixer is broken in rev 1817 clsid (don't know how for long)

Swap channels and custom matrix does not work correctly on 8 channel uncompressed input.

Trying with a m2ts containing a 8 channel lpcm track.

I played this sample with some late 16xx or early 17xx revision with a custom matrix.

ffdshow audio says it gets 8 channel uncompressed but the only channels with audio data in them is input channel front left and input channel front center. Tested by mapping all input channels one by one through front left output channel in the custom mixer matrix.


Also tried with rev 1782 xxl. Same problem but here it is input channel front left and input channel front right (instead of center) that have audio in them. All the other ones are dead.


6 channel uncompressed works just perfect. Channels are swapable (swap channels) and mapable (via custom matrix)


Edit:
8 channel uncompressed from madFlac plays and maps perfect. Maybe a connection between haali media splitter and ffdshow that causing the mixing problem? Haali Media Splitter does output all 8 channels. Tested with dump-filter in graphedit.


Sample 8 channel LPCM in ts container:
http://www.earselect.se/sample.ts

[ImAhNoBoDy]

Ok, I downloaded ac3filter and when I played the ac3 file with 6 channels the input is 6 channels but the output is only 2 channels. I played the wav file with 6 channels and the input and output is 6 channels. I am basing all of this on the volume tab and in the "Info & CPU" tab.

I uninstalled ac3filter and played the ac3 file with the 6 channels. I turned on mixer for an output of 2/0/0 and 6 channels are active in the volume tab. I turn off the mixer and only left and right speakers showing active.

I played the wav file with the mixer on for an output of 2/0/0 and also got 6 channels active in the volume tab. I thought when you put something for output, the volume levels are suppose to change too right?

EDIT: Ugh, I remembered I downloaded matrix mixer the other day and it messed up my ffdshow settings. Anyways, after installing ac3filter it got messed up again. I uninstalled ffdshow and installed it again. 6 channels is showing now for ac3 input and output without the mixer on. Still can't get 2008-01-20 (rev. 1803) to work right in graphedit.

[clsid]

The mixer works perfectly fine for me in the latest build. And afaik nothing has changed recently to that part of ffdshow.

Use this to reset your ffdshow settings to defaults:
Windows Registry Editor Version 5.00

[-HKEY_CURRENT_USER\Software\GNU\ffdshow]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_audio]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_vfw]

[rickardk]

The mixer works perfectly fine for me in the latest build. And afaik nothing has changed recently to that part of ffdshow.

Use this to reset your ffdshow settings to defaults:
Windows Registry Editor Version 5.00

[-HKEY_CURRENT_USER\Software\GNU\ffdshow]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_audio]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_vfw]

The Mixer works perfectly for me too. But not on 8 channel LPCM. Everything else I throw at it works. Please (if you have the time) try my sample and see if you can hear all 8 channels in the mixer. Easiest way to test is to map all channels one by one in one of your output channels. Thanks!

[clsid]

My response was to ImAhNoBoDy's 6 channel problem.

You are correct about the 8 channel problem.

[rickardk]

My response was to ImAhNoBoDy's 6 channel problem.

You are correct about the 8 channel problem.

Aha ok..

I know I had a working build from you before. Gone through old builds but can't find it.

[sheppaul]

What about adding a simple reset button in the preferences of ffdshow?

[Inventive Software]

There is one! Top of the audio config box, towards the right-hand side. Should be next to a drop-down box that's labelled "default"...

[sheppaul]

What I'm saying is to remove the whole registries of ffdshow. :p

[Inventive Software]

Do it 3 times? One for the video decoder, one for the VFW config, and one for the audio decoder? :p

[haruhiko_yamagata]

Seems like the mixer is broken in rev 1817 clsid (don't know how for long)

Swap channels and custom matrix does not work correctly on 8 channel uncompressed input.

Trying with a m2ts containing a 8 channel lpcm track.

I played this sample with some late 16xx or early 17xx revision with a custom matrix.

ffdshow audio says it gets 8 channel uncompressed but the only channels with audio data in them is input channel front left and input channel front center. Tested by mapping all input channels one by one through front left output channel in the custom mixer matrix.
That sample has weird channel mask.
WAVEFORMATEXTENSIBLE:
wValidBitsPerSample: 16
dwChannelMask: 0x00000003
SubFormat: {00000001-0000-0010-8000-00AA00389B71}
dwChannelMask: 0x00000003 is impossible for 8 channels. It's better to check if dwChannelMask is valid, but I can say that the file is incomplete.

WAVEFORMATEXTENSIBLE (http://msdn2.microsoft.com/en-us/library/ms713496(VS.85).aspx)

[rickardk]

All channels are written (and in correct blu-ray lpcm order) when I connect the dump filter to haali media splitter in graph edit. Same thing with another sample from another 8channel lpcm track.

I also have samples with 6 channel LPCM. Same dwChannelMask: 0x00000003 but all 6 channels is working with ffdshow!

Any ideas why?

Maybe earlier ffdshow had a workaround or did not depend on having a valid dwChannelMask?
Im 100% sure I played 8 channels from Crank Blu-ray with one of the first ffdshow revs with 8 channel support.

I know madFlac stopped working a while back with 8 channels and madshi had to change the dwChannelMask output to get all 8 channels working with ffdshow. http://forum.doom9.org/showthread.php?t=130498&page=9


Edit:
Just tested to playback 2 original m2ts files ripped from War Blu-ray [8 channel LPCM] disc and The Rock Blu-ray [6 channel LPCM]. dwChannelMask: 0x00000003 for both of them. ffdshow just gives me 2 (out of 8) channels for War but all 6 channels for The Rock. So this seems to just affect 8 channel inputs.


May this be a Haali Media Splitter thing?
Is a workaround possible?

[haruhiko_yamagata]

All channels are written (and in correct blu-ray lpcm order) when I connect the dump filter to haali media splitter in graph edit. Same thing with another sample from another 8channel lpcm track.

I also have samples with 6 channel LPCM. Same dwChannelMask: 0x00000003 but all 6 channels is working with ffdshow!

Any ideas why?

Maybe earlier ffdshow had a workaround or did not depend on having a valid dwChannelMask?
Im 100% sure I played 8 channels from Crank Blu-ray with one of the first ffdshow revs with 8 channel support.
It doesn't matter. dwChannelMask MUST be valid (or 0).
May this be a Haali Media Splitter thing?
I think Haali can tell you if the file is broken or he can fix it.

Is a workaround possible?It is possible. If Haali said he can't support you, please let me know.

[rickardk]

It doesn't matter. dwChannelMask MUST be valid (or 0).I think Haali can tell you if the file is broken or he can fix it.
It is possible. If Haali said he can't support you, please let me know.

I also tested 12 titles in original m2ts (2 titles with 8 ch LPCM and 10 titles with 6 ch LPCM). Using Haali Media Splitter they all use dwChannelMask: 0x00000003.

All 10 titles with 6 ch LPCM does work with ffdshow even though they use wrong channel mask. But the 2 titles with 8 ch will not work.

So I guess the files are not broken.

I have PM:ed Haali and waiting for response.


EDITED:

Haali will fix this in next release!

[ImAhNoBoDy]

The mixer works perfectly fine for me in the latest build. And afaik nothing has changed recently to that part of ffdshow.

Use this to reset your ffdshow settings to defaults:
Windows Registry Editor Version 5.00

[-HKEY_CURRENT_USER\Software\GNU\ffdshow]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_audio]

[-HKEY_CURRENT_USER\Software\GNU\ffdshow_vfw]

So basically just delete ffdshow folders with regedit? o.O

[rickardk]

I have a suggestion:

In ffdshow audio it would be great to have access to ffdshow video stuff in the preset autoload page. I will explain why...

Been working with my audio/video setup for a couple of days and found out that I have to use diffrent delays on audio to get perfect lip sync when using diffrent refresh rates on my graphic card.


I have a Pioneer lx608 Kuro display that can use all kinds of refresh rates. Diffrent refresh rates may introduce diffrent video delays. And someone may have additional video post processing taking place in the display for just a certain refresh rate. As an example there is 150ms diffrence in audio between setting the graphic card refresh rate at 23.976Hz and 59.94Hz. Not sure if the graphic card also introduces diffrent video delays at diffrent refresh rates.


So there is really a need to have diffrent audio presets loaded at diffrent frame rates (movie fps).
Also diffrent video codecs may give additional video delay I think. Because I have to switch audio delay when using CoreAVC and libav.

So if we could have video codec and fps available in ffdshow audio preset auto loading conditions it would help alot!

-----------------------------------------------------------
The author of Media Control working on having Vista Media Center switching the refresh rate to match the frame rate on video. Media Control gets the fps from ffdshow video.
So I guess ffdshow audio could also read the frame rate from ffdshow video?

Is this feature possible?

[haruhiko_yamagata]

I have a suggestion:

In ffdshow audio it would be great to have access to ffdshow video stuff in the preset autoload page. I will explain why...

Been working with my audio/video setup for a couple of days and found out that I have to use diffrent delays on audio to get perfect lip sync when using diffrent refresh rates on my graphic card.


I have a Pioneer lx608 Kuro display that can use all kinds of refresh rates. Diffrent refresh rates may introduce diffrent video delays. And someone may have additional video post processing taking place in the display for just a certain refresh rate. As an example there is 150ms diffrence in audio between setting the graphic card refresh rate at 23.976Hz and 59.94Hz. Not sure if the graphic card also introduces diffrent video delays at diffrent refresh rates.


So there is really a need to have diffrent audio presets loaded at diffrent frame rates (movie fps).
Also diffrent video codecs may give additional video delay I think. Because I have to switch audio delay when using CoreAVC and libav.

So if we could have video codec and fps available in ffdshow audio preset auto loading conditions it would help alot!

-----------------------------------------------------------
The author of Media Control working on having Vista Media Center switching the refresh rate to match the frame rate on video. Media Control gets the fps from ffdshow video.
So I guess ffdshow audio could also read the frame rate from ffdshow video?

Is this feature possible?
I think it is not possible to get fps from ffdshow audio at the time of loading preset.
It is done at the early stage of DirectShow filter connection.
It is possible to get fps during playback though.

You can use video delay (in "Queue & misc" page) instead.
Decrease video delay to increase audio delay.
In ffdshow audio configuration, set constant audio delay.

[rickardk]

You just made my day!!
I thought video delay was global.
One step closer to the perfect setup...Thanks!!!

[jeffy]

I don't know how to solve the following problem, therefore I am asking someone wise for help.

When I encode a standard definition PAL video (720 × 576 pixels) into the following format in VirtualDubMod, I keep getting some dropped frames on playback in Media Player Classic. My CPU is E6600 C2D, currently 3.51 GHz, 800 MHz DDR2 RAM (780 effective), GPU GF6200 (128 MB DDR onboard, non-Turbo Cache version), Windows XP SP2:

Codec: ffdshow video codec, MPEG-4, fourcc XVID, one pass quality 90, Quantization type H263, macroblock decision mode: fewest bits, ME compare functions: SAD+chroma, Subpixel ME...: SAD, Macroblock compare: SAD, EPSZ: 1 diamond, all other settings default, 1 thread, no b-frames, output format: bobbed video 720 × 576 × 50 fps. The harddrive should be no issue as well, according to HDTune benchmark approx. 60 MB/sec.

Can you please help me? What should I set in ffdshow to get the rid of the dropped frames? Or what else could I try? CPU usage is only 2-6%! I have even tried queue output samples in the ffdshow settings, yet MPC is still dropping some frames. Under decoder options I have even unchecked Drop frame on delay 1500 ms.

Thank you for your help.

[thuan]

@ffdshow_tryouts devs: I have a problem with ffdshow_tryouts Keys&Remote enabled which is descibed here http://forum.doom9.org/showthread.php?p=1094678#post1094678 and the forth post down. Please take a look.

[haruhiko_yamagata]

@ffdshow_tryouts devs: I have a problem with ffdshow_tryouts Keys&Remote enabled which is descibed here http://forum.doom9.org/showthread.php?p=1094678#post1094678 and the forth post down. Please take a look.Thanks, I can reproduce.

[chros]

@jeffy: what kind of renderer are you using? On secondary monitor?
Try Overlay in MPC, and see what's happening ...

(For me a 1280x720 59.94fps x264 didn't go well with Haali's+ffdshow+sec-monitor on Athlon X2 4600+, I had to switch to Overlay+CoreAVC and this time it wasn't perfect as well, but it's acceptable ...)

[taklubaba]

Hi,
I have built ffdshow source (ffdshow_beta1_20061211) on Microsoft Visual Studio 2005.
The ffdshow filters gets registered and can be seen in the Graph Edit Insert Filters list.
When I try to insert the registered ffdshow filter, ffdshow video encoder filter to be precise, the Graph Edit gives an error message "The filter cannot be added to the graph. No combination of intermediate filters could be found to make the connection. (Return code: 0x80040217) ".

Before building the ffdshow source and registering the filters, I had installed the ffdshow-rev1803_20080120_xxl.exe. Here too the ffdshow related filters were seen in Graph Edit, and in this case I was able to insert and use the ffdshow video encoder filter in Graph Edit.

What then am I missing in the building of the ffdshow source code?
Do I have to build it in some specified way?
Do I have to add some other code/project and build the same?
Do I have to install some other packages?

Please help.
Thanks,
Taklu.

[clsid]

The files libavcodec.dll and libmplayer.dll must be present in the same directory as ffdshow.ax.

[Delerue]

I want to thank you, dev guys, for the last two patches (H.264 deinterlace and DVD menu with resize bug). Very good things, IMO. ;)

[cc979]

i just tried to compile svn 1842 but got this error:

make[1]: Entering directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/subtitles'
gcc -c -mno-cygwin -mdll -fno-rtti -mthreads -pipe -D_WINGDI_ -DUCLIBCPP -D_GLIBCPP_HAVE_MBSTATE_T -D_WIN32_IE=0x0500 -DARCH_IS_IA32 -DARCH_IS_32BIT -DHAVE_MMX -mmmx -msse -mfpmath=sse -w -DNDEBUG -UDEBUG -DFFDEBUG=0 -I. -I.. -Iuclibc++ -Ibaseclasses -I../baseclasses -IimgFilters -I../imgFilters -Implayer -I../mplayer -Isettings -I../settings -Isettings/filters -I../settings/filters -Icodecs -I../codecs -Isubtitles -I../subtitles -Iconvert -I../convert -Idialog -I../dialog -IaudioFilters -I../audioFilters -Icygwin -I../cygwin -Iffmpeg -I../ffmpeg -Iacm -I../acm -Ifilters -I../filters -Imuxers -I../muxers -I/dx/Include -L/dx/MingLib -ldx9 -O2 -march=pentium-mmx -mtune=i686 -fomit-frame-pointer -finline-functions -finline -frename-registers -fweb -funit-at-a-time -MMD -o ffdshow_subtitles_all.o ffdshow_subtitles_all.cpp
TsubtitleText.cpp: In member function 'void TsubtitleTextBase<tchar>::prepareKaraoke()':
TsubtitleText.cpp:1242: error: there are no arguments to 'clear' that depend on a template parameter, so a declaration of 'clear' must be available
TsubtitleText.cpp:1242: error: (if you use '-fpermissive', G++ will accept your code, but allowing the use of an undeclared name is deprecated)
make[1]: *** [ffdshow_subtitles_all.o] Error 1
make[1]: Leaving directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/subtitles'
make: *** [lib] Error 2

[Thunderbolt8]

just read this:

Revision 1837 - Directory Listing
Modified Sat Feb 2 12:56:58 2008 UTC (34 hours, 18 minutes ago) by clsid2

Updated libavcodec.
H.264 decoding now supports interlaced + spacial direct mode.

does this mean that all H.264 features are implemented now, its 'complete' ? so there shouldnt be any more problems regarding features ffdshow cant decode properly?

[haruhiko_yamagata]

i just tried to compile svn 1842 but got this error:fixed at 1843. Thanks.

[leeperry]

sorry to be asking a prolly retarded question, but I'd like to have 2 profiles in the audio ffdshow filter.

one for music, with "supereq" enabled, and one for movies where it's disabled..........because its buffer is giving dropouts in MPC HC.

so I wanna setup the autoload option with "audio sampling frequency match", the first profile being for <47999 and the second for >47999.

I've tried many combinations, to no avaible :(

TIA,

[haruhiko_yamagata]

sorry to be asking a prolly retarded question, but I'd like to have 2 profiles in the audio ffdshow filter.

one for music, with "supereq" enabled, and one for movies where it's disabled..........because its buffer is giving dropouts in MPC HC.

so I wanna setup the autoload option with "audio sampling frequency match", the first profile being for <47999 and the second for >47999.

I've tried many combinations, to no avaible :(

TIA,Did you enter freq?
freq<47999

[Thunderbolt8]

just read this:

Revision 1837 - Directory Listing
Modified Sat Feb 2 12:56:58 2008 UTC (34 hours, 18 minutes ago) by clsid2

Updated libavcodec.
H.264 decoding now supports interlaced + spacial direct mode.

does this mean that all H.264 features are implemented now, its 'complete' ? so there shouldnt be any more problems regarding features ffdshow cant decode properly?
someone already mentioned there are still rainbow frames in some movies (which arent there when using coreavc). when I have time during the week ill send you a sample.

[leeperry]

Did you enter freq?

works perfect, thanks!

[taklubaba]

"The files libavcodec.dll and libmplayer.dll must be present in the same directory as ffdshow.ax."
Thanks CLSID.
Where can I get the dlls / source for building the dlls on VC++?

Will adding the dlls in the same directory as ffdshow.ax solve the issue?

[Leak]

Where can I get the dlls / source for building the dlls on VC++?
If you've checked out the whole ffdshow-tryout trunk you already have the source; check your src/ffmpeg and src/mplayer directories.

Dunno if they're built with the MSVC project files as IIRC (someone please correct me if I'm wrong) the inline assembly is GCC only so you'll get a build lacking all MMX/SSE code when you compile it with MSVC.

The good news is that GCC is free... :) You can either install cygwin and use it's gcc 3.3 that's a tad rusty, or you can get a newer mingw build here (http://forum.doom9.org/showthread.php?t=108215).

Will adding the dlls in the same directory as ffdshow.ax solve the issue?
Those two DLLs contain most codecs and color conversion code, which means without them ffdshow can hardly connect to anything - so I'd say that's a yes.

np: Pole - Hafen (2)

[leeperry]

ok forget about supereq, I've found THE ultimage winamp plugin :D

http://www.winamp.com/plugins/details/79374

it works perfectly in ffdshow audio, but each time I open a new file in MPC HC or mplayer2.exe, it crashes the player :(

is it possible to make the "ffdshow audio filter" not crash when you open a new file and have a winamp2 plugin enabled ? :(

and the ability to keep the plugin GUI hidden would also be AWESOME, like in winamp.

Thanks,

[Kado]

@leeperry

I can reproduce that, when track changes it triggers the cpu "Data Execution Prevention" and crashes mpc. That happens only for iZotope Ozone (I have the full version).
I use enhancer anyway!:D

The ability to hide the plugin gui depends on the plug in, because ffdshow invokes the plug in every time a track changes iZotope will appear, that also happens if you close and open Winamp using iZotope.
For enhancer the window will remain hidden but mpc will lose focus to that hidden window in track changes.

[swalker]

Should ffdshow default to connecting to WaveOut for only S/PDIF output because of the known problems when connecting to DirectSound?

http://forum.videolan.org/viewtopic.php?t=31776 - S/PDIF finally working like a charm!
http://forums.snapstream.com/vb/showthread.php?t=33014 - SPDIF cutting out
http://forum.inmatrix.com/index.php?showtopic=323 - DirectSound vs WaveOut

[leeperry]

@leeperry

I can reproduce that, when track changes it triggers the cpu "Data Execution Prevention" and crashes mpc. That happens only for iZotope Ozone (I have the full version).
I use enhancer anyway!:D

DFX you mean ? it's quite OK, but I prefer the "tube" sound of Ozone :D

and I also love the sound of the MPC mp3 decoder, so do I have any other alternative than MPC HC ? :(

I've written about this issue to the izotope tech support, but Ozone was coded in 2001.....I don't expect miracles...

haruhiko_yamagata : is there a way you could make ffdshow audio shutdown Ozone before it triggers the cpu "Data Execution Prevention" when a track change occur ?
..so it doesn't crash the player ?

also, I would guess that some argument to the winamp plugin would let it hide the GUI....when switching to new files at least ?!

..but if it didn't crash the player that'd already be too awesome!!!! :D

[taklubaba]

"If you've checked out the whole ffdshow-tryout trunk you already have the source; check your src/ffmpeg and src/mplayer directories."

Thank You Leak.
I have found these directories and have managed to build the mplayer in Visual Studio and get the libmplayer.dll.
I am trying to build the libavcodec project on Visual Studio but get errors like,
syntax error : identifier 'uint_fast8_t' in vorbis.h
error C2065: 'M_E' : undeclared identifier in opt.c
'int_fast32_t' : undeclared identifier in jfdctint.c
DCTELEM' : illegal use of this type as an expression in jfdctint.c

Regarding uint_fast8_t, on searching for it, I came to know that it is declared in inttypes.h. But I cannot find this header in the ffdshow code.
Are they declared in any other file?

Once I am able to get the libavcodec.dll, I will try to insert in GraphEdit. Let see how it goes.
Thanks,
Taklu.

[clsid]

Just download the ffdshow installer from the sourceforge website. That includes all the files you need.

[clsid]

A few assembly code additions/changes have been made in libavcodec in the past few days.

These changes should give about 5% performance improvement on an Intel Core2 when decoding H.264 video.

Revision 1846 includes all the changes. Everyone please test.

[fastplayer]

These changes should give about 5% performance improvement on an Intel Core2 when decoding H.264 video.
Just to avoid confusion: the performance improvements also affect other MMX/SSE2-capable CPUs.
Core2 has gotten some special treatment though by using SSSE3 instructions to improve H.264's motion compensation.

[leeperry]

guys, I'm using this DirectX plugin wrapper to winamp2 :

http://pagesperso-orange.fr/vb-audio/us/products/dxrack/dxrack.htm

the winamp version is freeware :
http://vincent.burel.free.fr/download/ffx4WinAmp_FullDemo.zip

I use it with the Ozone3 DX plugin.....but when I play new tracks, MPC HC tells me "Failed to query the needed interfaces for playback"

and mplayer2.exe crashes.

what gives ?

it seems that the ffdshow audio filter doesn't succeed to close the winamp plugins in a clean and safe way :(

edit : what is really *WEIRD* is that if I click on "configure" on the winamp DSP section of the ffdshow audio settings, then I can open new files as much as I want ?!! it doesn't crash anymore but the CPU utilization goes from 10% to 60/70% after I open 2/3 files in a row :eek:

problem is easy to reproduce with OzoneMP :
http://www.winamp.com/plugins/details/79374

or the DX wrapper above.
when I open a new track, it will open a new winamp plugin window, and everything works fine :)

so maybe a bug ?!

[wyrd]

A few assembly code additions/changes have been made in libavcodec in the past few days.

These changes should give about 5% performance improvement on an Intel Core2 when decoding H.264 video.

Revision 1846 includes all the changes. Everyone please test.
Hi, clsid
Thank you for your efforts.
I've tried beta4a,rev1827,rev1839 and rev1846 in timecodec.
Results (http://tirnanog.fate.jp/tmp/comparison/ffdshow_comparison_clsid_rev1846.txt)
test files (http://tirnanog.fate.jp/tmp/mp4_test/)

in C2D(E6600@2.4),XPsp2(32bit)
Regards

[cc979]

"I have found these directories and have managed to build the mplayer in Visual Studio and get the libmplayer.dll.
I am trying to build the libavcodec project on Visual Studio but get errors like

i think libavcodec compiles only with gcc 4.0.4 or the patched gcc 4.2.1

[fastplayer]

Thanks for the tests, wyrd! :)
I put the results of 2 tested trailers into a chart:

http://img2.freeimagehosting.net/uploads/939a41d37c.png

[clsid]

i think libavcodec compiles only with gcc 4.0.4 or the patched gcc 4.2.1
It can be compiled with Visual Studio. (assembly stuff won't be included in that case)

[Inventive Software]

Thanks for the tests, wyrd! :)
I put the results of 2 tested trailers into a chart:

http://img2.freeimagehosting.net/uploads/939a41d37c.png

That's some nice increasing over beta4a, and a small (1% and 2.2% respectively) but significant increase over rev1826. :)

[haruhiko_yamagata]

Hi,
I have built ffdshow source (ffdshow_beta1_20061211) on Microsoft Visual Studio 2005.You may have some compilation errors if you try to compile beta1-libavcodec by MSVC.
Please install tortoisesvn or your favorite svn client and check out the latest svn.

[haruhiko_yamagata]

@leeperry,swalker
I don't have time now. Please wait.

[chros]

That's some nice increasing over beta4a, and a small (1% and 2.2% respectively) but significant increase over rev1826. :)
Woow, thank you guys !!!
I hope I can get some improvement on my AthlonX2 too (I was using xll's 1738 build, which already had the multicore support) ... :)

[taklubaba]

Thanks to Leak, clsid, cc979, haruhiko_yamagata.

I have downloaded the ffdshow_beta3_rev1324_20071127 source and managed to build ffdshow, libavcodec and mplayer on Visual Studio 2005.
The ffdshow video encoder and decoder filter can now be inserted into the Graph Edit.

BUT :), there is are some problems.

1. When i check the property pages of ffdshow video decoder it shows that MJPEG, other MJPEG, H246/AVC, Raw Video formats are disabled. How do I enable them in the code?

2. Every time I run the graph edit from ffdshow project in Visual Studio and try to connect a yuv input file or some filter to the ffdshow encoder filter, the graph edit hangs for a long time and
then I get a message in Visual Studio saying

Unhandled exception at 0x01c41b02 in graphedt.exe:0xC0000005:Access violation reading location
0x00000040

The statement causing the problem is
mov eax,dword ptr [edi+40h]

Value of eax and edi is 0. The call stack shows this statement to be in the vp6dec.ax.
What could the problem be? Is it stack overflow?

3. Now if I directly open the graph edit (not through the ffdshow project in Visual Studio) and try to connect a yuv input file to the ffdshow encoder the graph edit again takes a long time before throwing out the following error messages

ASSERT Failed: graphedt.exe - Application Error
The instruction at "0x7482a522" referenced memory at "0x0c8e4cac". This memory could not be "read"

The instruction at "0x74829ea1" referenced memory at "0x0c8e4e94". This memory could not be "read"

Has anyone faced the above problems? Please suggest how to go about solving this in Visual Studio.

I am using the source code because I want to use the MJPEG part of the ffdshow code for an application.

Thanks,
Taklu.

[clsid]

The settings for which codecs are enabled is stored in the registry (HKCU\Software\GNU\ffdshow).

Why are you using an old version of the source code? If you install TortoiseSVN, then you can check out the lastest code from the SVN repository.

[_xxl]

http://img147.imagevenue.com/loc1070/th_16102_Capture_122_1070lo.jpg (http://img147.imagevenue.com/img.php?image=16102_Capture_122_1070lo.jpg)
Why is 250 mb ram used with x264 720p mkv samples?
CPU AMD XP 2000+, MPC rev 36 (internal mkv splitter) and ffdshow 1840.Only subtitles filter is active.Frames are dropped, because this cpu can't handle 720p.

[clsid]

It uses only 40MB RAM here.

[chros]

Here, it's 65 MB with a 1 hour 2GB 1280x720 x264 encode !
What is your renderer (mine is Haali's)?
Perhaps, the queueing feature of ffdshow is the cause ?

[Delerue]

It uses only 40MB RAM here.

Confirmed. Build 1846 + MPC (your lastest build) + VMR7 Windowed.

[_xxl]

For an unknown reason mpc rev 36 compiled by me is using 250mb.
Clsid version doesn't have this issue.Dir mpc/res/web is missing from svn?Can somebody confirm?I used web dir from original project.

[clsid]

Web dir should be fixed now. I had to delete part of my local repository, because that somehow erroneously marked it as committed files (even allowing me to delete/restore them).

[taklubaba]

Thanks clsid.

I have tried using the TortoiseSVN to check out the latest code from the SVN repo.
But TortoiseSVN gives the following error message

Error: PROPFIND request failed on '/viewvc/ffdshow-tryout/trunk'
Error: PROPFIND of '/viewvc/ffdshow-tryout/trunk': could not connect to server (http://ffdshow-tryout.svn.sourceforge.net)

Am I missing something?

[Eragon4ever]

Use this link:
https://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout

[taklubaba]

Eragon4ever,
It doesn't succeed.

I get the message
Error * PROPFIND request failed on '/svnroot/ffdshow-tryout' PROPFIND of '/svnroot/ffdshow-tryout': could not connect to server (https://ffdshow-tryout.svn.sourceforge.net)


Does it matter, if I am behind a proxy or firewall?
Is there any file where in I should mention the proxy address?

[Eragon4ever]

It seems to matter... take a look at Settings -> Network. That should help.

[Joniii]

What is the difference with libavcodec and wmv9 in VC1 and libavcodec and libmpeg2 in Mpeg-2 settings?

[clsid]

They are all different decoders.

[Leak]

What is the difference with libavcodec and wmv9 in VC1 and libavcodec and libmpeg2 in Mpeg-2 settings?
"libavcodec" is a decoding library for almost all file formats, as you can see from the fact that it's an option for almost every format supported by ffdshow. It's what mplayer, xine and several other players use, and ffdshow started out as "just" a DirectShow wrapper around it.

"wmv9" is using Microsoft's WMV decoding libraries that are already installed on your system.

"libmpeg2" is an MPEG2-decoding library that does only MPEG1/2, i.e. it's an alternative to libavcodec that in my experience works better with decoding DVDs.

np: Proem - Live @ Amoda Austin 03-12-2002 (Merck Fragments)

[oddball]

Not sure if already mentioned. VP6 decoding has a major problem. See screenshot. It plays like that for a few seconds then crashes. Does not matter which player I use (I tried about 5 different players).

Disabling it in ffdshow and using the default codec enabled proper playback.

Using ffdshow_rev1850_20080206_clsid_sse_icl10.exe but have not tried other builds.

[fastplayer]

Pics are worthless for the devs. Sample, please! :)

[georgevalkov]

I tried to encode a TV show to H264 using ffdshow,
but the output does not look properly. No errors are
reported in the player, but the image looks strange
and messed up. I tried mplayer.exe and directshow,
both same results.
Input filtering may be disabled or enabled, the same
problem exist either way, so it is not related to it.
Also when I try to stop or abort the encoding process,
the application will crash in the debugger
(Visual Studio 2008 beta 2). I tested VirtualDubMod
and graphedit, both produce same results and both
crash on aboort.

Versions tested to have the H264 encoding bug:
ffdshow-rev1846_20080205_clsid.exe
ffdshow-rev1850_20080206_clsid_sse_icl10.exe

Versions tested to be free of the bug and work
properly with the same configuration:
ffdshow-rev1798_20080119_clsid_sse_icl10.exe
and older.

This is the registry configuration of my computer:
http://gfc.my.contact.bg/tests/2008-02-08_HKCU_ffdshow.reg.7z

Please fix the problem! :thanks:


George Valkov

Please have a look at the screen-shot:
http://gfc.my.contact.bg/tests/H264-bug-FFDShow.png

[clsid]

My latest builds do not contain the x264 encoder (because x264VFW is preferred over it). So you possibly have an old version of the lib installed. Uninstall ffdshow and then re-install it.

[georgevalkov]

My latest builds do not contain the x264 encoder (because x264VFW is preferred over it). So you possibly have an old version of the lib installed. Uninstall ffdshow and then re-install it.

Okay, I can understand that VFW support for H264 is dropped, but FFDShow is a DirectShow filter. I don't see why would you remove the H264 encoder?
You can alert the user that H264 may not work good inside AVI, but VFW and AVI are not the only option. For example:
One can feed the encoder input directly from some DirectShow video decoder; there are still two reasonable ways for using the encoded H264 output, that does not relay on VFW.
~ 1st one is to connect to a MKV or a MP4 multiplexer and file writer. That way, the multiplexer filter will take care to generate a valid container for the H264 stream.
~ 2nd option is to set the FFDShow encoder to write a raw.264 file, that can be remuxed later to MP4 or MKV. The encoder output pin can be connected to Null renderer.

Both cases have the advantage of allowing the user to activate the built-in input filtering that FFDShow offers. So the user can cut|arrange the scenes, apply filtering and encode directly, all in 1 pass. Another advantage is that the RAW H264 output file stream produced by the FFDShow encoder can be remuxed with MP4box.exe on the fly during encoding, and the results can be previewed, without interrupting the process.

In my personal experience, I once took an MPEG file with variable frame rate (starts with 29.97 and changes to 23.97 after). This would usually break the A-V sync if VFW is used, but on the other hand, using graphedit and FFDShow to encode a H264 stream and feed the Haali Matroska Muxer, produced a valid MKV stream, with audio and video in sync. It plays just like the original MPEG file: (starts with 29.97 and changes to 23.97 after).

CLSID, please do not remove this great functionality. If anyone does not like it, that's fine, but still there are many people like me, that love it and use it every day!

:thanks:

George Valkov

[oddball]

Pics are worthless for the devs. Sample, please! :)

Sorry. No can do. It's a viral video preview I am not allowed to release to public yet.

[clsid]

And it is probably a bug in libavcodec. So there is not much we can do about it.

[Delerue]

Today I tried to encode a xVid file with x264 using VirtualDub 1.6.19 + FFDShow 1850, but the results were not good. As you can see, the original xVid has more red, and less green than the encoded x264. Using AutoMKV the x264 encode is fine. The samples:

http://www.zshare.net/download/73239907607157/ (8.5 MB)

I used 'one pass - quality' with quality set to 65.

[leeperry]

@leeperry,swalker
I don't have time now. Please wait.

ok cool, no problem :)

actually, I've updated my report about the winamp2 plugin support here :
http://forum.doom9.org/showpost.php?p=1096605&postcount=3111

basically it's giving audio jitter after 15 mins, and there's a bug that crashes the player when you open new files.

:thanks: for your help!

[Jeremy Duncan]

To the developers, Clsid, XXL, Harahiko, and Leak.
I ask, is all the major development done and now your just working on bugs and stability?

And I wanted to tell you about a bug for about a year now, but never bothered.

In the install, when it asks you what programs for audio and video.
I select mpc and a few other programs, about 4. And I get these popups asking me if I should use ffdshow on it or not. Things like my widgets and the ccc lcd overdrive.

But, if I don't pick specific programs during install, I never get these popups asking for permission.

heh heh. That bug has bothered my for maybe 2 years, I dunno. I forget now. :helpful:

[clsid]

That is not a bug. The installer lets you choose to enable the "whitelist" or not. If enabled it will show a prompt whenever an unknown application tries to use ffdshow.

You can't select specific applications from the list. The list that is shown is the entire whitelist.

[fastplayer]

You can't select specific applications from the list. The list that is shown is the entire whitelist.
Maybe someone could make the list "un-editable".
Or: Is that list even necessary? Does actually someone go through that list during setup? I think Joe Sixpack doesn't bother and just clicks "Next, Next, Next,...".

[clsid]

I agree, just a few checkboxes should be enough imo. However, Haruhiko preferred if the list itself is also displayed. I don't know if he still thinks that. He has to comment about that himself.

I would suggest two checkboxes (on the tasks page):
[ ] Restrict usage of the ffdshow filters to whitelisted applications
|--[ ]Show prompt for applications that are not on the whitelist/blacklist

[fastplayer]

I agree, just a few checkboxes should be enough imo. However, Haruhiko preferred if the list itself is also displayed. I don't know if he still thinks that. He has to comment about that himself.
Yes, he has put a lot of effort into that feature.

I would suggest two checkboxes (on the tasks page):
That's even better than what I had in mind: Putting both "Management of Comp..." into one page for both audio and video (separately configurable). That would reduce the number of steps to go through by 1. Your approach is better but it'll get crowded in the task page. Maybe it's time for a "recommended" and "custom" installation routine...

[compunett]

It might have been already answered, but i didnt found the post or the thread talking about it.

So the development of the EAC3 codec seems working and jruggle commited it in the mplayer repository. It seems the decode is called ffac3 (http://forum.doom9.org/showthread.php?p=1088908#post1088908) . I also saw that the open MLP / DTS TrueHD was commited in the repository.

However the latest installer of ffdshow don't show any information on them, and of course ffdshow doesn't decode them.

Is there a still some work to do on this matter, either on ffdshow, on mpc, or matroska splitter. (As they are part of my current configuration).

tia for your information. :)

Btw I express my utmost consideration to your work that has already been done, it's a fantastic work, credits goes as well to the mplayer team ;)

[clsid]

No work is being done on it. We use the decoding code from FFmpeg, and no EAC3 decoder has been committed to that repository yet.

[compunett]

Ok, thx for the information.

Just for my understanding, isn't ffmpeg and mplayer very related, I mean libavformat and libavcodec are shared or the same on these two projects ?

[Inventive Software]

Only in that sense. Otherwise, we use what's in ffmpeg's repository. :)

[compunett]

Ok, so probably there is some syncing to be done on both repositories, between mpleyer and ffmpeg.
I saw on mplayer svn a few commits from jruggle.

Anyway thx for the information

[leeperry]

is it possible to decode ac3/dts with the ffdshow audio decoder, and also get it through the winamp2 plugin ?
it seems to only work with external filters.

TIA,

[taklubaba]

"Error: PROPFIND request failed on '/svnroot/ffdshow-tryout/trunk'
Error: PROPFIND of '/svnroot/ffdshow-tryout/trunk': could not connect to server (https://ffdshow-tryout.svn.sourceforge.net)
"

I have tried mentioning the proxy address and port.
But TortoiseSVN is not able to connect to the server.
I have tried to checkout the code from SVN server using tortoiseSVN, from a PC with no firewall and no proxy. Yet I get the same error message.

Has anyone faced this problem?
Can anyone try checking out https://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout/trunk
using TortoiseSVN on a Windows XP platform?

Thanks,
Taklu.

[clsid]

Works fine for me. Are you able to visit the url with your browser?

[Inventive Software]

Works absolutely fine for me, running WinXP MCE2K5, TortoiseSVN, and successfully checked out ffdshow-tryouts. So it must be something your end.

[haruhiko_yamagata]

You can't select specific applications from the list. The list that is shown is the entire whitelist.You can edit the whitelist there, can't you?
I agree, just a few checkboxes should be enough imo. However, Haruhiko preferred if the list itself is also displayed. I don't know if he still thinks that. He has to comment about that himself.

I would suggest two checkboxes (on the tasks page):
[ ] Restrict usage of the ffdshow filters to whitelisted applications
|--[ ]Show prompt for applications that are not on the whitelist/blacklist

As I explained to clsid before, the installer show those pages to let users know about the whitelist feature.
Users have to know about it IMO.
I don't expect users to check/uncheck or edit anything in the pages (although they can do it).
People wouldn't be too happy if the dialog box suddenly pops up to prompt them to allow/reject the use of ffdshow.

The installer show two pages, for video and audio, which is too much. But the settings are separate, and it won't look very nice if I try to squash both in one page.

[leeperry]

oh btw haruhiko_yamagata, you can forget about my request for better winamp2 plugin support in ffdshow audio.

I've run many tests these past days and the only way to get a *rock stable* 24.000 FPS frame rate in 48 Hz in MPC HC is to use the following filters :

-ffdshow video decoder(for h264)
-haali's media splitter(gabest's splitter is not as accurate, and neither is KMP's splitter)
-intervideo audio decoder
-EVR

if I add the "ffdshow audio processor", after 30 to 50 minutes I get a lot of jitter in MPC HC :(

I guess the timing of the audio part of ffdshow is not really tight, and it shows :(

[fastplayer]

You can edit the whitelist there, can't you?
Nope, just an independent listing of the whitelist regkey.

As I explained to clsid before, the installer show those pages to let users know about the whitelist feature.

I think we need to look at this more from a "normal user"-perspective. "Normal users" just want things to work. They don't read and don't want to read anything during setup because they assume that the application is already using optimal settings.
But what may be optimal for a "normal user", may not be for us "advanced users".
So the question is: Who do we aim at with the current installer? The regular Joe Sixpack or the A/V enthusiast?
If it's the former then the current setup routine is too "much", if it's the latter then don't change a thing.

[haruhiko_yamagata]

Nope, just an independent listing of the whitelist regkey.It works for me :confused:.
So the question is: Who do we aim at with the current installer? The regular Joe Sixpack or the A/V enthusiast?
I think ffdshow is mainly for A/V enthusiast.
Why do the "normal users" use ffdshow rather than VLC?
Just because they find ffdshow in first or do they have some reason?
Of course both users are welcome, but the latter is more targeted.

Another choice is to disable the whitelist by default. Then we don't have to show the pages in the installer.

[SeeMoreDigital]

Why do the "normal users" use ffdshow rather than VLC?I must admit I'd use VLC player more if it had a more responsive "pause" button... and a "step" button?

[fastplayer]

It works for me :confused:.
If I make changes in the audio compat. page, then they get ignored.
If I change EXEs in the video compat. page, then they end up correctly in the whitelist regkey.


I think ffdshow is mainly for A/V enthusiast.
Hmm, the applications that get reported by the minute tell me a different picture. I see more entertainment-related applications reported than professional A/V apps.

Why do the "normal users" use ffdshow rather than VLC?
- People know ffdshow as a "codec pack",
- and they know it as one of the good-behaving ones.
- They get it recommend from A/V enthusiasts :D
- People like WMP but they know it lacks some codecs...

Of course both users are welcome, but the latter is more targeted.
Personally, I like ffdshow setup how it is now. It gives me the option to configure the most important stuff right during setup.
But I still think that we should target a broader base than just a niche group - meaning us :)

Another choice is to disable the whitelist by default. Then we don't have to show the pages in the installer.
I like clsid's idea mentioned above. I'd just change the wording:

[ ] Restrict usage of ffdshow to known, compatible applications
|--[ ]Show prompt for applications that are unknown to ffdshow

[Keepitsimple]

I think the reason many prefer ffdshow over vlc is better spdif support. The shaders available are also better.

[umaximus]

Im having problem with ffdshow clsid ver.1846 with mpeg2 of Lost cap. ffdshow is dropping a huge amount of frames in MPC, while Cyberlink Video/SP Decoder (PDVD7) is playing it without a single droped frame. Video is MPEG2.ts (~40mbit bitrate), and i did try with libavcodec & libmpeg2 under codecs in ffdshow, no difference.
Im having a lot of CPU power (Q6600@3.6Ghz), and cpu usage is arround 10% while playing this, so this is not a problem.


Here is the sample --> http://rapidshare.com/files/91239085/lost40mbit.ts.html

[fastplayer]

No frame drops with any combination of MPC (6.4.9.1, rev36) and ffdshow (rev1852).

[umaximus]

No frame drops with any combination of MPC (6.4.9.1, rev36) and ffdshow (rev1852).

I did upgrade both, but I still get droped frames on this. I didnt experience any problems with all other movies/shows.

Have you tried the sample that i posted?

Also, Im playing this of raid5 partiotion, with read speeds ~400mbit, so hdd is not the issue here either.

[fastplayer]

Have you tried the sample that i posted?
What would've been the point of my previous reply if I didn't try it? :)

What TS-splitter are you using? I'm using MPC's internal one and - again - no problems with playback. It surprises me a bit that my A64 3500 has no trouble keeping up considering it's 720p @ 60fps...

[umaximus]

Sorry, I understand that you have no problems with playing back your own files.

Im using Haali ver 1.7.401.3. I did try with System Default and there is no droped frames, but I prefer using Haali.
This is the first file that is cousing problems with dropped frames.
I can normaly play back more 'demanding' h264, vc1 files with no hickup. Any ideas why this hapens here?

[fastplayer]

Any ideas why this hapens here?
No idea. Haven't had any trouble so far with the internal splitter, so I never bothered using other ones.
I'd report the issue to Haali in the Media Splitter thread:
http://forum.doom9.org/showthread.php?t=80762

[Thunderbolt8]

didnt try the sample, but is it 720p @ 60fps? if yes, then its a know haali splitter/renderer (splitter? renderer? one of those i guess) problem, afaik I also reported that somewhere in the haali thread, but no solution has come up so far. the only thing you can do atm is choosing another renderer for playback, try vmr9 or overlay.

[umaximus]

Yea, it is 720p/60fps. I did try some other files that i have with 60fps and same story -droped frames with ffdshow/haali splitter.

It must be something in connection between those two, coz if i choose PDVD decoder with haali splitter everything works as it should.

Thanks for clearing this out!

[dimzon]

Why do the "normal users" use ffdshow rather than VLC?

1) Because I can use my favorite player with my favorite UI instead ugly VLC
2) Because I can play proprietary formats (VP7, RealVideo) in same player (using DirectShow filters)
3) Because I can use ffdhow for some post-processing and image tweaks even if I use third-party decoder

[Thunderbolt8]

despite that spatial thing and interlaced support have been added for H264 decoding, I still have trouble with some movies, in which rainbow frames or something like that occur from time to time. these are mainly remuxes from HD DVDs or blu-rays to .mkv and those problems occur (at least in this case here, will have to check up on the others a later time again) not only with ffdshow, but also with coreavc and vlc.

I dont know whether this problem is still caused by something which is unsupported yet or if the remuxing process of mkvmerge causes these problems. I uploaded a sample of this remux, I tried to cut a sample from the original .m2ts file, but apparently I always hit an open GOP or something like that and when I muxed it it looked worse than now, with green stuff in it and such.
would be nice if someone could have a look at that, thanks!

http://www.sendspace.com/file/q3rvfp

[nurbs]

Since my HD-DVD/Blu-Ray drive should be here any day I have one question.

Is the VC1 decoder in ffdshow any good or is it more adviseable to use a decoder from an other source?

[Richard Berg]

I'm having some trouble with the VFW plugin.

1) It won't decode MJPG files created with older versions of ffdshow.

2) MJPG and HFYU are not supported at all in by the encoder. At least, they're not in the dropdown list.

Are these known issues? I can provide a sample clip for the 1st one if necessary.

[taklubaba]

"Works fine for me. Are you able to visit the url with your browser?"

I am using a firefox 2.0.0.4 version browser.
I can visit the svn url with the browser and also see the contents of the files in the folder.

I have installed the TortoiseSVN. I then right click on a folder and select SVN Checkout.
I enter the url of the repository as http://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout/trunk
and the Checkout Directory as the folder on which I had right clicked to select SVN Checkout.

But I get a failed request on /svnroot/ffdshow-tryout/trunk.

Do I need to install any other software to work with TortoiseSVN? Or do I need to change some settings?

The system administrator of my company faced the same problems that I am facing.

Please suggest.
Thanks,
Taklu.

[Thunderbolt8]

despite that spatial thing and interlaced support have been added for H264 decoding, I still have trouble with some movies, in which rainbow frames or something like that occur from time to time. these are mainly remuxes from HD DVDs or blu-rays to .mkv and those problems occur (at least in this case here, will have to check up on the others a later time again) not only with ffdshow, but also with coreavc and vlc.

I dont know whether this problem is still caused by something which is unsupported yet or if the remuxing process of mkvmerge causes these problems. I uploaded a sample of this remux, I tried to cut a sample from the original .m2ts file, but apparently I always hit an open GOP or something like that and when I muxed it it looked worse than now, with green stuff in it and such.
would be nice if someone could have a look at that, thanks!

http://www.sendspace.com/file/q3rvfp
another sample from another AVC remux, same problem:

http://www.sendspace.com/file/xyyquw

[clsid]

@taklu

TortoiseSVN is all you need. It should just be working. You could try a different SVN program:
http://en.wikipedia.org/wiki/Comparison_of_Subversion_clients

[Leak]

I am using a firefox 2.0.0.4 version browser.
I can visit the svn url with the browser and also see the contents of the files in the folder.
Are you using a proxy in Firefox? If so, you'll need to set it up in TortoiseSVN as well. And if your proxy happens to be something Microsoft-y that wants NTLM authentication you'll probably need to use an NTLM auth proxy (http://ntlmaps.sourceforge.net/) on your machine...

I have installed the TortoiseSVN. I then right click on a folder and select SVN Checkout.
I enter the url of the repository as http://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout/trunk
and the Checkout Directory as the folder on which I had right clicked to select SVN Checkout.
The URL given at SourceForge for SVN access is actually https://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout - does that work? (Note the "https" in the URL).

The system administrator of my company faced the same problems that I am facing.
Why do I have the feeling that it's your firewall/proxy/whatever that keeps the nasty outside world away from your network is also keeping TortoiseSVN from connecting to the outside?

In that case, either try HTTPS (because that's usually passed through firewalls so banking sites et al work), or break out Wireshark (http://www.wireshark.org/) and look at TortoiseSVN's communication - I bet the error isn't on SourceForge's side (I'd expect a 404 Not found if the URL is wrong, but not "Could not connect to server"...) , but your own network edge...

np: Grad_U - Knob III (CMYK)

[Inventive Software]

I think ffdshow's decoding of elementary MPEG-2 files is broken somehow. It seems to output a SAR of 64:27 for a 16:9 720x576 video, when it's reporting internally that the SAR is 64:45. MPC's internal decoder has no such problems. Resize and aspect is disabled. Do you need a sample stream to confirm this?

[haruhiko_yamagata]

I like clsid's idea mentioned above. I'd just change the wording:

[ ] Restrict usage of ffdshow to known, compatible applications
|--[ ]Show prompt for applications that are unknown to ffdshowOK, I'll use this. Please wait a few more days until I can get back to development.

[haruhiko_yamagata]

I'm having some trouble with the VFW plugin.

1) It won't decode MJPG files created with older versions of ffdshow.Please send us a sample.

2) MJPG and HFYU are not supported at all in by the encoder. At least, they're not in the dropdown list.I can find them in the dropdown list.

[fastplayer]

OK, I'll use this. Please wait a few more days until I can get back to development.
No hurry!
I suggest that this option covers both audio and video. IMO there's no need to separate them - at least during the installation process. This'll make it less confusing for beginners.
Us advanced users can easily make further changes in the DirectShow control page which should remain unchanged.
:thanks:

[Leak]

I think ffdshow's decoding of elementary MPEG-2 files is broken somehow. It seems to output a SAR of 64:27 for a 16:9 720x576 video, when it's reporting internally that the SAR is 64:45. MPC's internal decoder has no such problems. Resize and aspect is disabled. Do you need a sample stream to confirm this?
Which decoder (libmpeg2 or libavcodec) are you using, and have you tried the other one just in case?

np: Grad_U - Knob III (CMYK)

[fastplayer]

@Haruhiko:

With rev1626 (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout?view=rev&revision=1626) you fixed a bug where subtitles with an "&" in the filename were incorrectly shown in ffdshow's context menu.
Unfortunately, your bugfix introduced another bug:
Scenario 1


Directory A & B
|--> A & B.mkv
|--> A & B.srt
Works. No problems at all.

Scenario 2

Directory A & B
|--> A & B.mkv
|--> A & B.srt
|--> AB.srt

Works until you change to the AB.srt subtitle (via right-click) because then ffdshow doesn't show anything. Even selecting back the "default" A & B.srt subtitle doesn't work anymore.

Looking at the code (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/TtrayIcon.cpp?annotate=1626&pathrev=1626#l437), I see that you simply escaped a single "&" with "&&" which is the correct way but you're escaping the path to the subtitle file too (variable files contains file names and the corresponding paths).
The reason why Scenario 1 and 2 work in the beginning is that the escaping doesn't occur until you open the context menu (makeSubtitlesMenu()). If you then change the subtitle (or even select the same one), processCmd() will operate with the incorrect, escaped path (which was passed via GetMenuString()).

My analysis may be wrong, so take it with a grain of salt.
Java FTW! :p

[EpheMeroN]

Are MKV files not supported in FFDShow? Or is there an option I need to tick somewhere in config to allow decoding support?

I use WMP w/ FFDShow Tryouts and it plays everything I throw at it lately but nothing in an Matroska container.

[Leak]

I use WMP w/ FFDShow Tryouts and it plays everything I throw at it lately but nothing in an Matroska container.
ffdshow is no splitter. Never was, never will be.

If you want to play files with a container format that isn't supported by DirectShow out of the box install a splitter.

np: Rumpistol - Dodvaegt (Intelligent Toys 4)

[Inventive Software]

Which decoder (libmpeg2 or libavcodec) are you using, and have you tried the other one just in case?

np: Grad_U - Knob III (CMYK)

Both, and for good measure, tried changing the "MPEG in AVI" type. Both decoders gimme a wrong SAR. :(

I suspect it's something to do with my MPEG-2 in MKV bug not being nice with different SARs, but if it's doing it with elementary files, that suggests otherwise?

And I can't remember if it happened last time with the elementary files.....

[leeperry]

@leeperry,swalker
I don't have time now. Please wait.

OK I think I have fixed my problems with ffdshow and reclock :)

if I check "drop frame on delay : 1 ms", I get perfect synchronisation with reclock resampling 23.976@24 fps in 48.000 Hz.....it stays right on 24.000 even after 2 hours straight :)

so basically if you could still fix the bug with winamp2 plugins explained here, that'd be really AWESOME :D

http://forum.doom9.org/showpost.php?p=1096605&postcount=3111

the ability to hide the winamp2 plugin GUI would also be fantastic....like in winamp.

TIA,

[Episode]

Are MKV files not supported in FFDShow? Or is there an option I need to tick somewhere in config to allow decoding support?

I use WMP w/ FFDShow Tryouts and it plays everything I throw at it lately but nothing in an Matroska container.

You need to install Haali's media splitter to get it working. You can get it from http://haali.cs.msu.ru/mkv/

[taklubaba]

Leak, clsid, Inventive Software and others.

Thanks a lot for your help. Really appreciate it.
Though the problem may seem small, your suggestions have helped me in checking out ffdshow code from the repository.

The problem was with the url.
As Leak had suggested I tried with
https://ffdshow-tryout.svn.sourceforge.net/svnroot/ffdshow-tryout
and it worked.

Thanks once again.
Taklu.

[leeperry]

I've made 2 profiles in ffdshow audio...one for movies and one for music.
but for WMV files, it never works.
the "movie" profile is supposed to be enabled when "ffdshow video decoder" is active(I've set ffdshow video to "all supported"), but it doesn't work for WMV files......even though "ffdshow video decoder" is loaded ?!

[clsid]

Why don't you use the default profile in this case? Because the condition "ffdshow video decoder is in the filter chain" will (or should at least) always be true. So it makes no sense to use that condition.

[leeperry]

well it works perfectly fine....except for WMV files.

the "movies" profile is used if "ffdshow video decoder" is loaded

otherwise, the "music" profile is used.

but for WMV files, it falls back to the "music" profile.....even though "ffdshow video decoder" is loaded.

[clsid]

Why do you have a "music" profile in the video settings of ffdshow? That makes no sense.

The audio and video settings (including the profiles) are fully separated in ffdshow!

[leeperry]

huh ?

I got these profiles in the ffdshow Audio Decoder Configuration.

[Leak]

Why do you have a "music" profile in the video settings of ffdshow? That makes no sense.

The audio and video settings (including the profiles) are fully separated in ffdshow!
Ummm... as far as I understand it he wants two different audio profiles to load under different circumstances:

* One ("movies") when the ffdshow video decoder is in the graph, since the audio decoder is then probably decoding a movie's soundtrack

* Another one ("music") when there's just the audio decoder in the graph, since that's probably the media player just playing some MP3 file or the like

(Personally, I'd just use a dedicated audio player (i.e. Foobar2000) for playing music, but that might be just me in that case...)

np: Proem - Live @ KVRX 10-09-2001 (Merck Fragments)

[clsid]

Ok, I have read it incorrectly. I thought you defined the profiles in the video decoder.

But there indeed is a bug. It also doesn't work with for example "Video Renderer".

Edit:
A workaround would be to use this condition: "ffdshow Video Decoder;DMO Wrapper Filter"

[leeperry]

w00t thanks, clsid!

I tried with "ffdshow Video Decoder;WMVideo Decoder DMO" as it says in the filters section....but that didn't work :eek:

well I got one "movie" profile that doesn't resample to 96 Hz to avoid problems with Reclock doing 23.976@24 resampling.
in the "music" profile, I've got noise shaping set to "heavy", and resampling in high quality to 96 KHz.
both profiles are processed with the "OzoneMP" winamp2 plugin in 32 bit float, and output in 24 bit integer through WaveOut to my M-Audio Audiophile USB and my Sony MDR-CD3000.

I've tried foobar, but the GUI sucks, and I very much like Gabest's mp3 decoder.....sounds really good to me :)

there's some bugs in the winamp2 plugin support as explained here : http://forum.doom9.org/showthread.php?p=1100930#post1100930
hopefully haruhiko_yamagata would fix them soon enough :D

[Thunderbolt8]

another sample from another AVC remux, same problem:

http://www.sendspace.com/file/xyyquw
http://www.sendspace.com/file/q3rvfp
could anyone please have a look at these? I really would like to know if this is maybe due to mkvmerge then I know Ill have to wait until a better remux solution comes up.

[clsid]

Well, ffdshow should show the actual name of the DMO filter, instead of the wrapper filter. So that is another little bug. Same probably happens with the ACM and VFW wrappers.

[haruhiko_yamagata]

could anyone please have a look at these? I really would like to know if this is maybe due to mkvmerge then I know Ill have to wait until a better remux solution comes up.I have no problem playing back clipboy.m2ts.
As for splighost.mkv, whichever I use CoreAVC or ffdshow, I have the same problem.

[haruhiko_yamagata]

I've made 2 profiles in ffdshow audio...one for movies and one for music.
but for WMV files, it never works.
the "movie" profile is supposed to be enabled when "ffdshow video decoder" is active(I've set ffdshow video to "all supported"), but it doesn't work for WMV files......even though "ffdshow video decoder" is loaded ?!ffdshow video decoder has to be loaded before ffdshow audio initialize its preset settings. Wmv splitter loads audio first.

[leeperry]

Well, ffdshow should show the actual name of the DMO filter, instead of the wrapper filter. So that is another little bug. Same probably happens with the ACM and VFW wrappers.

still not a foul proof solution, because the "movies" profile is loaded for .wma files :scared:

[clsid]

ffdshow video decoder has to be loaded before ffdshow audio initialize its preset settings. Wmv splitter loads audio first.
So that also means that the preset condition won't work for any filter that appears after ffdshow in the graph? Like for example "Video Renderer".

[Thunderbolt8]

I have no problem playing back clipboy.m2ts.
As for splighost.mkv, whichever I use CoreAVC or ffdshow, I have the same problem.
try to remux the clipboy.m2ts also to .mkv please, then you should see the same problems there. I gave you this one as .m2ts so you could have a look at the original structure.
so coming back to my initial question, are all the H264 features of those 2 movies are included in ffdshow meanwhile? if yes, then I can take it for granted its mkvmerge that causes these problems?

[skrowl]

Has a few issues playing (legally obtained, of course) EVO files. The video looks perfect, but the audio comes out garbled.

[Leak]

I've tried foobar, but the GUI sucks
Have you tried it lately? The GUI was completely revamped for it's 0.9.5 release, and even before that there were several other GUI plugins to replace the default one...

np: Pan Sonic - Lähetys / Transmission (Katodivaihe / Cathodephase)

[haruhiko_yamagata]

So that also means that the preset condition won't work for any filter that appears after ffdshow in the graph? Like for example "Video Renderer".It should work for video decoder because ffdshow is connected to downstream when it is initialized. It may or may not work for audio decoder.

[Inventive Software]

Has a few issues playing (legally obtained, of course) EVO files. The video looks perfect, but the audio comes out garbled.

This is most likely because you don't have an E-AC3, TrueHD, DTS-HD MA or DTS-HD HR decoder, depending on what audio type it is. If you want E-AC3 support (or the other HD audio formats), bug the ffmpeg devs, not us! ;)

[breez]

How complete is the ASS subtitle support? It seems to work fine most of the time, but vertical writing seems broken (unimplemented?) at least in a few anime op/ed sequences with karaoke kanji/kana.

e: Forgot to mention rev 1850

[DigitalDeviant]

I'd also like to add it doesn't make use of mkv's attached fonts like vsfilter does (I haven't tried to actually embed the font data in the ass file itself yet).

[breez]

I'd also like to add it doesn't make use of mkv's attached fonts like vsfilter does (I haven't tried to actually embed the font data in the ass file itself yet).

Works ok with Haali splitter. "Load embedded fonts" may need to be set in Haali's options. While testing, I had that setting disabled and I got subs with Arial font. Intended font showed up when enabling the setting. On subsequent tries with setting disabled right font was still used (fonts cached somewhere?).

[JarrettH]

Friend from IRC was asking...

<+Jenova266> Just wondering if it would be possible to add a simple tick box to the deinterlacing filter page which would run the frame doubler filter before the selected filter and if so please do so, Thanks in advance

[haruhiko_yamagata]

How complete is the ASS subtitle support? It seems to work fine most of the time, but vertical writing seems broken (unimplemented?) at least in a few anime op/ed sequences with karaoke kanji/kana.

e: Forgot to mention rev 1850Vertical writing is not implemented.
TODO (ASS related, if my motivation continues)

complete karaoke (fix a bug) - done
implement move
implement layer
implement collisions
implement clip
implement other minor stuffs (scrolling text, rotate/transformation, drawings, vertical writings and others)
check and fix compatibility issues
support style overrides from GUI

[Leak]

Friend from IRC was asking...
"Just wondering if it would be possible to add a simple tick box to the deinterlacing filter page which would run the frame doubler filter before the selected filter and if so please do so, Thanks in advance"
Considering that the frame doubler needs, if anything, progressive frames I really wonder what he hopes to achieve by that?

Also, proper full-rate deinterlacing is done by the bob filters already...

[Thunderbolt8]

btw. whats the reason that interlaced vc-1 content is not supported yet, do you have to wait for the ffmpeg guys to implement it or do you need more samples?

[_xxl]

new open source h.264 source:
The AMD Performance Library / Framewave covers a multitude of
> operations from simple math operations to media processing and
> optimizations for multi-core environments. Among the supported
> operations are H.264 video decoding. The Framewave project is housed
> over at SourceForge and at the AMD Developer Center.
http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2008-February/042298.html

[Yong]

guys, any fix about this? :p
DllEntry.c: In function 'DllMain':
DllEntry.c:48: warning: 'av_free_static' is deprecated (declared at libavcodec/avcodec.h:2493)
DllEntry.c:58:1: error: pasting "LIBAVCODEC_VERSION_MAJOR" and "." does not give a valid preprocessing token
DllEntry.c:58:1: error: pasting "." and "LIBAVCODEC_VERSION_MINOR" does not give a valid preprocessing token
DllEntry.c:58:1: error: pasting "LIBAVCODEC_VERSION_MINOR" and "." does not give a valid preprocessing token
DllEntry.c:58:1: error: pasting "." and "LIBAVCODEC_VERSION_MICRO" does not give a valid preprocessing token
make: *** [DllEntry.o] Error 1

[dimzon]

Dear ffdshow developers, please, take look @ this thread (http://forum.doom9.org/showthread.php?p=1103680). Maybe it's ffdshow problem? Or maybe it's possible to add some workaround into ffdshow (at least - complete lock flip option via yet another option)

[nekrosoft13]

could someone let me know if ffdshow can handle EAC3? also know as DD+

[clsid]

EAC3 is not supported yet. It is unknown when or if it will be supported in the future.

[dimzon]

talking about my problem (http://forum.doom9.org/showthread.php?p=1103680) seems like this is ffdshow issue (read this (http://forum.team-mediaportal.com/codecs-external-players-55/strange-flip-behavior-using-ffdshow-35689/#post237734))

Dear developers, can you add some logging feature to understand what happens? Thanks.

[dimzon]

maybe MediaPortal broadcast some magic WM_XXX message for it's own needs every time playback started and independend message with same code are in ffdshow (occurs message number collision)?

Anyway I need some assistance from ffdshow developers

[nekrosoft13]

EAC3 is not supported yet. It is unknown when or if it will be supported in the future.

thanks for responce

[nautilus7]

Thanks for graphedit bug fix.

[leeperry]

any chance seeing the winamp2 plugin bugs fixed ? :cool:

basically it doesn't seem to be closed properly, so you can't open 2 files in a row, it crashes the player(mplayerc.exe and mplayer2.exe).

but if you click on "CONFIGURE" in the winamp2 section of ffdshow audio, then it works....but after 3 files in a row it's using 70% of CPU time on my 3Ghz C2D :(

problem is easy to reproduce with this plugin :
http://www.winamp.com/plugins/details/79374

if you could also offer the option to hide the plugin GUI(like winamp does), that would be supreme coolness :p

could someone let me know if ffdshow can handle EAC3? also know as DD+

use the sonic audio decoder.

[clsid]

Please stop repeating your requests over and over again. It won't make a fix appear faster. Its just plain annoying.

Haruhiko already said that he will have a look at it when he has time.

[leeperry]

sorry, didn't mean to be annoying..

just wanted to give a clearer explanation in a few words, that's all.

:thanks:

[Inventive Software]

Since your requests seem only to be with this one particular plugin, have you tried with other DSP plugins? ;)

[leeperry]

I did.

http://forum.doom9.org/showpost.php?p=1096605&postcount=3111

[dustobub]

Has anyone noticed that the XXL builds with the cabac patch have broken libavcodec's VC1 support. Video does play, but has significant artifacting and can lead to a program crash. Any idea's, or should I be asking over at FFMpeg. Thanks.

Dustin

[dustobub]

FFMpeg devs now are aware of the bug. It looks like it was introduced around build 1858.

[_xxl]

Has anyone noticed that the XXL builds with the cabac patch have broken libavcodec's VC1 support.
That patch has nothing to do with VC-1.

[dustobub]

You are correct, the timing of the bug however made me think it was the cabac patch, my bad.

[Jeremy Duncan]

Is the libmpeg codec updated each time libavcodec is updated?

I look at the svn and I see all these updates for libavcodec, and Iwonder about this.

Also, can somebody tell me if the newer ffdshow is faster when comparing the libmpeg codec to previous versions of ffdshow?

[ACrowley]

any chance seeing the winamp2 plugin bugs fixed ? :cool:

basically it doesn't seem to be closed properly, so you can't open 2 files in a row, it crashes the player(mplayerc.exe and mplayer2.exe).

but if you click on "CONFIGURE" in the winamp2 section of ffdshow audio, then it works....but after 3 files in a row it's using 70% of CPU time on my 3Ghz C2D :(

problem is easy to reproduce with this plugin :
http://www.winamp.com/plugins/details/79374

if you could also offer the option to hide the plugin GUI(like winamp does), that would be supreme coolness :p



use the sonic audio decoder.

Problme is :

Sonic Decoder applies a ugly Mix of DynamicRangeCompression+DialogNorma.+gain

The Decoder System Settings have no Effect.

Nero AudioDecoder2 applies DRC+DialNorm too

So we have no Decoder fro Realtime with the ability to disable DRC/DialNorm = bad

So lot of People are waitng for a ffdshow with EAC3/THD support

[fastplayer]

Is the libmpeg codec updated each time libavcodec is updated?
No, different projects.
libmpeg2 development was finished years ago. From time to time minor bugfixes keep trickling in though.

[Thunderbolt8]

I have a problem with a H.264 broadcast here with ffdshow (1870 xxl), the picture is pretty much distorted, while its fine with CoreAVC and VLC (0.8.6d).

I made a short sample: http://www.sendspace.com/file/rk4epu

[fastplayer]

Clip plays fine here with rev1868.

[clsid]

Fine too with 1872.

[Thunderbolt8]

tested it with 1868 clsid, but strangely for me it looks the same as before, like this: http://www.sendspace.com/file/tixf46

what could be the reason for that??

[dustobub]

FFMpeg devs now are aware of the bug. It looks like it was introduced around build 1858.

I know this is more of an FFMpeg issue, but here is a link to a 1080p VC1 sample if you are curious.

[dustobub]

The VC1 bug looks like it was fixed last night. Build 1873 fixes it.

[ImAhNoBoDy]

I recently ripped subtitles from a Japanese DVD using VSRip. There are 4 different languages in the idx/sub file. The default sub is in Japanese but when I try to choose English, WMP crashes all the time. The sub loads by itself cause it has the same filename, but when I manually choose the subtitle myself, I am able to choose English. I mux the subs in mkv and it works just fine. Vobsob from VSfilter works just fine too when I enable it (I mostly use ffdshow to display subs).

Now I have tested other idx/sub files that has multiple languages and they are just fine. I have a sample: http://www.sendspace.com/file/inqfew

[foxyshadis]

tested it with 1868 clsid, but strangely for me it looks the same as before, like this: http://www.sendspace.com/file/tixf46

what could be the reason for that??

What splitter do you use?

[Thunderbolt8]

the haali splitter & haali renderer. I already tried it with mpcs internal splitter and also with overlay mixer and VMR9 as well, but the result stays the same.

The VC1 bug looks like it was fixed last night. Build 1873 fixes it.
before im going to post on this as well, could this be a bug with funny blue like colours, which appear then and when?

[leeperry]

hey guys,

simple question if you don't mind.

I was told by Seb.26 that checking the "high quality YV12 to RGB conversion" was actually converting to YUY2 then to RGB ?!

atm I'm decoding h264 with ffdshow from xxl_, then upscaling in lanczos 4X and sometimes deinterlacing.

is there any point to output from ffdshow in YUY2 over YV12 ?

I then send it to EVR with MPC HC, use some PS scripts in MPC HC, and either EVR or my ATi drivers are doing the RGB32 conversion in hardware.

I've made some screenshots, the CRC is different but I can't see any difference between YV12 and YUY2 with h264 decoded by ffdshow ?! :eek:

TIA,

[haruhiko_yamagata]

hey guys,

simple question if you don't mind.

I was told by Seb.26 that checking the "high quality YV12 to RGB conversion" was actually converting to YUY2 then to RGB ?!Yes, he's correct. YV12->RGB converter by Xvid project is fast, while YUY2->RGB converter by AviSynth project is high quality.


atm I'm decoding h264 with ffdshow from xxl, then upscaling in lanczos 4X and sometimes deinterlacing.

is there any point to output from ffdshow in YUY2 over YV12 ?
If you don't see any difference, there's no point.
Implementation of YV12 by the downstream filter may be buggy. YUY2 output is useful only in such cases.
Another use of YUY2 output is MJPEG. Most MJPEG files are decoded to YUY2 (strictly speaking, 422P).
If you use text subtitles, RGB output make sense.

[leeperry]

ok thanks for the reply haruhiko, so to be on the safe side I can just output in YUY2 from ffdshow I guess ;)

[soulkeeper]

I want to update my current ffdshow version(1485.xxl 26/09/08) to 1870.xxl...
Do i simple select uninstall and install new version?
No need for registry cleaning e.t.c?
Thanks...

[clsid]

I am having some resizing troubles with DVD playback. The height of the video is too small. Happens with both libavcodec and libmpeg2.

Input: 720x576, SAR: 64/45, DAR: 16:9
Output: 1280x1024

With MPEG-4 and H.264 video the resizing is done correctly. Those files have a SAR of 1/1, so I suspect the problem related to that. The height of the (wrongly) resized dvd video is approximately cunsuming 45/65 of the height of my screen.

[clsid]

I want to update my current ffdshow version(1485.xxl 26/09/08) to 1870.xxl...
Do i simple select uninstall and install new version?
No need for registry cleaning e.t.c?
Thanks...
You can just install on top of your current installation. No need for uninstallating or registry cleaning.

[soulkeeper]

You can just install on top of your current installation. No need for uninstallating or registry cleaning.

Thanks!!

[fofwp]

I have this problem when using ffdshow for decoding the audio of DVDs. At certain parts the audio sounds as though it is being clipped. I have found a particular scene where this happens and played it back using ffdshow on two different PCs. Both produced the same clipping sound.

I have tried the internal AC-3 decoder of MPC-HC and the NVidia Audio Decoder and neither display this behavior.

The only thing that is checked in ffdshow is the Mixer. All other processing is disabled. Can someone tell me what can be causing this?

[Px]

I am having some resizing troubles with DVD playback. The height of the video is too small. Happens with both libavcodec and libmpeg2.
NVIDIA Videocard?

[Thunderbolt8]

the haali splitter & haali renderer. I already tried it with mpcs internal splitter and also with overlay mixer and VMR9 as well, but the result stays the same.
well, any idea what I could do now? ive already installed & uninstalled other versions, but its still the same :S

[wyrd]

Hi,Thunderbolt8
I can reproduce when i uncheck "skip deblocking always".
(skip deblocking checked then work fine.)
snapshot (http://tirnanog.fate.jp/tmp/snap/test-001.mkv.snapshot.jpg)
regards

[clsid]

NVIDIA Videocard?
Nope. ATI 9200 se.

[Thunderbolt8]

Hi,Thunderbolt8
I can reproduce when i uncheck "skip deblocking always".
(skip deblocking checked then work fine.)
snapshot (http://tirnanog.fate.jp/tmp/snap/test-001.mkv.snapshot.jpg)
regards
thank you very much, that was indeed the reason! :)

so this basically means deblocking works faulty for this movie and therefore needs to be turned off here? luckily is this the only h264 movie with this issue I had so far.

[nm]

thank you very much, that was indeed the reason! :)

so this basically means deblocking works faulty for this movie and therefore needs to be turned off here? luckily is this the only h264 movie with this issue I had so far.
It means that there is a bug somewhere and it should be fixed. As you previously noticed, the clip plays fine with deblocking on other decoders and players, even ones based on libavcodec (MPlayer, VLC).

[haruhiko_yamagata]

I have a problem with a H.264 broadcast here with ffdshow (1870 xxl), the picture is pretty much distorted, while its fine with CoreAVC and VLC (0.8.6d).

I made a short sample: http://www.sendspace.com/file/rk4epuWorks if I set "Number of decoding threads:" = 1.

[clsid]

This is one of those cases that shows why the multithreading patch is still experimental.

[Thunderbolt8]

ok then ill just use the clsid version to decode this one :P

Ive got another question, since I un- and re-installed ffdshow I dont know any more which was the default renderer for mpeg1 and 2 video - libmpeg2 or libavcodec?
or better asked whats the difference between those 2 anyway and which one offers the better PQ for HD mpeg2 files (TV/blu-ray)?

[nm]

Ive got another question, since I un- and re-installed ffdshow I dont know any more which was the default renderer for mpeg1 and 2 video - libmpeg2 or libavcodec?
or better asked whats the difference between those 2 anyway and which one offers the better PQ for HD mpeg2 files (TV/blu-ray)?
I don't know which one is the default, but libavcodec's decoder has much better error tolerance when dealing with damaged TV caps (it beats most other MPEG-2 decoders in this, as can be seen in MSU's comparison: http://www.compression.ru/video/codec_comparison/mpeg-2_2006_en.html). With good streams, output quality is the same for both.

[dimzon]

Some stupid question
I can see ffdshow changelog here (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/?view=log)
Where I can see libavcodec changelog? Since libavcodec is 90% of ffdshow I must look at them to decide - download or not latest version.

[fastplayer]

Where I can see libavcodec changelog?
http://svn.mplayerhq.hu/ffmpeg/trunk/libavcodec/?view=log

[clsid]

ok then ill just use the clsid version to decode this one :P
My version also has multithreading because the patch was committed to SVN. You must set the number of threads to 1. Either permanently or just when you encounter a problematic file.

Ive got another question, since I un- and re-installed ffdshow I dont know any more which was the default renderer for mpeg1 and 2 video - libmpeg2 or libavcodec?
libmpeg2 is the default one for MPEG-2.
Disabled is the default for MPEG-1. Then the Microsoft decoder will be used that is part of DirectX.

[Thunderbolt8]

I don't know which one is the default, but libavcodec's decoder has much better error tolerance when dealing with damaged TV caps (it beats most other MPEG-2 decoders in this, as can be seen in MSU's comparison: http://www.compression.ru/video/codec_comparison/mpeg-2_2006_en.html). With good streams, output quality is the same for both.
interesting comparison indeed, thanks!
about the relationship of libmpeg2 and libavcodec for mpeg2 decoding, how is it, was libavcodec written by the same people who already made libmpeg2 and was meant to be an improvement?
My version also has multithreading because the patch was committed to SVN. You must set the number of threads to 1. Either permanently or just when you encounter a problematic file.
so I guess you and xxl are both using the same basis now and from now on it shouldnt matter any more which one I download? or do the xxl version still include some, well 'beta' material or however I should put this?

[clsid]

Afaik libmpeg2 and libacodec are two completely separated projects.

Current builds are all just plain SVN, no extra patches.

[rickardk]

Something is wrong with AC3 decoding I think.
Just watched Michael Clayton (m2ts from Blu-ray) and I get "pop/click" sounds all through the movie. I had this problem with other titles before but never so frequent as with this one.

If I use AC3filter or Sonic Audio decoder the sound plays without pops/clicks.

A small sample can be found here:
http://www.earselect.se/sample.mkv
http://www.earselect.se/sample.ac3

The click/pop is between 00:00:02 and 00:00:03. Located in the left front speaker.

[fastplayer]

The click/pop is between 00:00:02 and 00:00:03. Located in the left front speaker.
Confirmed.

If I use AC3filter or Sonic Audio decoder the sound plays without pops/clicks.
Add MPC's internal AC3 filter to that list.

[rickardk]

Confirmed.


Add MPC's internal AC3 filter to that list.

+ Nero and Cyberlink

[nm]

interesting comparison indeed, thanks!
about the relationship of libmpeg2 and libavcodec for mpeg2 decoding, how is it, was libavcodec written by the same people who already made libmpeg2 and was meant to be an improvement?
They have different authors, licenses and codebases. Libmpeg2 was mostly written by Aaron Holtzman and Michel Lespinasse and libavcodec's mpeg12 decoder was written by Fabrice Bellard and Michael Niedermayer. Libmpeg2 is licensed under GPL and lavc is LGPL, so the latter couldn't borrow code from the former without permission from the original authors.

I don't know the history well, but I might guess that the FFmpeg project (or Fabrice) wanted to have their own MPEG-1/2 decoder, and since some of the code could be shared with other decoders (or was already written for the MPEG-4/DivX decoder), it was best to write it from the ground up.

[fastplayer]

http://img03.picoodle.com/img/img03/4/2/29/f_NewBitmapImm_a6c4d0e.png

Shouldn't the "DVD decoding" check box be greyed-out when MPEG2 decoder is set to "disabled"? Or am I missing something here?

[Thunderbolt8]

maybe its meant as exception so that you dont want to use ffdshow to decode mpeg2 files in general, but still could use it only for DVDs.

[Leak]

maybe its meant as exception so that you dont want to use ffdshow to decode mpeg2 files in general, but still could use it only for DVDs.
Ah, but using which library? libmpeg2 or libavcodec? ;)

np: The Orb - Mother Nature (The Dream)

[clsid]

The internal AC3 decoder in MPC uses liba53, same as ffdshow. So perhaps the error is caused by the Mixer or any of the other audio filters in ffdshow?

[fastplayer]

The internal AC3 decoder in MPC uses liba53, same as ffdshow. So perhaps the error is caused by the Mixer or any of the other audio filters in ffdshow?
Even with everything turned off (Volume, Mixer and Dynamic range compression), it still clicks.

[clsid]

Just tested the file. I am getting a pop between second 4 and 5 instead of between 2 and 3. But only with WMVideo DMO decoder. With libavcodec everythings seems fine.

The .ac3 file play without any unusual sounds here. Unless I am getting deaf :)

[Yong]

if i select libtheora as decoder for theora,
avisynth filter "buffer back/ahead" value equal or bigger than 1 will cause ffdshow.ax crash.
its happen only if i select yv12 as input colorspace.
im using latest xxl build.

[rickardk]

Just tested the file. I am getting a pop between second 4 and 5 instead of between 2 and 3. But only with WMVideo DMO decoder. With libavcodec everythings seems fine.

The .ac3 file play without any unusual sounds here. Unless I am getting deaf :)


I'm getting the pop regardless of video decoder and also when I just play the AC3 file. Also tested muxing it into mka.

All decoders I tested except ffdshow (liba52) play this without pop/click. So I guess it must be a ffdshow/liba52 problem.

It's a very discrete pop though. But it's there...

[SeeMoreDigital]

The .ac3 file play without any unusual sounds here. Unless I am getting deaf :)I don't hear unusual sounds here either...

To confirm. I'm passing the (FFdshow) decoded bit-stream to an external DSS amplifier.


Cheers

[rickardk]

You should not pass the bit stream I think. You should decode the AC3 stream with ffdshow.

Also, if you don't hear the pop you can try with another AC3 decoder and then switch back to ffdshow. That might help you spot the pop/click.

I'm 100% sure it's there. Tested latest ffdshow (clsid) on clean XP SP2 and clean Vista Ultimate 32 (both with and without SP1). Same problem...

[fastplayer]

Just ripped a DVD and wanted to play it back with MPC and libavcodec's MPEG2 decoder. I double-clicked the VIDEO_TS.IFO and all I got was stutter. The video would literally freeze every half second.
Now when I directly start DVD playback (circumventing the intro and menu) by loading the VTS_0x_0.IFO, then all is OK.

I compared MPC's filter list and noticed that the stutter would only occur when the "DVD Navigator" filter was used. This filter loads with the VIDEO_TS.IFO but it does not with the VTS_0x_0.IFO (instead MPC's MPEG2 splitter kicks in).

If I use libmpeg2 - whether it's MPC's internal or ffdshow's version - then I have no problems at all.

So why does the "DVD Navigator" take precedence over MPC's internal MPEG2 splitter? :confused:

[clsid]

Without it you won't get DVD menus and thus you need it for proper DVD playback. Libavcodec just doesn't work very well in combination with DVDs.

[fastplayer]

Libavcodec just doesn't work very well in combination with DVDs.
The stutters mentioned above cause frame drops even though CPU load is less than 20%. :confused:

Everytime I give libavcodec a chance at DVD playback, it fails.
But as long as I have a working alternative, I won't complain. :)

[Thunderbolt8]

Ah, but using which library? libmpeg2 or libavcodec? ;)

np: The Orb - Mother Nature (The Dream)
cant tell, it was just a guess ;)

[fofwp]

Something is wrong with AC3 decoding I think.
Just watched Michael Clayton (m2ts from Blu-ray) and I get "pop/click" sounds all through the movie. I had this problem with other titles before but never so frequent as with this one.

If I use AC3filter or Sonic Audio decoder the sound plays without pops/clicks.

A small sample can be found here:
http://www.earselect.se/sample.mkv
http://www.earselect.se/sample.ac3

The click/pop is between 00:00:02 and 00:00:03. Located in the left front speaker.

I can hear what you are talking about when using ffdshow for decoding. When I use the AC3 decoder built into MPC-HC I don't hear it.

I have also experienced something similar which I had posted about earlier, but have not received a response for. I have quoted it below.


I have this problem when using ffdshow for decoding the audio of DVDs. At certain parts the audio sounds as though it is being clipped. I have found a particular scene where this happens and played it back using ffdshow on two different PCs. Both produced the same clipping sound.

I have tried the internal AC-3 decoder of MPC-HC and the NVidia Audio Decoder and neither display this behavior.

The only thing that is checked in ffdshow is the Mixer. All other processing is disabled. Can someone tell me what can be causing this?

If someone can give me instructions about how to cut a part of the ac3 track and where to upload it. I would be glad to do it if it means that someone will look into this. A thing to note is that I hear this when using the analog out (not digital) from two computers with different sound cards

[haruhiko_yamagata]

@Haruhiko:

With rev1626 (http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout?view=rev&revision=1626) you fixed a bug where subtitles with an "&" in the filename were incorrectly shown in ffdshow's context menu.
Unfortunately, your bugfix introduced another bug:
Sorry for late reply and thank you.
I have fixed at rev 1877.

[Leak]

The stutters mentioned above cause frame drops even though CPU load is less than 20%. :confused:

Everytime I give libavcodec a chance at DVD playback, it fails.
Well, for me I get the video playing at twice the speed for half a second followed by a half-second pause and so on.

I'd hazard a guess that in the DVD case the timestamp generation from libavcodec is somehow whack, but I'm not really familiar with it's innards...

np: The Orb - Codes (The Dream)

[fastplayer]

Sorry for late reply and thank you.
I have fixed at rev 1877.
No worries. I'm patient :)
I noticed that your fix now replaces "&&" with "&". I wonder what'll happen when you have actually 2 ampersands in the path/filename... I'll test the "special" cases as soon as a build gets released.
Anyway, :thanks:

Well, for me I get the video playing at twice the speed for half a second followed by a half-second pause and so on.
This might actually be a better explanation of what I noticed. I was too fixed on the pauses itself.

By the way, the "DVD decoding" check box "adds" the MEDIATYPE_DVD_ENCRYPTED_PACK media type to the decoder. That's how far I got looking at the code. I remember that in earlier ffdshow releases you had to add it manually to MPC in order to get ffdshow to decode DVDs.

[cc979]

You should not pass the bit stream I think. You should decode the AC3 stream with ffdshow.

Also, if you don't hear the pop you can try with another AC3 decoder and then switch back to ffdshow. That might help you spot the pop/click.

I'm 100% sure it's there. Tested latest ffdshow (clsid) on clean XP SP2 and clean Vista Ultimate 32 (both with and without SP1). Same problem...

try decoding the audio with ffdshow and load it into an audio editor you maybe able to see it

[haruhiko_yamagata]

try decoding the audio with ffdshow and load it into an audio editor you maybe able to see it
I did and I can see it. Still it's difficult to fix. I don't know where to start.

[_xxl]

Something is wrong with AC3 decoding I think.
Just watched Michael Clayton (m2ts from Blu-ray) and I get "pop/click" sounds all through the movie. I had this problem with other titles before but never so frequent as with this one.
#if defined(ARCH_X86) && defined(__GNUC__)
...
#endif //ARCH_X86
Liba52 has some MMX, 3DNOW! and SSE code from mplayer for (down)mix available only for MINGW GCC builds.
if(mm_accel & FF_CPU_SSE)
a52_imdct_512 = imdct_do_512_sse;
And some SSE for imdct available only in MSVC or ICL builds.
Try to disable them from ffdshow?

[haruhiko_yamagata]

if(mm_accel & FF_CPU_SSE)
a52_imdct_512 = imdct_do_512_sse;
And some SSE for imdct available only in MSVC or ICL builds.
Try to disable them from ffdshow?Good suggestion. It works without imdct_do_512_sse :).

[haruhiko_yamagata]

Something is wrong with AC3 decoding I think.
Just watched Michael Clayton (m2ts from Blu-ray) and I get "pop/click" sounds all through the movie. I had this problem with other titles before but never so frequent as with this one.

If I use AC3filter or Sonic Audio decoder the sound plays without pops/clicks.

A small sample can be found here:
http://www.earselect.se/sample.mkv
http://www.earselect.se/sample.ac3

The click/pop is between 00:00:02 and 00:00:03. Located in the left front speaker.
Mplayer have the same click.
The click is located at 3.122 seconds from the start.
If I use a52_imdct_512_C instead of imdct_do_512_sse, there is no click. Could you post a report to mplayer project?
imdct_do_512_sse is mplayer specific, so please do not post a report to lia52.

[_xxl]

I would like to remove that MMX,3dnow and SSE from liba52?
First is used only in gcc builds, rare and second doesn't work.

[clsid]

If you have some time, maybe you could also attempt to add the libavcodec AC3 decoder.

[rickardk]

Mplayer have the same click.
The click is located at 3.122 seconds from the start.
If I use a52_imdct_512_C instead of imdct_do_512_sse, there is no click. Could you post a report to mplayer project?
imdct_do_512_sse is mplayer specific, so please do not post a report to lia52.

Sorry, I don't know where to post a report to mplayer project.
(don't know if your response was directed to me?)

[haruhiko_yamagata]

(don't know if your response was directed to me?)
Yes.

Bug reports about MPlayer belong on "MPlayer-users" mailing list (http://www.mplayerhq.hu/design7/mailing_lists.html).
Writing a good bug report is a lot of work. Please help us.
And it will help not only us but also Mplayer development.

You can get the latest binary from MPlayer on Win32 (http://sourceforge.net/project/showfiles.php?group_id=205275).

[rickardk]

Read through the bug report guide for mplayer.
This seems to be way over my head...

Before I start dig in to this (I'm not that familiarised with development of ffdshow....yet):

Is mplayer or part of it used by ffdshow to decode AC3?
Or is it just some small part of code from mplayer?

----------------
"If I use a52_imdct_512_C instead of imdct_do_512_sse, there is no click. Could you post a report to mplayer project?
imdct_do_512_sse is mplayer specific, so please do not post a report to lia52."


Two diffrent AC3 decoders used by mplayer?

[chros]

I am having some resizing troubles with DVD playback. The height of the video is too small. Happens with both libavcodec and libmpeg2.

Input: 720x576, SAR: 64/45, DAR: 16:9
Output: 1280x1024

With MPEG-4 and H.264 video the resizing is done correctly. Those files have a SAR of 1/1, so I suspect the problem related to that. The height of the (wrongly) resized dvd video is approximately cunsuming 45/65 of the height of my screen.
r1846 is working perfectly (both libmpeg and libav) with "Set aspect ratio..." checked, MPCHC with builtin splitter, Haali's renderer ...
I tried with decrypted files on HDD:
- VIDEO_TS.IFO : aspect is correct, but stutters like hell
- VTS_01_0.IFO : no problem at all
- VTS_01_1.VOB : no problem at all

Sample file ?

[haruhiko_yamagata]

if i select libtheora as decoder for theora,
avisynth filter "buffer back/ahead" value equal or bigger than 1 will cause ffdshow.ax crash.
its happen only if i select yv12 as input colorspace.
im using latest xxl build.I can reproduce this.
@Leak, Could you take a look at this? Do you need a sample?

[fastplayer]

Sorry for late reply and thank you.
I have fixed at rev 1877.
I noticed that your fix now replaces "&&" with "&". I wonder what'll happen when you have actually 2 ampersands in the path/filename... I'll test the "special" cases as soon as a build gets released.
Sorry for my doubts! It works perfectly! :)
You can now have as many ampersands in the path as you like.
:thanks:

[Leak]

I can reproduce this.
@Leak, Could you take a look at this?
Sure, but probably no sooner than this weekend.

Do you need a sample?
Fire away, seeing as I have exactly zero theora files here... :)

[Thunderbolt8]

though more H.264 features have been implemented in ffdshow meanwhile, I still have the same problem with raindow frames with the 1st EP of the band of brothers HD DVD, from which I already posted a sample some months ago. here it is again:

http://www.sendspace.com/file/g2m1jt

tested it with xxl 1878 and the rainbows are in both, the source evo and a .mkv remux. in contrast both (a remux to .mkv and this .evo source sample here) work fine with coreavc. so could anyone please have a look at this again?

I have the same problem with the Babel HD DVD, though there the rainbow frames occur only at some places, while for this 1st band of brothers episode it goes throughout the whole rest of the movie (I guess once you hit them it stays until you close/open the mpc again). couldnt make a sample from that one though, because its in the mid of the movie and I dont know how to avoid to hit open GOPs etc. so that the sample would most likely not play properly.

[clsid]

Test it with a recent MPlayer build. If that has the same problem, then it is a FFmpeg bug and should be reported to FFmpeg.

[Thunderbolt8]

theres a bit of a problem here, I downloaded the lastest version I could find (2008-02-10), but both .mkv remux version from band of brothers and babel stopped right after few seconds playing. I also was not able to jump forward or backwards at all in the stream.

and when I tried to open the .evos from both movies no picture occured at all.
so is it already possible to say from this point that it could be an ffmpeg bug or not?

[clsid]

The file seems to play ok on my old cpu, same images as with coreavc, but there are too many frame skips to be absolutely sure. So you should test with SSE/SSE2/etc disabled in ffdshow debug options. You can keep MMX/MMXext enabled. I suspect the bug might be in the assembly code of libavcodec.

[Thunderbolt8]

just to be sure I disabled all those of the SIMD instructions in the 'info & cpu' field, I guess thats what you meant. but it looks still the same as before.

[clsid]

Can you post a screenshot of what I am supposed to see?

[Thunderbolt8]

left the osd activated accidently, but shouldnt matter.
this is from the original .evo source, the sample and .mkv remux looks the same:

http://www.image-load.eu/out.php/i30091_snapshot20080304232408.PNG
http://www.image-load.eu/out.php/i30093_snapshot20080304232414.PNG
http://www.image-load.eu/out.php/i30094_snapshot20080304232422.PNG

[B.F.]

ffdshow don't load external ass subtitles.
If I change extension to ssa, subtitles work fine.

[thuan]

Add ass to the accepted subtitle extensions.

[B.F.]

It don't work.

[haruhiko_yamagata]

ffdshow don't load external ass subtitles.
If I change extension to ssa, subtitles work fine.
How did you configure ffdshow? A screen shot of the dialog should be enough.

// EDIT
rev 1884 should work. Please test.

[DMD]

You excuse the ignorant person question.
And' possible currently to use ffdshow as post processing with Power DVD also acting on the register of windows?

Cordially I thank
Domenico

[leeperry]

you can force Reclock in PDVD on XP, but not ffdshow I think.

that would be the shiznit, though :eek:

[DMD]

you can force Reclock in PDVD on XP, but not ffdshow I think......
have confirmed me that Power DVD to it has the protected runs, and that the postprocessing is not permitted
http://www.avsforum.com/avs-vb/showthread.php?p=13297813#post13297813

regards
Domenico

[leeperry]

on Vista, you can't force Reclock with PDVD 7.3 Ultra

on XP you can, with Reclock 1.4

XP doesn't have the same security enforcment that Vista has :eek:

[Shirokuu]

I've been inactive for a while, but nonetheless I have a question.

Is it possible to automatically download the SVN and nightly builds? I am configuring my NAS to grab every night the last software updates, the news (as in plain ASCII text) and i thought it was an interesting idea to grab the latest FFDSHOW also.

I can schedule the downloads by providing a download path on either a website or FTP site (or local drive if I wanted to do some backing up). So basically my question would be what path should I point to, to grab the nightly builds on SourceForge for FFDSHOW? (If it is possible, that is).

[clsid]

Install a Subversion (SVN) client and you can use that to update to the latest source code.

[Thunderbolt8]

any idea what else I could do about my band of brothers & babel video problem?

[clsid]

1) Ask Celtic_Druid if he can compile a new version of FFmpeg (which includes ffplay.exe).
http://celticdruid.no-ip.com/xvid/
2) Play your samples with FFplay.exe to see if that fails as well.
3) Report it to FFmpeg:
http://ffmpeg.mplayerhq.hu/bugreports.html
https://roundup.mplayerhq.hu/roundup/ffmpeg/
4) Wait

[LotharZ]

I'm having problems with an audio codec (sowt: 16-bit signed little-endian). If I enable it, ffdshow consume all my CPU (Sempron 2800+).

thx

ffdshow 1874 + MPC HC 1.0.11.8

Trailer: http://files.ea.com/downloads/eagames/cc/tiberium/KanesWrath_shortcineTrailer_HD.mov

[wyrd]

It seems fine here.
http://tirnanog.fate.jp/tmp/snap/KanesWrash_CPUusage.jpg
rev1882 icl10 on C2D.
Thanks

[haruhiko_yamagata]

@LotharZ
It works for me. My CPU is Core2Duo.

[fastplayer]

@LotharZ
It works for me. My CPU is Core2Duo.
Are you sure you're using ffdshow to decode the video too?
For me, it's just unplayable.

[Eragon4ever]

Plays perfectly fine in WMP11 here. CPU useage is ~10-20%.
In MPC-HC the sound is garbage, the video plays very slow and CPU is at 50%.

Using Core 2 Duo @ 2,66GHZ

Edit: The problem is MPC-HC's internal mp4 splitter. Disabling it and using Haali instead fixes the playback for me.

[clsid]

@LotharZ
Report the above findings to Casimir, then maybe he can fix the internal splitter in MPC-HC.

[Thunderbolt8]

1) Ask Celtic_Druid if he can compile a new version of FFmpeg (which includes ffplay.exe).
http://celticdruid.no-ip.com/xvid/
2) Play your samples with FFplay.exe to see if that fails as well.
3) Report it to FFmpeg:
http://ffmpeg.mplayerhq.hu/bugreports.html
https://roundup.mplayerhq.hu/roundup/ffmpeg/
4) Wait
I pmed celtic_druid about it, hopefully he will have time to compile a new version soon. currently I used the latest one available to test (from february 5th, 2008) and when I test this band of brothers sample with this version the same problems are present there as well.
(though I could only test it with the .mkv remux, since nothing happened when I chose the .evo as input file)

[B.F.]

How did you configure ffdshow? A screen shot of the dialog should be enough.

// EDIT
rev 1884 should work. Please test.
It work in rev1890.
:thanks:

[dimzon]

Some feature request (http://forum.doom9.org/showthread.php?p=1109647#post1109647)

[Thunderbolt8]

there could to be a little bug with libavcodec regarding some mpeg2 streams, the blu-ray video from 'memento' has approx. 0,5 seconds video corruption after seeking at some places. using libmpeg2 its all fine.
its not a big problem, as its no problem with the actual video and doesnt occur at normal playback, only after seeking. nevertheless I made a 20mb sample here in case someone would like to have a look at it:

http://www.sendspace.com/file/vbg4o0

[clsid]

I think that is a "known" problem/limitation.

[clsid]

FFplay rev12362 (http://www.zshare.net/download/86216137fe177c/)

[Thunderbolt8]

thanks! tested it already with the .mkv remux and the problem is still present there (still couldnt test it with the .evo, though one time the playback box even came up, but then nothing was displayed there other than black).
guesss Ill write that report to ffmpeg then tomorrow evening. the sample has a size of 25 mb, could this be a problem, as it was written there it should be trimmed if its larger than 10mb? I cant really make it shorter, since it takes a little until the problem shows itself.

I have registered at roundup.mplayer, I guess what Ill have to do then is to create a new task and describe my problem in that big box? theres also a bar to upload a file I guess, should I upload my sample there then or at that upload.mplayer ftp and just mention I in that commentary field I uploaded the sample there?

[clsid]

Upload to the ftp. I don't think 25mb will be a problem. The 10 mb size is not a hard limit, just a note to prevent people from uploading huge files instead of cutting a sample.

I don't think FFmpeg has a working EVOB parser yet, so that would explain why the .evo won't play. Here ffplay.exe just freezes.

[Eragon4ever]

I've got a lot videos which play with disorted colours here. If I change the output colorspace from YV12 to RGB32 everything is fine.

Sample: http://rapidshare.com/files/98092875/Ragnarok_2.avi.html

[fastplayer]

I've got a lot videos which play with disorted colours here. If I change the output colorspace from YV12 to RGB32 everything is fine.

Sample: http://rapidshare.com/files/98092875/Ragnarok_2.avi.html
No probs here with latest MPC/ffdshow and VLC in WXP SP2 with X800GTO/CAT8.3.

[Leak]

if i select libtheora as decoder for theora,
avisynth filter "buffer back/ahead" value equal or bigger than 1 will cause ffdshow.ax crash.
its happen only if i select yv12 as input colorspace.
im using latest xxl build.
Ick. Turns out libtheora produces YV12 images with a negative pitch... :eek:

I thought that was only ever done with RGB (which was what I fixed last time), but obviously the same fix needs to be applied to YV12 as well... :o

EDIT: should be fixed in rev. 1893

np: Tocotronic - Kapitulation (Kapitulation Live)

[Yong]

Ick. Turns out libtheora produces YV12 images with a negative pitch... :eek:

I thought that was only ever done with RGB (which was what I fixed last time), but obviously the same fix needs to be applied to YV12 as well... :o

EDIT: should be fixed in rev. 1893

np: Tocotronic - Kapitulation (Kapitulation Live)

yup, just compiled and tested latest build, no more crash ;)
:thanks:

[Jeremy Duncan]

They are telling me that using this configuration, and mpc clsid, that the screen resizes like it's flickering. ffdshow 1870 xxl, mpc 20080127.zip.
video and audio codec: ffdshow.
They say the problem is not present with beta 4.

________ Seesaw 2. For Seesaw 2, you need to use VMR9 Windowed to be able to calibrate the colors using your videocard ________

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

Avisynth tab
YV12 checked,
Add FFdshow Video source checked,
3:2 Pulldown box: ignore Pulldown checked,
Check Buffer back/Ahead: 1, 15

setmemorymax(1024)
FluxSmoothT(3)
MT("
a= last
b=a.Spresso(limit=1, limitC=1, bias=27, biasc=27, rgmode=17, rgmodec=17)
SeeSaw(a, b, NRlimit=0, NRlimit2=5, Sstr=0.90, Szp=1, Slimit=9, Sdamplo=16, Spower=1, sootheT=95, SdampHi=39)
",2)
ConvertToRGB24(matrix="pc.601")
chromaresample="Lanczos"

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

Resize & aspect tab
Multiply by:
1.780 (for 1280x720 16:9 aspect ratio)
2.668 (for 1920x1080 16:9 aspect ratio)
Process Pixel aspect ratio internally checked
No aspect ratio correction checked
Lanczos
Taps: 2
Luma Sharpen: 0.00
Accurate rounding checked

Queue & Output tab
Queue output samples checked

Output tab
rgb32 checked

[leeperry]

sorry for repeating, but I'm still eagerly hoping for a winamp2 plugin fix :(

here's what Reclock says when ffdshow audio crashes MPC HC :

298.59s 000690 ERROR **********************************************
298.59s 000690 ERROR Exception detected in player (not caused by reclock but logged here for your information)
298.59s 000690 ERROR Base address: 0x59410000 (my dll at 0x02030000)
298.59s 000690 ERROR Exception type:
298.59s 000690 ERROR Access violation at address 0x59471bd8 (base+rva)
298.59s 000690 ERROR Tried to read address 0x59471bd8
298.59s 000690 ERROR Exception flags: 0
298.59s 000690 ERROR More detailed information:
298.59s 000690 ERROR [00]: 0x00000000
298.59s 000690 ERROR [01]: 0x59471bd8
298.59s 000690 ERROR [02]: 0x0001003f
298.59s 000690 ERROR [03]: 0x00000000
298.59s 000690 ERROR [04]: 0x00000000
298.59s 000690 ERROR [05]: 0x00000000
298.59s 000690 ERROR [06]: 0x00000000
298.59s 000690 ERROR [07]: 0x00000000
298.59s 000690 ERROR [08]: 0x00000000
298.59s 000690 ERROR [09]: 0xffff027f
298.59s 000690 ERROR [10]: 0xffff0000
298.59s 000690 ERROR [11]: 0xffffffff
298.59s 000690 ERROR [12]: 0x00000000
298.59s 000690 ERROR [13]: 0x00000000
298.59s 000690 ERROR [14]: 0x00000000
298.59s 000690 ERROR Stack trace:
298.59s 000690 ERROR at 0x59471bd8 (0x00061bd8 relative): P:\winamp\plugins\dsp_iZOzone.dll
298.59s 000690 ERROR at 0x7c91378b (0x0000378b relative): C:\WINDOWS\system32\ntdll.dll

[haruhiko_yamagata]

sorry for repeating, but I'm still eagerly hoping for a winamp2 plugin fix :(

here's what Reclock says when ffdshow audio crashes MPC HC :iZotope Ozone Free works for me.
Do I have to combine anything to it to make ffdshow crash?

Recently I've tried very hard to fix VST host compatibility, with no success.

[leeperry]

well Ozone works, until you open a new file :(

then ffdshow audio crashes MPC HC, the issue seems to be with the way ffdshow closes the winamp2 plugin.

which means that if I have movies in 2 parts or mp3 playlists, I can't use them AT ALL :(

but if you go to the ffdshow audio configuration, select OZone then click on configure and close the Ozone GUI....then ffdshow audio doesn't crash MPC HC each time you open a new file, but after you've opened 3 files in a row it's sucking 70% of the CPU :(

both the Dx host and Ozone exhibit the same problem, so you might just kill 2 birds with one stone :)

thanks for your help, hopefully you can fix it as it seems like a minor bug ?!?! :eek:

and if you could offer an option to hide the winamp2 plugin GUI, that'd be really great :)

[rickardk]

Recently I've tried very hard to fix VST host compatibility, with no success.

WOW...
VST host compatibility would be great!!

Just tested IKs (Audyssey actually provides the core) room correction suite (includes mic) plugin for my studio. It bests even the Audyssey MultEQ XT in my Onkyo 905.

If you could get VST support in ffdshow working ok, we could have room correction that even could be preset based. A level of flexability not found in A/V receivers to date.

Plugin that handles 8channels is in the pipeline (rumors)

[leeperry]

you can use any winamp plugin....well if the bug that crashes the player can be fixed.

there is some VST/DX wrappers in winamp2 plugin forms ;)

[Atak_Snajpera]

Is it possible to implement TrueHD support in FFDShow?

[Inventive Software]

Good question. If the patch is *easily* compilable with libavcodec on Windows, then I don't see why not. I'd attempt it, but that means re-installing MinGW and re-compiling a load of stuff that I can't remember how to, having not done it seriously in about a year. :D

EDIT: It also depends on a TrueHD parser having been written and working on Windows.

[clsid]

I made a custom MinGW installer a few days ago. Here it is:
Download (http://www.zshare.net/download/8749833fa15553/) (8.8 MB)

With this you'll have a fully working and up-to-date MinGW/MSYS environment in a matter of seconds :)

Are there already any (open-source) DirectShow splitters that can parse TrueHD? Without those a decoder in ffdshow would not yet be useful.

If you want you could also attempt to add E-AC3 support. I think that is more wanted than TrueHD.

[Inventive Software]

Wow, thanks clsid. Wasn't expecting that! :)

AFAIK, no on the splitter front. A parser exists in ffmpeg for TrueHD as a patch, but I'd have to look at eac3to to see how it integrates E-AC3 and TrueHD decoding, and how easy it would be to add the relevant patches to ffdshow.

[SeeMoreDigital]

Is there any way that files containing DTS-HD or True-HD audio can be "down-converted" to their DTS or AC3 cores on-the-fly?


Cheers

[Inventive Software]

If eac3to can do it in faster than real-time, then there must be a way. I dunno about on the fly though.

[nautilus7]

DTS-HD on the fly to DTS isn't needed, because every DTS-HD track contains a simple DTS @ 1536 kbps core.

BTW, the new version of mkvtoolnix supports DTS-HD muxing and haali media splitter can pass the complete track to the decoder (not only the dts core). So using sonic audio decoder, dts-hd playback should work.

@ Inventive Software

You should talk to madshi... He said once that he might make a dshow filter for truehd and eac3 based on libav/ffmpeg decoder.

[Thunderbolt8]

what was the reason for disabling multithreading with mbaff/paff? was it buggy?

[haruhiko_yamagata]

well Ozone works, until you open a new file :Thanks. I can reproduce now.

[haruhiko_yamagata]

what was the reason for disabling multithreading with mbaff/paff? was it buggy?Multithreading worked for interlaced H.264 only if B frames were not used.

[leeperry]

Thanks. I can reproduce now.

OMG, I hope so much that you can fix this problem and add an option to always hide the winamp2 plugin GUI http://forum-images.hardware.fr/images/perso/kornichon.gif

my HTPC works pefectly now with MPC HC/EVR/Reclock/pstrip....the only last remaining issue is ffdshow crashing with Ozone :(

I'm far from a coder, but I've tried to find how to hide the plugin GUI, and it might have to do with the "Plugin.MainWindow" command :
http://www.codeproject.com/KB/audio-video/winampoutput.aspx

if you could make it always disabled, except if we click on "configure" in the ffdshow audio winamp2 plugin section, that would be marvelous :)

TIA,

[chros]

if you could always make it disabled, except if we click on "configure" in the ffdshow audio winamp2 plugin section, that would marvelous :)
+1 :)

Thanks

[georgevalkov]

Hello clsid, I have noticed that the H264 encodding feature is back to ffdshow, so I came to say "thank You" for that, and thanks for spending Your time on that great ffdshow project!

:thanks:

[Px]

And another one "thank You" for h264 encoder from me :)

[Inventive Software]

You have xxl to thank for that, not clsid. ;)

[georgevalkov]

:) Thank You, xxl! :thanks:

By the way where can I read more about the advantages and differences between the two builds: xxl's vs. clsid's SSE

I've been using clsid's SSE version in the past year, because of the SSE instruction set in its title. Or may be I should run a few benchmarks, to see, which is better? :rolleyes:

[Mangix]

SSE only applies to filter AFAIK

[fastplayer]

By the way where can I read more about the advantages and differences between the two builds: xxl's vs. clsid's SSE
http://ffdshow-tryout.sourceforge.net/html/en/faq.htm

[clsid]

It is also explained in the very first post of this topic.

[fastplayer]

Can ffdshow be updated to use SSE4.1 instructions on 45nm Core 2(Penryn core) cpus to speed up decoding of H.264? I saw from an earlier reply in this thread about Core 2(65nm Conroe) getting SSSE3 support to improve motion compensation.
And if you've looked further you would've noticed that this is not in our hands. Talk to the ffmpeg guys...

[clsid]

Read the other topics here on Doom9 that are about SSE4. The general conclusion is that using SSE4 will not bring any significant performance improvements compared to the existing implementations (which includes some SSSE3 code).

Furthermore, as mentioned several times before in this topic, we don't develop libavcodec. That is done by the FFmpeg project.

End of discussion.

[cc979]

anyone tried compile ffdshow-tryout svn1898 with gcc-4.3

i get this error:

make -C baseclasses
make[1]: Entering directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/baseclasses'
gcc-4.3 -c -DRELEASE -mno-cygwin -mdll -fno-rtti -mthreads -pipe -D_WINGDI_ -DUCLIBCPP -D_GLIBCPP_HAVE_MBSTATE_T -D_WIN32_IE=0x0500 -DARCH_IS_IA32 -DARCH_IS_32BIT -DHAVE_MMX -mmmx -w -DNDEBUG -UDEBUG -DFFDEBUG=0 -I. -I.. -Iuclibc++ -Ibaseclasses -I../baseclasses -IimgFilters -I../imgFilters -Implayer -I../mplayer -Isettings -I../settings -Isettings/filters -I../settings/filters -Icodecs -I../codecs -Isubtitles -I../subtitles -Iconvert -I../convert -Idialog -I../dialog -IaudioFilters -I../audioFilters -Icygwin -I../cygwin -Iffmpeg -I../ffmpeg -Iacm -I../acm -Ifilters -I../filters -Imuxers -I../muxers -I/dx/Include -L/dx/MingLib -ldx9 -O2 -march=pentium-mmx -mtune=i686 -fomit-frame-pointer -finline-functions -finline -frename-registers -fweb -funit-at-a-time -MMD -o baseclasses_all.o baseclasses_all.cpp
make[1]: Leaving directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/baseclasses'
make -C acm
make[1]: Entering directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/acm'
gcc-4.3 -c -mno-cygwin -mdll -fno-rtti -mthreads -pipe -D_WINGDI_ -DUCLIBCPP -D_GLIBCPP_HAVE_MBSTATE_T -D_WIN32_IE=0x0500 -DARCH_IS_IA32 -DARCH_IS_32BIT -DHAVE_MMX -mmmx -w -DNDEBUG -UDEBUG -DFFDEBUG=0 -I. -I.. -Iuclibc++ -Ibaseclasses -I../baseclasses -IimgFilters -I../imgFilters -Implayer -I../mplayer -Isettings -I../settings -Isettings/filters -I../settings/filters -Icodecs -I../codecs -Isubtitles -I../subtitles -Iconvert -I../convert -Idialog -I../dialog -IaudioFilters -I../audioFilters -Icygwin -I../cygwin -Iffmpeg -I../ffmpeg -Iacm -I../acm -Ifilters -I../filters -Imuxers -I../muxers -I/dx/Include -L/dx/MingLib -ldx9 -O2 -march=pentium-mmx -mtune=i686 -fomit-frame-pointer -finline-functions -finline -frename-registers -fweb -funit-at-a-time -MMD -o Tacm.o Tacm.cpp
In file included from ../imgFilters/avisynth/Tavisynth.h:5,
from Tacm.h:5,
from Tacm.cpp:21:
../imgFilters/avisynth/avisynth.h:697: error: conflicting type attributes specified for 'virtual GenericVideoFilter::~GenericVideoFilter()'
../imgFilters/avisynth/avisynth.h:549: error: overriding 'virtual IClip::~IClip()'
make[1]: *** [Tacm.o] Error 1
make[1]: Leaving directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/acm'
make: *** [lib] Error 2


cheers

[Leak]

anyone tried compile ffdshow-tryout svn1898 with gcc-4.3

i get this error:

In file included from ../imgFilters/avisynth/Tavisynth.h:5,
from Tacm.h:5,
from Tacm.cpp:21:
../imgFilters/avisynth/avisynth.h:697: error: conflicting type attributes specified for 'virtual GenericVideoFilter::~GenericVideoFilter()'
../imgFilters/avisynth/avisynth.h:549: error: overriding 'virtual IClip::~IClip()'

That has to be a bug in gcc, seeing as GenericVideoFilter doesn't even declare it's own destructor:

// instantiable null filter
class GenericVideoFilter : public IClip {
protected:
PClip child;
VideoInfo vi;
public:
GenericVideoFilter(PClip _child) : child(_child) { vi = child->GetVideoInfo(); }
PVideoFrame __stdcall GetFrame(int n, IScriptEnvironment* env) { return child->GetFrame(n, env); }
void __stdcall GetAudio(void* buf, __int64 start, __int64 count, IScriptEnvironment* env) { child->GetAudio(buf, start, count, env); }
const VideoInfo& __stdcall GetVideoInfo() { return vi; }
bool __stdcall GetParity(int n) { return child->GetParity(n); }
void __stdcall SetCacheHints(int cachehints,int frame_range) { } ; // We do not pass cache requests upwards, only to the next filter.
};

np: Proem - Live @ LSR 03-19-2006 (Merck Fragments)

[leeperry]

ok, I'm exchanging PM's with haruhiko concerning the winamp2 plugin crashing problem.

he told me he would need some open source winamp2 plugin with the problem, so he can find what's wrong.

I'm in contact with Vincent Burel, his FFX4 plugin has the problem as explained here :
http://forum.doom9.org/showpost.php?p=1096605&postcount=3111

he told me that he would help to find the problem, and I asked him if he could PM haruhiko and apparently he's willing to :)

also, haruhiko can't find how to hide the winamp2 plugin GUI, so if anyone knows ?!

I'm also in contact with Seb.26 to see if he can help.

I'm far from a coder, but I've tried to find how to hide the plugin GUI, and it might have to do with the "Plugin.MainWindow" command :
http://www.codeproject.com/KB/audio-video/winampoutput.aspx

if anyone knows of an open source winamp2 plugin that crashes ffdshow audio, please help :p

at least I'm not sitting on my ass moaning clsid :D

[Inventive Software]

If the plugin has a GUI hard-coded into it, then it's likely that it's near impossible to hide it. If Winamp can hide it, then debugging that and seeing what calls it gives out to the window would be a step closer. :)

[leeperry]

well yeah, you can hide any DSP plugin window with winamp :)

[leeperry]

ok I've talked to Vincent Burel.

he's told me the best way would be to get back to him with details on the crash, and he would do his best to help.

his email is vincent.burel____vb-audio.com (replace ____ with @)

the problem is the same with Ozone or with his FFX4 plugin as explained here :
http://forum.doom9.org/showpost.php?p=1096605&postcount=3111

as soon as you open a new track in MPC HC, ffdshow crashes.....but if you click on CONFIGURE in the fffdshow audio winamp2 section, then it doesn't crash anymore but after 3 files in a row it's using 70% of CPU time :eek:

hope something can be worked out :(

[leeperry]

OK, Vincent Burel has given me some infos about his FFX4 winamp2 plugin(that also crashes with ffda).

FFX4 is using the SDK 0.9 for winamp2.
it's a very basic SDK that only defines 4 functions :

-init : to create the plugin
-modifysamples : to process the audio
-quit : to kill the plugin

and finally "config" to open the plugin GUI.
but nothing's been done to communicate from the host, which doesn't even know the handle name of the plugin GUI(if that was the case, a SW_HIDE would be fine to hide the window)

so he believes that the only way to never have the plugin GUI showing up is to never call the "config" function.

once this function has been called, to hide the GUI again, you would have to kill the effect(QUIT) and recreate it(without using "config" of course)

[leeperry]

calling CONFIG twice in a row might also hide the GUI, considering the plugin knows its GUI is already being shown

[leeperry]

somehow I can sense Ozone will keep on crashing with ffdshow audio for a long long time :D

Seb.26 is gonna take a look on this week end, but he told me ffdshow audio is basically thousands of code lines w/o any comments :(

[EpheMeroN]

Would it be possible to implement ReClock's capabilities into the new ffdshow-tryouts?

[dimzon]

Would it be possible to implement ReClock's capabilities into the new ffdshow-tryouts?

+1

:)

[clsid]

Such functionality is not the job of a decoder. Its probably even technically impossible. So imo the answer is NO. In its current state ffdshow is already large and complex enough.

Also, instead of re-inventing the wheel, the existing wheel should just be improved.

[leeperry]

besides Reclock works perfectly fine :)

some friends of mine very much enjoy the "Frame Rate doubler" in the deinterlacer section of ffdshow.

you force 24fps with Reclock and you get "DNM-like" 48fps this way.

you run your display in 48Hz with pstrip, and you're good to go :)

it doesn't work too well on my PC though(random freezes+artifacts with scene changes), so I prefer to use only Reclock in 23.976@24 in 48Hz :cool:

[Leak]

Such functionality is not the job of a decoder. Its probably even technically impossible. So imo the answer is NO. In its current state ffdshow is already large and complex enough.
Well, it should be possible to twiddle the video timestamps to some other frame rate and resample/timestretch the audio accordingly.

But synchronizing the whole shebang with the monitor's refresh rate? Yeah, that's out of ffdshow's league and needs to be done by the video renderer.

np: Autechre - Nofour (Quaristice (Versions))

[clsid]

True, but that would require ffdshow to decode both audio and video, which is not necessarily always the case. But anyway, lets just quickly forget this topic.

[fano]

I have a feature request now if I have for example a 5.1 AAC file I must encode it to AC3 5.1 @ 640 Kbit/s to send it to my receiver via spdif... is possible to add DTS encoding @ 1.5 Mb/s?
I think we can have more quality in this way (less compression!).

Are you working in hdmi uncompressed PCM support (maybe is too early... no real HDMI soundcard for now) and new DTS and AC3 HD codec?

I think is very important to have free solution for audio HD decoding...

Thanks for the answers,
fanoI

[clsid]

I don't think an open-source DTS encoder currently exists. So the answer would be no.

[Kurtnoise]

\o/ (http://svn.mplayerhq.hu/soc/dcaenc/)...

[bwahn]

hi~ ffdshow tryout team.

I have a question about dxva.

in svn, I checkout ffdshow-tryout.rev1913 and /branches/dxva.

then, I copied dxva/src/*.* files into ffdshow-tryout.rev1913/src directory.

using ffdshow_2005.vcproj, I tried compile ffdshow.

I have some error.

....
fatal error C1083: Cannot open include file: 'dxva2api.h': No such file or directory
...

trying find dxva2api.h , but can't found it.
where can I download dxva2api.h?

or something sdk install?
I installed "Windows Server 2003 R2 Platform SDK.3790.2075" and "Microsoft DirectX SDK (April 2007)".

thanks for any help.

Eric Ahn.

[_xxl]

DXVA is not supported. Try MPC-HC for h.264 and vc-1 dxva decoders.

[fano]

\o/ (http://svn.mplayerhq.hu/soc/dcaenc/)...

So a free DTS encoder is on the way... good.
We have to wait when it is completed and integrated in ffdshow!

Maybe will arrive first the PCM multichannel conversion for usage on real HDMI output... the problem is that for now there is not HDMI soundcard :eek:

fano

[CruNcher]

is it possible to watch elementary .264 files with ffdshow? it seems ffdshow can't parse the Mpeg-2 data from Mainconcept based parser correctly (detecting its AVC flag and so redirect to the H.264 Decoder instead of the Mpeg-2) that stops the connection? :(


MainConcept MPEG Demultiplexer::AVC

Media Type 0:
--------------------------
Video: MPEG2 Video 1280x720 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B}
formattype: FORMAT_MPEG2_VIDEO {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA}


ahh i see when Mpeg-2 under Codecs is set to Libavcodec then it works (tough if someone doesn't want Mpeg-2 to be decoded with it that's no real solution) :)

and it's still not correctly playing (alot of frame skips with -b-pyramid and -bime jumps to the next i-frame to many b-frames in general seem to be the problem)


Filter : MainConcept MPEG Demultiplexer - CLSID : {79DF8A70-E7D5-49DB-ADC4-A4BE0E193AA2}

- Connected to:

CLSID: {04FE9017-F873-410E-871E-AB91661A4EF7}
Filter: ffdshow Video Decoder
Pin: In

- Connection media type:

Video: MPEG2 Video 720x576 (4:3) 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B}
formattype: FORMAT_MPEG2_VIDEO {E06D80E3-DB46-11CF-B4D1-00805F6CBBEA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 184

VIDEOINFOHEADER:
rcSource: (0,0)-(720,576)
rcTarget: (0,0)-(720,576)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 4
dwPictAspectRatioY: 3
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

MPEG2VIDEOINFO:
dwStartTimeCode: 0
cbSequenceHeader: 48
dwProfile: 0x0000004d
dwLevel: 0x0000001f
dwFlags: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 720
biHeight: 576
biPlanes: 0
biBitCount: 0
biCompression: 0
biSizeImage: 0
biXPelsPerMeter: 2868
biYPelsPerMeter: 3128
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 d0 02 00 00 40 02 00 00 ........Ð...@...
0010: 00 00 00 00 00 00 00 00 d0 02 00 00 40 02 00 00 ........Ð...@...
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: 40 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 @...............
0060: 34 0b 00 00 38 0c 00 00 00 00 00 00 00 00 00 00 4...8...........
0070: 00 00 00 00 30 00 00 00 4d 00 00 00 1f 00 00 00 ....0...M.......
0080: 00 00 00 00|00 00 00 01 67 4d 40 1f b9 04 01 68 ........gM@.¹..h
0090: 24 d0 80 00 00 03 00 80 00 00 19 71 30 00 0d 59 $Ѐ....€...q0..Y
00a0: f8 00 06 ac fd 02 8c c7 51 a2 d8 88 00 00 00 01 ø..¬ý.ŒÇQ¢Øˆ....
00b0: 68 e9 39 f2 00 00 00 00 hé9ò....

- Enumerated media type 0:

Set as the current media type


i added the {8D2D71CB-243F-45E3-B2D8-5FD7967EC09B} to ffdshow i think that makes the connection possible @ all

[Inventive Software]

.264 playback needs a parser. I'm in the process of seeing how integrating most of libavformat to play said files to a DShow filter, but I cannot for the life of me find a skeleton DShow filter to do this. :(

[clsid]

You could have a look in the MPC-HC/Guliverkli2 projects. That contains a 'BaseSplitter'. Maybe that's useful.

[SeeMoreDigital]

Donald Graft's DGAVCIndex (http://forum.doom9.org/showthread.php?t=122598) application is able to play elementary AVC streams, with libavcodec.dll. Could any of his software be incorporated/used?

[clsid]

Unless that application uses DirectShow, it's useless.

[georgevalkov]

Hello ffdshow developers!
There's a minor problem with the Logoaway filter, when an image mask is used (Mode:Shape XY). Please fix it, if You find time! :thanks:
1. The picture under the black area of the mask is shifted up by 1 line. It should be left unchanged.
2. The colours under the black area of the mask are modified. They should be left unchanged.

I tried interlaced material: 720x576@25i PAL; 704x576@25i PAL;
YUY2 and RGB24 - all produced the same problem in the output.


The mask is in its original size. The other images below are zoomed 200% or 400% (nearest neighbour in Photoshop).

[CruNcher]

There seems to be a Multithreading Slice Decoding bug in the H.264 Decoder in the current ffdshow, you can see the problem it couses here with this stream.

http://mirror05.x264.nl/CruNcher/force.php?file=./chromaprob/ateme-high.mp4

http://s6.directupload.net/images/080329/lo7lapez.png

MPCs Internal Decoder has the same problem it helps to set the threads to 1, to avoid it tough ffdshow tryout has no such option (didn't found any)

The stable build works without problems, it seems to be existing allready very long as build 1890 (i used before) also had these decoding problems.

[clsid]

You can change the number of threads in "Decoder options" section of ffdshow config.

[Shinigami-Sama]

holy shit I was wondering about that

though with MPC-HC + d3d mode I don't notice it
but with anything else I see it

now that you mention it though it only started happening when I changed to try-outs...

[GrofLuigi]

Was the change to Inno Setup 5.2.3 really necessary?

GL

[CruNcher]

You can change the number of threads in "Decoder options" section of ffdshow config.

yep missed that thx that fixes it also with the nightlies :)

[leeperry]

I've forced ffdshow audio for all supported streams, and it's doing 24/96 resampling with the Ozone winamp2 plugin enabled(32 bit float processing).

it works like a charm in MPC HC, and even in WinDVD9+

and you can force Reclock in PowerDVD......

wouldn't that be possible to force ffdshow video as well in PowerDVD/WinDVD ?

at least on XP, because ogo told me Vista had some very strong stream protections :eek:

if it works for audio, why not video ?!
maybe cheating like Reclock, stating ffdshow as the default renderer ?

you can force EVR + Reclock in PowerDVD, and this is the most stable HD player on PC atm......so all it needs is some "unsharp masking" sharpening from ffdshow to become first class :eek:

[clsid]

wouldn't that be possible to force ffdshow video as well in PowerDVD/WinDVD ?If enabling raw video doesn't work, then it is not possible, unless the player specifically adds the filter to the graph, which of course will never be done by PowerDVD.

Was the change to Inno Setup 5.2.3 really necessary?Please elaborate.

[fastplayer]

Please elaborate.
Probably complaining because innounp (http://innounp.sourceforge.net/) can't extract it.

[vlada]

I'm sorry if this has been asked before, I haven't found any post regarding this. I'm wondering if the DXVA decoder from MPC-HC could be used in ffdshow. If I want the internal DXVA decoder in MPC, then I can't use ffdshow for postprocessing anymore.

[clsid]

That question has already been answered several times. When DXVA is used, there can be no intermediate filters in between the decoder and the video renderer.

IT IS NOT POSSIBLE TO USE DXVA TOGETHER WITH FFDSHOW AS POST-PROCESSOR

[Yong]

I'm sorry if this has been asked before, I haven't found any post regarding this. I'm wondering if the DXVA decoder from MPC-HC could be used in ffdshow. If I want the internal DXVA decoder in MPC, then I can't use ffdshow for postprocessing anymore.

from http://tibrium.neuf.fr/DXVASupport.html :p
The MPC HC Video decoder must be connected directly to the renderer. That means no intermediate filters such as DirectVobSub or ffdshow can be inserted between the decoder and the video renderer.

[GrofLuigi]

Probably complaining because innounp (http://innounp.sourceforge.net/) can't extract it.
Yes, but not complaining, only asking. I would complain only if the answer is 'no'. ;)

GL

[vlada]

@clsid, Yong
Yes, I knew this. I probably wrote my question incorrectly.

I rather split it in two parts:

1) How much work would it be to port the DXVA decoder fro MPC-HC to ffdshow?

2) Would it be then possible to use ffdshow's filters (with the DXVA decoder integrated)?

So is it a general problem that a DXVA decoder needs to be connected to video renderer directly? Or is this just a limitation of MPC's decoder?

[Shinigami-Sama]

2) Would it be then possible to use ffdshow's filters (with the DXVA decoder integrated)?

So is it a general problem that a DXVA decoder needs to be connected to video renderer directly? Or is this just a limitation of MPC's decoder?

not possible with DXVA1

with DXVA2 clsid or casmir said its theoreticly possible to get it to post process

basicly DXVA1 goes source -> render -> hardware -> display
hardware/renderer might be reversed
but thats the jist of it

[CruNcher]

tough of course you can post-process on DXVA1 too but "only" with Shaders Nvidia and ATI do it all the time in their Drivers (Noise Reduction,Sharpening,Deinterlacing) :D and they do it pretty fast (software cant holdup with these speeds anymore (1080i25->50p in realtime is no problem for their Shader solutions). Compared vs the quality the difference got also allot smaller then some generations before,nvidias newest (geforce8,9) motion adaptive deinterlacing seems at least on the quality level of yadif even above more into direction of mcbob) and efficient :) (you could even ask you the question is Software Deinterlacing and Post-Pro anymore needed??? if we can use it in realtime on the GPU.

PS: Would be a sad thing if ffdshow misses this revolution of Post-Processing as it started allready (Commercialy), ofcourse if your target is editing it is a different thing couse that involves alot more work getting the stuff out of the GPU into a file ;)

[clsid]

This discussion is pointless. There is nobody that is willing to code this. Furthermore, DXVA should not be added to ffdshow with its many incompatible filters, instead (theoretically) compatible post-processing functions should be added to the standalone filter from MPC.

[leeperry]

you can use any number of PS scripts in MPC HC.

deinterlacing/sharpen/etc.....

[Inventive Software]

This discussion is pointless. There is nobody that is willing to code this. Furthermore, DXVA should not be added to ffdshow with its many incompatible filters, instead (theoretically) compatible post-processing functions should be added to the standalone filter from MPC.

Wrong! I'm willing to code it, but my knowledge of C++ is limited, and I don't have much time these days, so summer at the absolute earliest! ;)

[clsid]

I should have said "is going to" instead of "is willing to" :)

[RadScorpion]

I have a small question regarding the ffdshow interface. I'm trying to make graphstudio handle ffdshow and I'd like to be able to set e.g. resizing filter or deinterlacing via external XML file.... so far I've been unable to pass these parameters via the ffdshow interface :(. Does anyone have a clue ?

[clsid]

Send a PM to Albain. He has created most of the API for controlling ffdshow from external apps.

[iron2000]

Is it possible to implement auto colorspace selection in ffdshow?

Or does Haali Renderer overwrites whatever comes before it?

[leeperry]

HR only deals with YUY2 colorspaces.
for RGB32 it won't change a damn thing, you can make profiles in ffdshow like I do.

RGB32/BT601 for upscaled SD, and YUY2 for HD.

I'm actually the one who asked Haali to offer auto colorspace in HR ;)

[leeperry]

hey guys, my stupid question of the day :

I see you can change the order of the filters in ffdshow, but is it only GUI related......or does it actually change the processing order ?! :eek:

Thanks,

[fastplayer]

The filter order, of course.

[leeperry]

kthx, makes sense :D

[leeperry]

ok, concerning the winamp2 plugin crashes.......haruhiko_yamagata has exchanged emails with Vincent Burel....but still can't find what's wrong.

Vincent has kindly offered to run a debug session with ffdshow and his FFX4 plugin, to see what's up ?!

he asked me for a non-protected build....I can tell him to use the latest clsid/xxl ?

ffdshow crashes like a 100 times/day on my PC.....that'd be really great to see this bug fixed :(

TIA,

[leeperry]

ok Vincent Burel told me that it doesn't crash on win2k.
but it does crash on XP for him as well.

and when I open a new file on XP, it's like it was gonna work.....then it crashes, like if there was some missing component ?!

and sometimes MPC HC says "failed to query the needed interfaces for playback"

any chance someone finding out what's missing in ffdshow audio ?

please ? please ? please ? http://pix.nofrag.com/8/6/d/8b5de77fde2c5d3219ee4e549f77e.jpg

[leeperry]

ok Vincent Burel got back to me with some debug infos :

with his FFX4 plugin, available here :
http://vincent.burel.free.fr/download/ffx4WinAmp_FullDemo.zip

when you skip to a new song in MPC HC, his DLL is released before the destruction function was over.
the closing procedure of the plugin keeps on doing its work, BUT the DLL is not available on the system anymore.

so the FreeLibrary call is done too early, you need to wait for the QUIT call to be finished before closing the library.

sometimes it works(eg when you click on CONFIGURE in the ffdshow winamp plugin options), because the library is overwritten before the QUIT call was made, and that seems to work fine....but after 3 files in a row it's using 70% of CPU time.

any chance for a fix please ? :thanks:

[JarrettH]

new beta soon? eh eh :p

[leeperry]

ok so Seb.26(another ffdshow occasional coder) pointed me to this url :
http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/audioFilters/winamp2/Twinamp2.cpp?view=markup

I've sent it to Vincent Burel.

here's what he said :

all I can say is that at some point they have a builder that creates a DLL object :

Twinamp2dspDll::Twinamp2dspDll
dll=new Tdll(flnm.c_str(),NULL);

and in another destroyer, they call it in the end :

Twinamp2dsp::~Twinamp2dsp()
{
mod->Quit(mod);
}

now they need to check that the DLL object(shown above) is destroyed AFTER the Quit(mod) , and not BEFORE.

it's translated from french to english, so bear with me :D

:thanks:

[leeperry]

BTW, I also have a problem with 1440* videos

they are encoded in 1440 and need to be expanded to 1920

like this one :
http://download.microsoft.com/download/c/e/a/ceaf0def-8889-4462-9616-7ce4246569ae/Taxi_3_1080.exe

it says

Input Size : 1440*816
Input SAR : 4/3, DAR : 40/17

in the resize options, it's set to 2.35:1 AR

yet it's outputting 1.78 :(

I've tried with MPC HC and KMPlayer.....they both show 1.78 content

I have to select 2.35 AR in KMP so it's shrunk

happens with WMV-HD and MKV/x264 videos.

and also on that video(and all 6 channels WMA), ffdshow doesn't output the channels properly.....it says 24 bit 6 channels input/output.....yet I've got a pair of headphones on the PC and I'm hearing the rear speakers in the headphones :(

:thanks:

[SeeMoreDigital]

Hi Lee,

Just to let you know, the 1440x816 WMV sample is displayed at the correct aspect ratio for me.

I'm using MediaPlayer Classic with FFDshows "libavcodec" decoder.


Cheers

[leeperry]

so it's 2.35 ?!
I also run MPC HC with ffdshow, and it's 1.78 :(

[haruhiko_yamagata]

BTW, I also have a problem with 1440* videos
I can't reproduce. Could you paste a screen shot of the resize dialog?

[leeperry]

Hi haruhiko,

http://pix.nofrag.com/e/3/9/241c036639de8ee20cb6f5a66789d.png

I resize to 1280*768, but that shouldn't mess with the AR....if I disable the RESIZE filter, ffdshow still outputs 1.78

and I've tried to toggle "set aspect ratio in output media type", it doesn't change anything.

and the WM audio is not processed correctly in ffdshow audio, I get the rear channels in the headphones ?!

but that's no big deal :D

I hardly ever watch WMV HD movies....

did the infos that Vincent Burel gave me would finally help you to fix the Ozone winamp plugin bug ?

tell me so.....please :D

Thanks for all your hard work,

[haruhiko_yamagata]

@leeperry
I still can't reproduce.
Are you using overlaymixer?

As for winamp plugin, I said I didn't have any good news. Please don't repeat.

[leeperry]

well, I've just given you infos on how to fix the problem I think ?!

http://forum.doom9.org/showthread.php?p=1122857#post1122857

http://forum.doom9.org/showthread.php?p=1122966#post1122966

You've told me yesterday by PM that you were willing to fix it if I was giving you clues ?!

isn't that clues ?

Vincent Burel checked your code through debug, isn'it it possible that you check that it's done in the right order please ?

and now you talk to me like I was a spammer :eek:

if you don't care and think I'm harassing you, please allow me to apologize reporting bugs.....I'm sincerely sorry to waste your time repeating myself like a scratched LP and I shall stop.

[Inventive Software]

@leeperry: Check the Aspect Ratio settings in MPCHC.

[SeeMoreDigital]

Personally I don't use any of FFdshow's "Resize and Aspect" options: -

http://i29.tinypic.com/2s6352c.png


Cheers

[leeperry]

even If I don't resize in ffdshow, I get 1.78 with YUY2 and EVR/VMR9/Overlay in KMPlayer/MPC HC

I've tried to play around with all the AR options in MPC HC, to no available.

anyhow, I've made a 1.78>2.35 pan&scan preset in MPC HC, and it works perfectly :D

[Inventive Software]

What about when you use YV12 video?

[haruhiko_yamagata]

@leeperry
Sorry, I overlooked your #3428. I'll look into it.

[leeperry]

ok cool, thank you :)

also #3430 please :D

What about when you use YV12 video?
same old, same old.

in VMR7, in YV12, in RGB........but that's OK :p

[Inventive Software]

OK, a trial and error session if you're willing.... :D

What about just the resize filter enabled?

[leeperry]

OK I'll try :D

BTW, I got one question if you don't mind.

I understand x264 is always in YV12, for WMVHD I'm not sure ?

but then I resize to 1280*768 in spline/spline in ffdshow, then apply unsharp mask sharpening.

do I have any reason to output in YUY2 over YV12 ?

thanks,

[Inventive Software]

Only if the renderer won't take YV12 video. For the most part, WMV-HD is YV12, like x264. The only way you'd know whether WMV-HD is YV12 or YUY2 is check the "Info" panel during playback. Don't apply any filters to it, and see what colour-space is being used. Check the output mode too, and make sure no conversion is being used. :)

[leeperry]

ok but even if the input is YV12, don't the resize and sharpen filters work in YUY2 ?

then this would yield better results in YUY2 over YV12 ?
like oversampling in 64bit for audio ?!

I've tried to enable/disable the sharpen and resize filters.........still 1.78 :D

[Inventive Software]

Right, please double check your splitter and make sure that ffdshow is reporting the correct SAR. If it is, and it's still not correctly displaying it, try a different media player. I'd play the sample, but I am limited to 720p by my display (1280x800).

Regarding YUY2 vs YV12, there's a quality loss between converting to/from colourspaces, regardless of whether your supposed upsampling from YV12 to YUY2 works correctly. The filters may well work ONLY with YUY2, and if that's the case, then I dunno what to suggest. :D

[leeperry]

oh right.

anyone knows if the resize and unsharp mask sharpening filters work in YV12 then ? :D

well I'm also limited to 1280*768, yet it displays as 1.78 even though ffdshow knows its AR is 2.35 :eek:

OK I'll run more tests ;)

[Inventive Software]

Right, very quick and very dirty lesson in AR......

You're resizing a 1440x816 video to the same AR, but 1280 width, so that it displays correctly. That's 1280x544 as your destination. The SAR STAYS as 4:3, because the frame size is still the same dimension multiple. Try it yourself. 1440/816 = 960/544. The SAR is reported as 1.77 because it's the closest to 1440/816. ffdshow isn't reporting the SAR to the renderer, hence why it appears to be 1.77.

Try this: resize JUST the horizontal to 1280, check "Keep Aspect Ratio" down the bottom of the box, and then see. ;)

BTW: 1280/768 = 1.666666666, AKA 5:3. :)

[SeeMoreDigital]

Lee,

I think your issue may be related to MediaPlayer Classic-HD.... Can you try the "non HD" version of MPC and report back?


Cheers

[leeperry]

well resizing only x to 1280 still outputs 1.78

I know I'm resizing to 1.67, yet I've never ever had this kind of problem before.

I would guess that the WMV splitter does the job right, considering ffdshow show the AR as 2.35

I've just tried with mpc2kxp6490, and it's............the same :D

so what's better then ?

YV12 resize and sharpening because there's no colorspace conversion

OR

YUY2 resize and sharpening because it's 16 bit instead of 12 ?

We usually process audio in 32/64 bit to get better overall results :D

[Inventive Software]

Right, I'm downloading this WMV-HD video, to try and figure out what the hell's going on. :D

Regarding YV12 vs YUY2, YV12 is YUV 4:2:0, YUY2 is 4:2:2. There's no difference between the output, unless your filters only work in YUY2.

Can you use MPC 6.4.9.1 from now on in this scenario, as that's what I'm running, and that fixes a few bugs and that's it. MPC-HC is too bleeding-edge to be considered reliable, because most new features in that end up breaking something, and I do read that thread occasionally, so I know what I'm staying away from.

One last thing. In the "Output" section of ffdshow's config, can you confirm the following settings...

- Planar YUV (mine's all enabled);
- Packed YUV (mine's all enabled);
- RGB (mine's all enabled, except YV12 to RGB) (if enabled, please disable YV12 to RGB conversion);
- Select Closest matching colorspace (mine's enabled);
- Set pixel aspect ratio in output media type (mine's enabled);
- Allow output format changes during playback - half-filled, not ticked.

Last thing, what's your ffdshow revision? I use 1856, and haven't updated it since.

[leeperry]

well, even if the source is 4:2:0, it's better to process in 4:2:2......ain't it ?
12 Vs 16 bit :eek:

and the "high quality YV12 to RGB conversion" does exactly that........YV12 > YUY2 > RGB32(haruhiko confirmed it) :D

I think I'll stick to YUY2 output to get upsampled chroma :D

but I'm not sure if resize and sharpening are processed depending on the input or the output colorspace ? :eek:

if it's done in YV12 like the source, then converted to YUY2 in the end.........it's pointless :(

if so, then I should use CoreAVC in YUY2........to make sure that ffdshow works in YUY2 as well :D


I use the latest build from XXL, everything's unchecked in the output section......except for YUY2

I've tried to enable the other options, didn't change a damn thing :D

[Inventive Software]

Well the easy way to determine whether your sharpen and resize work in YV12 or not is to do exactly that.... make it output YV12 ONLY! ;)

And the 4:2:0 to 4:2:2 then filtering, then converting back to 4:2:0 is a waste of CPU time because there's 2 (4?) bits not being used. Keep everything in the chain YV12 to be sure it's doing its job.

It's not actually up-sampling, because it's creating new data based on nothing! ;)

And I tried downloading that source that you have problems with, but my internet's so slow at my halls that I'll wait until uni tomorrow. So I can't do anything until then. Sorry.

[leeperry]

ohhhhhhhhhhh, so even with YUY2 output from CoreAVC, the extra chroma bits are not used at all ? :eek:

well video processing such as resize/sharpening would put those bits to use, or am I missing something ?

and I don't even know if they are processed in the colorspace of the input or of the output ?!

like sampling a vynil audio source in 24/96 and doing 64/192 post-processing on top of it ?!

[Inventive Software]

The only way CoreAVC would output usable YUY2 video is if the H.264 video was encoded as YUY2. Most of the time it's YV12. YUY2 is useful for connecting to a couple of renderers and filters that can't accept YV12 video, hence that option.

If you're recording at 24/96 and doing 64/192 post-processing, I'd ask why you're not recording at 64/192 in the first place. The extra precision may help if there's interpolation in the post-processing, but it's the same deal when encoding from that. Unless you're doing the final encode at 64/192 you lose what was calculated, which in audio is sometimes actually useable. So do the whole chain in 24/192, and don't convert inbetween.

If you have some time on your hands, do a test with ffdshow's video filters, and enable JUST YV12, YUY2 and RGB output, conversion enabled. If the filters work in all 3, you're fine and dandy. If you can see that it's better, great, all the better for you. If they don't work in YUY2 or RGB or YV12, please state which ones so that others don't make that mistake.

:)

[leeperry]

well the whole point of doing 64/192 post processing on 24/96 material is to get more precise roundings.

even if in the end you downsample to 24/96 again with noise shaping, you still end up with better roundings.......and the more filters you use, the better it gets :D

doing YUY2 resize/sharpening on YV12 input follows the same idea.........more usable bits....which then are converted to RGB32 by the videocard mixer, and then to RGB24 to the DVI output.

I just dunno if the ffdshow filters are processed in the input or the output colorspace ?!

I'm colorblind so I can't "see" changes.........but the smoother the roundings, the smoother the color fades.....and the reward would be less banding in the end :D

[SeeMoreDigital]

Lee,

Please try the "non HD" version of MPC and report your findings?


Cheers

[leeperry]

I did :D

I've just tried with mpc2kxp6490, and it's............the same :D

[SeeMoreDigital]

I did :DIn that case...

I think some of your MPC/FFdshow settings must be in-conflict with each other :eek:

The WMV sample is displayed at the correct aspect ratio for me.... ie: MAR=2.352941176:1 (or 40:17), ARS=4:3

[leeperry]

well, If I disable ffdshow I get correct AR.

what do u have enabled in ffd ?

[haruhiko_yamagata]

Is "Allow output format changes during playback" checked?

[haruhiko_yamagata]

ok Vincent Burel got back to me with some debug infos :

with his FFX4 plugin, available here :
http://vincent.burel.free.fr/download/ffx4WinAmp_FullDemo.zip

when you skip to a new song in MPC HC, his DLL is released before the destruction function was over.
the closing procedure of the plugin keeps on doing its work, BUT the DLL is not available on the system anymore.

so the FreeLibrary call is done too early, you need to wait for the QUIT call to be finished before closing the library.

sometimes it works(eg when you click on CONFIGURE in the ffdshow winamp plugin options), because the library is overwritten before the QUIT call was made, and that seems to work fine....but after 3 files in a row it's using 70% of CPU time.

any chance for a fix please ? :thanks:The dll is unloaded after QUIT call.
But finally I found that QUIT had to be called before the Window class was destroyed.
Rev 1928 should improve stability of Winamp plug-in feature, yet there are some other problems.

[leeperry]

OMG OMG OMG, finally a fix for Ozone ?! :D

can you play several files in a row w/o ffdshow audio crashing ? :eek:

thank you haruhiko!

can someone PLEASE compile rev1928 ? :D

Is "Allow output format changes during playback" checked?

just tried, doesn't change anything.

I'll try to reset ffdshow to defaults :devil:

[Inventive Software]

(Changed a few things)
One last thing. In the "Output" section of ffdshow's config, can you enable the following settings...

- All Planar YUV;
- All Packed YUV;
- All RGB except YV12 to RGB;
- Enable "Select Closest matching colorspace";
- Enable "Set pixel aspect ratio in output media type";
- Enable "Allow output format changes during playback", but make it not ticked, just filled.


Please either make a profile in ffdshow like that or make your settings like that currently, and then play it again.

And methinks you might have to wait a while (hours, possibly) to get your 1928 build. ;)

[leeperry]

well Seb.26(an occasional ffdshow coder) will build it for me in a few hours if noone else does in the meantime:D

ok you nailed it!

I need to have "Set pixel aspect ratio in output media type" and "Allow output format changes during playback" filled(not ticked)........then it's OAR :)

only one of the two options and it's 1.78.

should I fill these 2 options for all my profiles(SD,720p,1080p) ?

Thanks!

and BTW concerning my YV12 upsampled to YUY2, Seb.26 has told me that the sharpen filter of ffdshow works in YV12 :eek:

[Inventive Software]

If that ends up not working, and you're ready and willing to sort this, get GraphEdit and I'll give you my MSN addy via PMs. I got things to sort in the mean time, so gimme a couple hours. ;)

[leeperry]

I've edited my post actually :D

[haruhiko_yamagata]

OMG OMG OMG, finally a fix for Ozone ?! :D

can you play several files in a row w/o ffdshow audio crashing ? :eek:Yes, but not hundred. It doesn't crash anymore, but there are occasional freezes.

[SeeMoreDigital]

well, If I disable ffdshow I get correct AR.

what do u have enabled in ffd ?I have "libavcodec" selected.

[Inventive Software]

well Seb.26(an occasional ffdshow coder) will build it for me in a few hours if noone else does in the meantime:D

ok you nailed it!

I need to have "Set pixel aspect ratio in output media type" and "Allow output format changes during playback" filled(not ticked)........then it's OAR :)

only one of the two options and it's 1.78.

should I fill these 2 options for all my profiles(SD,720p,1080p) ?

Thanks!

and BTW concerning my YV12 upsampled to YUY2, Seb.26 has told me that the sharpen filter of ffdshow works in YV12 :eek:

About bloody time! Keep it that way! :D

[leeperry]

ok great, Seb.26 just gave me that new rev http://forum-images.hardware.fr/images/perso/mamy_furax.gif :
http://membres.lycos.fr/sebfr26/PCHC/.public/ffdshow.ax

so good news, now you can open two files in a row w/ Ozone.......but it start freezing at the third, and the more you open the more it freezes and the fifth refuses to open :(

like there was a deadlock ?! :eek:

it seems that the DX wrapper works fine as you said :)

but more importantly, you can take off the dsp wrapper from the blacklist coz it works perfectly fine now.......it even hides its GUI :)

at this point it's my best bet to find some similar plugin to Ozone in VST : http://www.savioursofsoul.de/Christian/Programs/WinAmp_VST_Bridge.exe

perfect VST support in ffdshow, AWESOME :eek:

thanks for all!

[leeperry]

to get back on the colorspace output differences with yv12 x264 input(with resize+sharpen)

the test video is available here :
http://www.megaupload.com/fr/?d=UE3O6G2J

yuy2:
http://www.freepicshost.net/pview.php?fid=9012&fname=yuy2.png

yv12:
http://www.freepicshost.net/pview.php?fid=9013&fname=yv12.png

rgb32:
http://www.freepicshost.net/pview.php?fid=9010&fname=rgb32.png

rgb32hq:
http://www.freepicshost.net/pview.php?fid=9011&fname=rgb32hq.png

YUY2 looks smoother(extra chroma bits?) than YV12, and same goes for RGB32 over RGB32HQ :eek:

[Px]

Thumbs images unavailable
http://www.freepicshost.net/public/thumb_47fd4dc8542c9353503495.png
Forbidden
You don't have permission to access /public/thumb_47fd4dc8542c9353503495.png on this server.

Additionally, a 404 Not Found error was encountered while trying to use an ErrorDocument to handle the request.

[leeperry]

fixed, click the links :D

ok this VST plugin seems to sound better than Ozone :

http://static.kvraudio.com/i/b/spectralive2.jpg

it works fine with ffdshow and the VST wrapper, but I've just opened like I dunno.........200 files in 3/4H and now it's freezing for each new file even if MPC HC was closed ?

I don't plan on rebooting my PC everyday :(

[foxyshadis]

Check that mpc is gone and not just closed. Sometimes something refuses to unload even though the GUI is gone, and it turns out there's a few copies of MPC still open in task manager.

[Shinigami-Sama]

I agree with foxy; I've had that problem a few times - also my enter key is no longer functioning here on doom9...

[leeperry]

Check that mpc is gone and not just closed. Sometimes something refuses to unload even though the GUI is gone, and it turns out there's a few copies of MPC still open in task manager.

yes I did, thanks for the tip!

I've removed the dsp stacker plugin, even though it wasn't in use, it was freezing to death ?! :eek:

ok nothing sounds as good as OZone :(

any chance to add a delay when you QUIT to avoid deadlock ?! :(

BTW, look at Amelie Poulain in EVR + unsharp mask 33 :

http://pix.nofrag.com/4/2/3/2ce84a7dbbfc9bab48e19a16b91d7tt.jpg (http://pix.nofrag.com/4/2/3/2ce84a7dbbfc9bab48e19a16b91d7.html) http://pix.nofrag.com/7/9/2/3fd6ab8948752efeb2a8e5d1b46aatt.jpg (http://pix.nofrag.com/7/9/2/3fd6ab8948752efeb2a8e5d1b46aa.html)

[jmartinr]

I would like the crop filter to respect PAR.

I always use "resize and aspect" to get video perfect on my TV.

I use "specify horizontal and vertical size", "keep original aspect ratio", "resize always" and "process pixel aspect ratio internally". Works great. But if I crop before, ratio isn't correct.

Cropping via the Avisynth filter does work. But I have an old computer attached to my TV, so if I could use the crop filter straightaway, speed would be helped and it would be more handy too.

Only thing needed I think/hope is that crop doesn't change PAR.

[clsid]

You could move crop below resize.

[cyberbeing]

I don't know if this is a ffdshow bug or a flv splitter bug but I'm suspecting there is a similar bug in both.

Sample1: http://www.mediafire.com/?cd2tyy2w9ew 640x360 30.000fps (reported by MediaInfo)
Sample2: http://www.mediafire.com/?bvgsdywr5gh 720x405 29.970fps (reported by MediaInfo)

FFDShow (latest clsid build) reports 25.000fps (renderer uses 30.000/29.970fps anyways) and outputs 640x368/720x416 even when using the Nero splitter (which reports 640x360/720x405) which results in seeing a 8/11 pixel blurry bar on the bottom of the video.

MPC-HC FLV decoder outputs 640x368/720x416 25.000fps (renderer uses 30.000/29.970fps anyways) even when using Nero Splitter.

FLVsplitter 1.0.0.4 reports 640x368/720x416

Nero FLV Splitter reports 640x360/720x405

Mplayer reports and outputs 640x360/720x405 30.000/29.970fps

Adobe Media Player outputs 640x360/720x405

Streaming players output 640x360/720x405

Any ideas? (16:9 MOD16 bug?)

[Placio74]

I don't know if this is a ffdshow bug or a flv splitter bug but I'm suspecting there is a similar bug in both.
...
Hmm...

But GSpot, VirtualDub with FLV input plugin (and ffdshow VfW as decoder) and Avidemux show 640x368/720x416.

After used FLV Extract - AVI files have 640x368/720x416 (MediaInfo show too).
With blurry bar on the bottom of the video when play or edit.
Same when change container used Avidemux.

After change container to AVI (used FLV Extract or Avidemux) and FourCC to VP62 - when play or edit with original VP6 codec - still blurry bar on the bottom of the video (exactly not bottom, but top - because image video is flip) and of course 640x368/720x416.

Used FLVMDI i'm see 640x360/720x405 in metadata info.

When change container to AVI used FFmpeg... most tools show 640x360/720x405.
VirtualDub with FLV input plugin and ffdshow VfW as decoder show 640x360/720x405.
When play (system AVI splitter and ffdshow) i'm see 640x368/720x416 and GSpot show too - but when change FourCC and use VP6 codec it's... 640x360/720x405. Except... when use ffdshow instead VP6 codec... it's again 640x368/720x416.

Odd...:confused:

[jmartinr]

You could move crop below resize.
Yes I could, but I want to end at a fixed size. Using Nvidia tv out that's wanted (if you want to know why see: http://www.jillesvangurp.com/2006/11/06/nvidia-tv-out-aspect-ratio-trouble-workaround/).

[clsid]

The FLV problem has nothing to do with ffdshow. The video decoder must crop off the blurry part, however afaik there is no way that the FLV splitter can signal the correct resolution to the video decoder.

[Mercury_22]

It's possible to add E-AC3 support, sice there is already a FFmpeg patch ? Look here http://forum.doom9.org/showthread.php?t=129050
Please ! :thanks::helpful:

(Even just experimental :stupid:)

[Inventive Software]

:search: It's not happening any time soon. Until it makes it into ffmpeg's SVN trunk, it won't make it into our trunk. ;)

[cyberbeing]

The FLV problem has nothing to do with ffdshow. The video decoder must crop off the blurry part, however afaik there is no way that the FLV splitter can signal the correct resolution to the video decoder.
That confuses me a little. So why are the extra blurry pixels (which I'm assuming don't even exist) being created on the bottom when FFDshow is used? What about the wrong reported framerate of 25fps? FFDshow is being passed 640x360 by the Nero FLV Splitter, FFDshow's video decoder info & cpu shows 640x360, yet it outputs 640x368 to the renderer. I was assuming there was some (GPL?) code being shared between multiple decoders/splitters that was wrongly forcing mod16 on flash video content.

FFDShow output pin (RGB & YV12 looks almost identical to YUY2 below)
- Connection media type:

Video: YUY2 640x368 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_YUY2 {32595559-0000-0010-8000-00AA00389B71}
formattype: FORMAT_VideoInfo2 {F72A76A0-EB0A-11D0-ACE4-0000C0CC16BA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 471040
cbFormat: 112

VIDEOINFOHEADER:
rcSource: (0,0)-(0,0)
rcTarget: (0,0)-(640,368)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 40
dwPictAspectRatioY: 23
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 640
biHeight: -368
biPlanes: 1
biBitCount: 16
biCompression: YUY2
biSizeImage: 471040
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 80 02 00 00 70 01 00 00 ........€...p...
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 28 00 00 00 17 00 00 00 ........(.......
0040: 00 00 00 00 00 00 00 00 28 00 00 00 80 02 00 00 ........(...€...
0050: 90 fe ff ff 01 00 10 00 59 55 59 32 00 30 07 00 �þÿÿ....YUY2.0..
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................

- Enumerated media type 0:

Video: YUY2 640x360 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_YUY2 {32595559-0000-0010-8000-00AA00389B71}
formattype: FORMAT_VideoInfo2 {F72A76A0-EB0A-11D0-ACE4-0000C0CC16BA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 460800
cbFormat: 112

VIDEOINFOHEADER:
rcSource: (0,0)-(640,360)
rcTarget: (0,0)-(640,360)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 16
dwPictAspectRatioY: 9
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 640
biHeight: 360
biPlanes: 1
biBitCount: 16
biCompression: YUY2
biSizeImage: 460800
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0010: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
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 10 00 00 00 09 00 00 00 ................
0040: 00 00 00 00 00 00 00 00 28 00 00 00 80 02 00 00 ........(...€...
0050: 68 01 00 00 01 00 10 00 59 55 59 32 00 08 07 00 h.......YUY2....
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................

- Enumerated media type 1:

Video: YUY2 640x360 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_YUY2 {32595559-0000-0010-8000-00AA00389B71}
formattype: FORMAT_VideoInfo {05589F80-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 460800
cbFormat: 88

VIDEOINFOHEADER:
rcSource: (0,0)-(640,360)
rcTarget: (0,0)-(640,360)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

BITMAPINFOHEADER:
biSize: 40
biWidth: 640
biHeight: 360
biPlanes: 1
biBitCount: 16
biCompression: YUY2
biSizeImage: 460800
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0010: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0020: 00 00 00 00 00 00 00 00 80 1a 06 00 00 00 00 00 ........€.......
0030: 28 00 00 00 80 02 00 00 68 01 00 00 01 00 10 00 (...€...h.......
0040: 59 55 59 32 00 08 07 00 00 00 00 00 00 00 00 00 YUY2............
0050: 00 00 00 00 00 00 00 00 ........

- Enumerated media type 2:

Video: 640x360 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_RGB32 {E436EB7E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_VideoInfo2 {F72A76A0-EB0A-11D0-ACE4-0000C0CC16BA}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 921600
cbFormat: 112

VIDEOINFOHEADER:
rcSource: (0,0)-(640,360)
rcTarget: (0,0)-(640,360)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

VIDEOINFOHEADER2:
dwInterlaceFlags: 0x00000000
dwCopyProtectFlags: 0x00000000
dwPictAspectRatioX: 16
dwPictAspectRatioY: 9
dwControlFlags: 0x00000000
dwReserved2: 0x00000000

BITMAPINFOHEADER:
biSize: 40
biWidth: 640
biHeight: 360
biPlanes: 1
biBitCount: 32
biCompression: 0
biSizeImage: 921600
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0010: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
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 10 00 00 00 09 00 00 00 ................
0040: 00 00 00 00 00 00 00 00 28 00 00 00 80 02 00 00 ........(...€...
0050: 68 01 00 00 01 00 20 00 00 00 00 00 00 10 0e 00 h..... .........
0060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................

- Enumerated media type 3:

Video: 640x360 25.00fps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Video {73646976-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_RGB32 {E436EB7E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_VideoInfo {05589F80-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 921600
cbFormat: 88

VIDEOINFOHEADER:
rcSource: (0,0)-(640,360)
rcTarget: (0,0)-(640,360)
dwBitRate: 0
dwBitErrorRate: 0
AvgTimePerFrame: 400000

BITMAPINFOHEADER:
biSize: 40
biWidth: 640
biHeight: 360
biPlanes: 1
biBitCount: 32
biCompression: 0
biSizeImage: 921600
biXPelsPerMeter: 0
biYPelsPerMeter: 0
biClrUsed: 0
biClrImportant: 0

pbFormat:
0000: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0010: 00 00 00 00 00 00 00 00 80 02 00 00 68 01 00 00 ........€...h...
0020: 00 00 00 00 00 00 00 00 80 1a 06 00 00 00 00 00 ........€.......
0030: 28 00 00 00 80 02 00 00 68 01 00 00 01 00 20 00 (...€...h..... .
0040: 00 00 00 00 00 10 0e 00 00 00 00 00 00 00 00 00 ................
0050: 00 00 00 00 00 00 00 00 ........

[akupenguin]

After change container to AVI (used FLV Extract or Avidemux) and FourCC to VP62 - when play or edit with original VP6 codec - still blurry bar on the bottom of the video (exactly not bottom, but top - because image video is flip) and of course 640x368/720x416.
It just so happens that I have written a VP6 encoder and had to deal with exactly that issue.
VP62 is not identical to VP6F. One of the differences is that VP6F has an extra header field specifying the exact video resolution for the decoder to crop it to, while VP62 doesn't and can only derive resolution from number of macroblocks. Either your extractor removed that header, or the decoder ignored it since it's not supposed to exist in the new format.
FFDshow is being passed 640x360 by the Nero FLV Splitter, FFDshow's video decoder info & cpu shows 640x360, yet it outputs 640x368 to the renderer.Stuff like that can happen if the container disagrees with the video stream itself about the resolution.
Some players might also choose to crop the decoded video to the container's resolution (and might or might not crop the correct edge).

[jmartinr]

You could move crop below resize.
Though not my preferred solution I tried it.

I put Crop after Resize & Aspect and that doesn't do anything. :-(

[still using: "specify horizontal and vertical size", "keep original aspect ratio", "resize always" and "process pixel aspect ratio internally"]

[clsid]

I have uploaded two different builds of libavcodec (compiled with different GCC settings) here:
http://www.zshare.net/download/105865011265b406/

Everyone feel free to run some benchmarks and let me know if there is any noticeable performance difference between the two builds.

[wyrd]

I've tried with TheGreatestGame(H.264) (http://tirnanog.fate.jp/tmp/mp4_test/TheGreatestGame_HD_AVC.mp4).
(on C2D(E6600),filter not use,skip deblocking always)
result(timecodec)
(http://tirnanog.fate.jp/tmp/comparison/ffdshow_comparison_clsid_20080415_test.txt)Regards

[Px]

C2D 6300@2.8 GHz, renderer - null

H264_24Mbps_final.mkv, H.264, 1920x1080, 24 Mbps
dll_1
User: 2s, kernel: 0s, total: 2s, real: 32s, fps: 459.3, dfps: 35.7
User: 1s, kernel: 0s, total: 2s, real: 32s, fps: 534.2, dfps: 35.8
dll_2
User: 2s, kernel: 0s, total: 2s, real: 31s, fps: 465.0, dfps: 37.2
User: 1s, kernel: 0s, total: 2s, real: 31s, fps: 570.7, dfps: 36.9

Doom.720p.avi, DivX, 1280x720, ~4 Mbps
dll_1
User: 12s, kernel: 0s, total: 12s, real: 11s, fps: 145.1, dfps: 163.4
User: 12s, kernel: 0s, total: 12s, real: 11s, fps: 146.8, dfps: 164.3
dll_2
User: 11s, kernel: 0s, total: 11s, real: 10s, fps: 152.0, dfps: 164.7
User: 12s, kernel: 0s, total: 12s, real: 10s, fps: 146.0, dfps: 165.9

cornell_m1080p.mov, H.264, 1920x1080, ~10 Mbps
dll_1
User: 70s, kernel: 0s, total: 70s, real: 63s, fps: 50.8, dfps: 56.5
User: 68s, kernel: 0s, total: 69s, real: 63s, fps: 52.1, dfps: 56.5
dll_2
User: 60s, kernel: 0s, total: 60s, real: 62s, fps: 59.0, dfps: 57.3
User: 66s, kernel: 0s, total: 66s, real: 61s, fps: 54.4, dfps: 58.1

[cyberbeing]

AMD X2 4800+ 2.4Ghz
1440x1080 x264 16:9 anamorphic ~6Mbps

rev1925
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 634.7, dfps: 52.6
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 617.5, dfps: 52.0
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 519.3, dfps: 52.5
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 557.3, dfps: 53.0

libavcodec_1
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 531.3, dfps: 52.4
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 431.1, dfps: 52.6
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 439.4, dfps: 53.0
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 714.0, dfps: 52.8

libavcodec_2
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 439.4, dfps: 51.9
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 496.7, dfps: 52.9
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 634.7, dfps: 52.6
User: 0s, kernel: 0s, total: 0s, real: 6s, fps: 519.3, dfps: 51.8

[wyrd]

with rev1943 (http://tirnanog.fate.jp/tmp/comparison/ffdshow_comparison_clsid_rev1943_test.txt)

[Yong]

Vista x86, amd athlon 6000+, yv12 output colorspace, null renderer.

mpeg1 video, 800x600, 24fps, 8mbps
libmpeg2, msvc 2008 express
User: 10s, kernel: 0s, total: 10s, real: 10s, fps: 315.1, dfps: 306.2
User: 10s, kernel: 0s, total: 10s, real: 10s, fps: 316.0, dfps: 307.6
User: 10s, kernel: 0s, total: 10s, real: 10s, fps: 314.6, dfps: 305.3
libavcodec, gcc4.2.1 r1938
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2278.1, dfps: 292.0
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2410.5, dfps: 284.8
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2094.0, dfps: 292.0
libavcodec_1.dll
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2094.0, dfps: 289.1
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 1974.3, dfps: 287.9
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2205.4, dfps: 290.4
libavcodec_2.dll
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2528.1, dfps: 289.5
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 1901.9, dfps: 286.7
User: 1s, kernel: 0s, total: 1s, real: 11s, fps: 2229.1, dfps: 287.1

XviD, 704x396 29.97fps 3mbps
xvidcore.dll, gcc4.2.1
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 498.8, dfps: 486.7
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 495.7, dfps: 486.7
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 495.7, dfps: 486.8
xvidcore.dll, msvc2008 express
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 451.4, dfps: 445.2
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 459.1, dfps: 452.7
User: 5s, kernel: 0s, total: 5s, real: 5s, fps: 453.9, dfps: 445.2
libavcodec, gcc4.2.1 r1938
User: 3s, kernel: 0s, total: 3s, real: 4s, fps: 638.7, dfps: 624.0
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 651.6, dfps: 633.7
User: 3s, kernel: 0s, total: 3s, real: 4s, fps: 646.4, dfps: 626.3
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 646.4, dfps: 633.7
libavcodec_1.dll
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 651.6, dfps: 638.7
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 656.9, dfps: 636.3
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 665.0, dfps: 641.3
libavcodec_2.dll
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 659.6, dfps: 633.7
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 654.3, dfps: 641.3
User: 3s, kernel: 0s, total: 3s, real: 3s, fps: 670.5, dfps: 651.6

h264(x264-680m, high profile), 1920x1088, 29.97fps 10mbps
libavcodec, gcc4.2.1 r1938
User: 3s, kernel: 0s, total: 4s, real: 13s, fps: 124.0, dfps: 37.6
User: 3s, kernel: 0s, total: 3s, real: 13s, fps: 125.9, dfps: 37.5
User: 3s, kernel: 0s, total: 4s, real: 13s, fps: 123.5, dfps: 37.2
libavocdec_1.dll
User: 3s, kernel: 0s, total: 3s, real: 13s, fps: 128.5, dfps: 36.6
User: 3s, kernel: 0s, total: 4s, real: 13s, fps: 121.6, dfps: 36.6
User: 3s, kernel: 0s, total: 3s, real: 13s, fps: 134.4, dfps: 36.3
libavcodec_2.dll
User: 3s, kernel: 0s, total: 3s, real: 13s, fps: 129.5, dfps: 36.6
User: 3s, kernel: 0s, total: 3s, real: 13s, fps: 132.2, dfps: 36.4
User: 3s, kernel: 0s, total: 4s, real: 13s, fps: 118.1, dfps: 36.6
libavcodec, gcc4.2.1 r1945
User: 1s, kernel: 0s, total: 1s, real: 10s, fps: 377.8, dfps: 47.2
User: 0s, kernel: 0s, total: 1s, real: 10s, fps: 479.3, dfps: 47.0
User: 1s, kernel: 0s, total: 1s, real: 10s, fps: 396.5, dfps: 46.6

[clsid]

H.264 decoding should be a bit faster since rev1941.

[fastplayer]

H.264 decoding should be a bit faster since rev1941.
Xvid's IDCT uses now SSE2 instead of MMX. Maybe it's faster too...

[EpheMeroN]

Has anyone had issues decoding DivX5 video files with ffdshow tryouts? I have a DivX5 file that when opened up in Media Player Classic only the audio plays. Tested video in VLC and plays back fine.

I'm using ffdshow tryouts rev1897 nightly build from Mar 9 2008.

[leeperry]

strangest things happen sometimes :)

I've updated to the latest MPC HC/ffdshow from clsid/EVR

if I enable the luma sharpen to 0.10 in my spline/spline resize, then MPC HC will stay in memory when I close it if I open 2 WMV files in a row.

if I disable the luma sharpen, then MPC HC never hangs at shutdown.

also happens with 1080p MKV files randomly :eek:

[clsid]

@EpheMeroN

ffdshow can decode divx5 just fine. So there is probably something wrong with your file. Try another AVI splitter. Or remux your file with VirtualDub/Avidemux.

[georgevalkov]

H.264 decoding should be a bit faster since rev1941.

:thanks: I have also noticed that at mode: one pass quality=52; min quantizer=18; max quantizer=31:
~ the average quantizer during encoding dropped from 29-31 to 26-27 resulting in better output quality. :)
~ I have not encoded one and same content using the old and the new versions, in order to make an accurate comparison, but the average file size on the new encodings doesn't seem larger at first sight.

[georgevalkov]

I am encoding from an interlaced PAL 704x576 source with ffdshow to H.264 format;
output=store frames to external file: RAW frames.264,
and then I use MP4box to mux it into a MP4 container.

MP4Box.exe -fps 25 -add "%1" -new "%1.v.mp4"

I can also open the RAW .264 stream directly in MPlayer, with the same results, as described below.

If I on the ffdshow encoder I enable both:
[x] Interlaced
[x] B frames
with B frames prediction = [spatial, temporal or auto]

The file plays properly in MPlayer, but it will complain outputting this warning on every frame:
[h264 @ 00D4FB70]MBAFF + spatial direct mode is not implemented

With B frames enabled I save 5% file size, but I don't like that warning to appear.
The file plays properly in both MPlayer and DirectShow using the ffdshow video decoder.

What would You recommend me?
A: ignore the warning
B: disable B frames
C: ?

Thanks for any advice! :thanks:

[clsid]

The same warning applies to ffdshow, although it isn't displayed anywhere.
When you get the warning it means that it is very likely that not everything will be decoded properly. It may look visually OK if your are lucky.
Decoding interlaced video requires more processing power than progressive video. So my advice would be to deinterlace the video before encoding.

[georgevalkov]

Thank You for the advice. :thanks:
I will preserve interlacing on interlaced sources,
but without B frames, to be safe. And I will use
the B frames on progressive sources.

Decoding interlaced video requires more processing power than progressive video. So my advice would be to de interlace the video before encoding.
Yes, de interlaced video will also requires lower bit-rate, but at the cost of:
~ blurring (reduced vertical resolution);
~ motion is no longer smooth.

The advantages of interlaced video are:
~ perfect for TV output;
~ preserves high resolution, when there's less or no motion;
~ preserves smooth motion;
~ when the screen is set to 720x576@100Hz, the interlacing artefacts are less noticeable and usually no deinterlacing is needed;
~ some interlaced 704x576 materials look good at 1152x864@75Hz, and no deinterlacing is needed, artefacts are less noticeable at normal watching distance;
~ setting the BOB flag in the output format of ffdshow, will cause the render to take care of the deinterlacing, preserving the smooth motion. If used with VMR9 it will play at double frame rate 50Hz, preserving both smooth motion and details.



The same warning applies to ffdshow, although it isn't displayed anywhere.
I followed this guide to write my first :) DirectShow console application:
http://msdn2.microsoft.com/en-us/library/ms783787(VS.85).aspx

that simply renders a file:
hr = pGraph->RenderFile(L"M:\\interlaced+b frames.mp4", NULL);
It used the ffdshow video and audio decoders, but did not output any warnings in the console window.

A few days ago I tested the same code to render an AVI:
hr = pGraph->RenderFile(L"M:\\AXE-effect-01.avi", NULL);
And the ffdshow audio decoder showed a single warning message in the console output:
"mpg123: Can't rewind stream by 642 bits!".

Who would have thought! :cool: My first step in DirectShow and it's getting interesting!
Seems like the DirectShow filters can output debugging information to the standard output. :)

[nm]

Yes, de interlaced video will also requires lower bit-rate, but at the cost of:
~ blurring (reduced vertical resolution);
~ motion is no longer smooth.
Both of these problems can be avoided by using a good motion-compensating framerate-doubling deinterlacer like MCBob, MVBob (AviSynth) or yadif+mcdeint (MEncoder). The cost is a higher bitrate (for the same quality) and slower encoding/decoding (due to the higher bitrate and framerate). If you don't want to double the framerate, blurring can still be avoided by using these better deinterlacers and dropping half of the output frames.

The advantages of interlaced video are:
~ perfect for TV output;
But hard to get to the TV from a PC without deinterlacing at some point. Standalone players are better at this.

~ when the screen is set to 720x576@100Hz, the interlacing artefacts are less noticeable and usually no deinterlacing is needed;
I don't see such an effect. At least bobbing is still needed to make it look half-decent.

~ some interlaced 704x576 materials look good at 1152x864@75Hz, and no deinterlacing is needed, artefacts are less noticeable at normal watching distance;
That's because scaling the video vertically less than 2x blends the fields together, so it works like a cheap deinterlacer.

[georgevalkov]

:thanks: Thank You nm, for your opinion and for the names of the filters, I will test them, whenever I have time for experiments.

Both of these problems can be avoided by using a good motion-compensating framerate-doubling deinterlacer like MCBob, MVBob (AviSynth) or yadif+mcdeint (MEncoder). The cost is a higher bitrate (for the same quality) and slower encoding/decoding (due to the higher bitrate and framerate). If you don't want to double the framerate, blurring can still be avoided by using these better deinterlacers and dropping half of the output frames.


Decoding 50 fps is twice as hard as 25 fps. The double bitrate is very hight. It's much cheaper to encode interlaced 25 fps, and to deinterlace after the decoder. In the future, my de-interlaced recordings will look the same... But there will be faster computers and better deinterlacing filters, so my interlaced recordings will look better :cool:.

If I set the BOB flag in ffdshow, EVR and VMR9 will do the hardware BOB deinterlacing, so I can get the smooth 50 fps at the cost of lower CPU usage. Athlon XP 1700+ is not strong enough to do this with software de-interlacing filters. It's also not fast enough to decode H.264@50p, but it can do H.264@25i.

[thuan]

I've just tried the subtitle filter in ffdshow-tryouts again, and it seems like the karaoke rendering timing is still off. And there're still differences in rendering position of some subs. Do you need a sample?

[draggoon01]

i use ffdshow to play subtitles. i'm trying to play a mkv file which contains multiple subtitles (utf-8). how do i select different subtitle languages with ffdshow?

[vlada]

You can select subtitle streams from menu by right clicking on ffdshow or Haali's splitter icon in tray.

[draggoon01]

You can select subtitle streams from menu by right clicking on ffdshow or Haali's splitter icon in tray.

ahh, i normally disable tray icons. enabling ffdshow tray icon let's me select subtitle.

is there any way to select subtitle from within ffdshow properties page?

[leeperry]

Right, very quick and very dirty lesson in AR......

You're resizing a 1440x816 video to the same AR, but 1280 width, so that it displays correctly. That's 1280x544 as your destination. The SAR STAYS as 4:3, because the frame size is still the same dimension multiple. Try it yourself. 1440/816 = 960/544. The SAR is reported as 1.77 because it's the closest to 1440/816. ffdshow isn't reporting the SAR to the renderer, hence why it appears to be 1.77.

Try this: resize JUST the horizontal to 1280, check "Keep Aspect Ratio" down the bottom of the box, and then see. ;)

BTW: 1280/768 = 1.666666666, AKA 5:3. :)

to get back on that matter.

it was OK with MPC HC + EVR on XP.

but it's not with HR :(

with this video :
http://download.microsoft.com/download/c/e/a/ceaf0def-8889-4462-9616-7ce4246569ae/Taxi_3_1080.exe

HR says : 1080x816.....instead of 1920*1080

seems there's some AR hiccup again :(

[Inventive Software]

What's HR?

EDIT: Forget that, I figured it'd be Haali Renderer. I think I need caffeine! :D

Have you tried other renderers, like VMR7, VMR9 (windowed and renderless)? Do they exhibit the same flaw? If so, then call it an ffdshow problem. If they're correct, head to the Haali Renderer thread in Software Players.

[leeperry]

right....thanks for the reply.

I made more tests, it might be due to KMPlayer's own AR implementation.

If I check "Original AR" I get 1.78....I have to enforce 2.35 in KMP

any chance someone could make the Ozone winamp2 plugin avoid deadlocks at shutdown in ffdshow audio ?

you can open 2 files a row, then it starts freezing more and more for each new file.....and after 6 files everything's frozen :(

Haruhiko is the only active ffdshow coder ? :(

[whiterabbit99]

can anyone help me with using the automatic presets?

ffdshow tryout beta4a_rev1723_20071224_clsid

im trying to setup some automatic presets for the video decoder

i have a dvd one -> dimensions must be 720x576
tv -> dimensions must be 704x576
other -> no conditions

"other" preset doesnt do anything, the other 2 do resizes, blurs etc.

i ticked the "automatic preset loading". tried it with "try to load preset from file" ticked and unticked - dont know what that is.

whatever i try the preset loading conditions are ignored - the default preset in bold is always loaded.

i dont understand what im doing wrong - i used the osd to make sure the resolutions above are correct.

please help.

[Leak]

i dont understand what im doing wrong - i used the osd to make sure the resolutions above are correct.
It could be that the image dimensions aren't known when the filters connect, but only when the first frame is sent - and that's after the preset has been loaded.

Preset loading based on FPS doesn't work for DVDs either, for instance.

In that case, preset loading would have to be done whenever the video streams properties (dimensions, frame rate, aspect ratio, etc.) change for this to work.

np: Bill Laswell - Vessel (Y34RZ3R0R3MIX3D)

[leeperry]

BTW, why is the "linear blending" deinterlace so blurry in ffdshow ? :(

I can't use MPC HC's pixel shaders scripts anymore coz I'm back with HR.......but MPC HC's "blend deinterlace" PS script is not blurry at all

any chance of getting the same kind of performance in ffdshow ?

TIA,

[Inventive Software]

Use kernel deinterlace or cubic blending. Kernel recommended, as it made my interlaced videos look very nice. :D

[leeperry]

ok sounds like a plan, I'll try thanks!

any settings that work for everything with Kernel ?

it seems that setting it to 10 and checking "sharp" gives nice results :
http://leak.no-ip.org/AviSynth/LeakKernelDeint/LeakKernelDeintHelp.html

[Leak]

any settings that work for everything with Kernel ?
Well, I personally use ffdshow's KernelBob (about the same code as my AviSynth plugin) with a threshold of 0 - as long as your display has the correct framerate (i.e. 50Hz for PAL material, or 59.940Hz for NTSC) it looks great.

np: Justus Köhncke - Wo Bist Du (Doppelleben)

[whiterabbit99]

so how do i use the presets - if i want specific presets to apply just to dvds for example.

i tried using just decoder match as mpeg2video but it still uses which ever preset is the default :(

[wyrd]

for dvds, i've been ticking both of dimensions and fourcc.
http://tirnanog.fate.jp/tmp/snap/preset_autoload1_t.jpg (http://tirnanog.fate.jp/tmp/snap/preset_autoload1.png)

[Leak]

so how do i use the presets - if i want specific presets to apply just to dvds for example.
What works for me is setting "DirectShow filter presence" to "DVD Navigator".

np: Nine Inch Nails - Ghosts I (Ghosts I-IV)

[rebkell]

How is the Input bitrate calculated, the one in the OSD option for the video decoder?

[draggoon01]

request/suggestion:

for mkv files with multiple embedded subtitles, the only way to choose different languages is by right clicking on the tray icon which shows "subtitle sources" menu under "subtitles".

could the "subtitle sources" menu be added to ffdshow configuration/properties as a sub-menu item under "subtitle" along with "text", "font", "vobsub" ?

[leeperry]

is it me or "accept embedded subtitles" is checked by default in the subtitles section of ffdshow ?

that's very annoying, because I couldn't get subtitles to work in KMPlayer with "Haali's Media Splitter" because that option was checked.

I've never touched this section of ffdshow, and it was disabled :eek:

[Jong]

Hi everyone. New to this thread, but a long time user of ffdshow. Thanks for all the hard work.

As the RGB Converter is just about vital now when upscaling SD, as drivers assume BT.709 when passed HD, is there any plan to multi-thread RGB Conversion, as was done with the resize filter.

It is really punishing right now!

[leeperry]

hey Jong ;)

to get back on deinterlacing, kernel bob required double frame rate, so not pratical at all.

kernel deinterlace makes ugly artefacts.

cubic blending doesn't really do its job.

only "linear blending" works fine, but it's blurry as hell.

I really miss the "blending" PS script from MPC :(

anyone knows a non-blurry good blending interlace algorithm ? D-Scaler plugin maybe ?

thanks!

[haruhiko_yamagata]

request/suggestion:

for mkv files with multiple embedded subtitles, the only way to choose different languages is by right clicking on the tray icon which shows "subtitle sources" menu under "subtitles".

could the "subtitle sources" menu be added to ffdshow configuration/properties as a sub-menu item under "subtitle" along with "text", "font", "vobsub" ?
It is not possible. You can select subtitle track by right clicking Haali's icon.

[haruhiko_yamagata]

is it me or "accept embedded subtitles" is checked by default in the subtitles section of ffdshow ?

that's very annoying, because I couldn't get subtitles to work in KMPlayer with "Haali's Media Splitter" because that option was checked.

I've never touched this section of ffdshow, and it was disabled :eek:
I don't understand. Could you send me the sample?

[leeperry]

well, it happens with any movie in KMPlayer.

is this option checked by default ?

because even if the "subtitles" section of ffdshow is disabled, it still blocks the subs from reading in KMP :(

so anyone knows of a way to get nice deinterlacing in ffdshow w/o jaggies and bluriness ?

[vlada]

haruhiko_yamagata> I think I understand what leeperry is trying to say. He wants KMPlayer to render subtitles, but the they are treated by ffdshow. The checkbox "accept embedded subtitles" used to be unchecked. I haven't tried any of latest versions if this has changed.

leeperry> Kernel bob is doing a good job for me. It doubles the framerate which is the best thing to do. It would be nice to have yadif ported to ffdshow.

Any blending deinterlacer is supposed to create a blurry image. If you want a sharp picture, you have to use motion adaptive bob.

[leeperry]

yes, that's what I meant.

well I've just tried everything on some strongly interlaced 29.97 video, and nothing looks as good and sharp as doubling the "deinterlace(blend)" PS script from MPC.

kernel bob does terrible artefacts, and anyway I run my CRT in 90Hz.

KMP also supports stacking PS scripts with EVR on XP, so that's pretty cool :cool:

[haruhiko_yamagata]

@leeperry
Then just uncheck that option. I won't support KMPlayer.

[leeperry]

yes I did, but wondering why ffdshow was "keeping" subs to itself, even though the "subtitles" ffdshow section was disabled and I never touched it ?!

it might occur with other players ;)

I understand KMPlayer breaks the GPL, I'm sorry for that.....it's a very good player(basically MPC on steroids) and a lot more frame stable than MPC HC(never misses the VSYNC, even after 60/90').....I regret its author decided to convert MPC code to Delphi and call it his :(

I think Reclock is also breaking GPL, but I won't stop using it either :p

[haruhiko_yamagata]

yes I did, but wondering why ffdshow was "keeping" subs to itself, even though the "subtitles" ffdshow section was disabled
Because ffdshow has to show subtitles if subtitles (in ffdshow configuration) are enabled during playback.
ffdshow has to get subtitles from input pin for that purpose.
Haali's media splitter can connect to both of MPC's internal renderer and ffdshow at the same time.

[MfA]

Any chance of more tarballs? SVN checkout is tortuously slow. (Or is that just divine punishment for using TortoiseSVN instead of a command line version?)

[puffpio]

is ffdshow still respecting aspect ratios?

I have an mp4 file w/ an h264 video stream in it
the video stream is 720x480 pixels and mp4box is reporting a PAR of 32:27
this should get me a 16:9 for a typical anamorphic widescreen DVD

ffdshow is reporting:
Input aspect ratio: SAR: 1/1, DAR: 3/2
Output aspect ratio: SAR: 1/1, DAR: 3/2

I'm using a 19" LCD monitor 1280x1024 so the DAR maybe coming from there

but why is the SAR not being picked up from the PAR in the mp4 stream?

I'm playing it in media player classic home cinema using the EVR renderer...I've tried in w/ haali's renderer and plain overlay with similar results

anything I'm missing?

[SeeMoreDigital]

I'm playing it in media player classic home cinema using the EVR renderer...I've tried in w/ haali's renderer and plain overlay with similar results

anything I'm missing?Hi puffpio,

You're not alone... I experienced this very issue a few weeks ago and reported it here: -

http://forum.doom9.org/showthread.php?p=1109947

In the meantime I've gone back to the regular version of MediaPlayer Classic.

[leeperry]

Any blending deinterlacer is supposed to create a blurry image. If you want a sharp picture, you have to use motion adaptive bob.

Seb.26 has improved the PS deinterlacing script so it doesn't need to be doubled anymore.

maybe it can be improved even further, we'll see about that......

but on strongly interlaced files, it looks better than anything I've seen from ffdshow....it doesn't blur at all and 95% of the lines are killed :eek:

feel free to comment ppl ;)

this works with MPC and KMPlayer as long as you choose "3D surface rendering" in VMR9 or EVR.

could this script be translated to ffdshow, somehow ?! :eek:

// deinterlace (blend).txt=ps_3_0
sampler s0 : register(s0);
float4 p0 : register(c0);
float4 p1 : register(c1);

#define width (p0[0])
#define height (p0[1])
#define counter (p0[2])
#define clock (p0[3])
#define one_over_width (p1[0])
#define one_over_height (p1[1])

#define PI acos(-1)

float4 main(float2 tex : TEXCOORD0) : COLOR
{

float2 h = float2(0, one_over_height);
float4 c1 = tex2D(s0, tex-h);
float4 c2 = tex2D(s0, tex+h);
float4 c0 = (c1+c2)/2;

return c0;
}

[Shinigami-Sama]

Hi puffpio,

You're not alone... I experienced this very issue a few weeks ago and reported it here: -

http://forum.doom9.org/showthread.php?p=1109947

In the meantime I've gone back to the regular version of MediaPlayer Classic.

so I wasn't going crazy when it refused to give me the proper resolution of some files I have...

[cc979]

is there a problem ffdshow-tryout svn1954 being compiled with gcc-4.0.4

make[1]: Entering directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/imgFilters'
gcc -c -mno-cygwin -mdll -fno-rtti -mthreads -pipe -D_WINGDI_ -DUCLIBCPP -D_GLIBCPP_HAVE_MBSTATE_T -D_WIN32_IE=0x0500 -DARCH_IS_IA32 -DARCH_IS_32BIT -DHAVE_MMX -mmmx -msse -mfpmath=sse -w -DNDEBUG -UDEBUG -DFFDEBUG=0 -I. -I.. -Iuclibc++ -Ibaseclasses -I../baseclasses -IimgFilters -I../imgFilters -Implayer -I../mplayer -Isettings -I../settings -Isettings/filters -I../settings/filters -Icodecs -I../codecs -Isubtitles -I../subtitles -Iconvert -I../convert -Idialog -I../dialog -IaudioFilters -I../audioFilters -Icygwin -I../cygwin -Iffmpeg -I../ffmpeg -Iacm -I../acm -Ifilters -I../filters -Imuxers -I../muxers -I/dx/Include -L/dx/MingLib -ldx9 -O3 -march=i686 -fomit-frame-pointer -MMD -o ffdshow_imgfilters_all.o ffdshow_imgfilters_all.cpp
gcc.exe: -ldx9: linker input file unused because linking not done
make[1]: *** No rule to make target `../ffmpeg/libavcodec/i386/skl_dct_sse.o', needed by `.all'. Stop.
make[1]: Leaving directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/imgFilters'
make: *** [lib] Error 2

cheers

[clsid]

Try this (untested) patch:
Index: Makefile
===================================================================
--- Makefile (revision 1954)
+++ Makefile (working copy)
@@ -4,7 +4,8 @@

SRCS_C=ffdebug.c

-SRCS_NASM=cpuid.asm memcpy.asm subtitles/fontRenderer.asm convert/colorspace_rgb_mmx.asm
+SRCS_NASM=cpuid.asm memcpy.asm subtitles/fontRenderer.asm convert/colorspace_rgb_mmx.asm \
+ imgFilters/Skal/skl_dct_sse.asm imgFilters/Skal/skl_fdct_mmx.asm

include makefile.inc

@@ -58,9 +59,8 @@
baseclasses/baseclasses_all.o \
ffmpeg/libavcodec/golomb.o \
audioFilters/TaudioFilterConvert.o audioFilters/dotproduct.o \
- imgFilters/simple_idct_mmx.o ffmpeg/libavcodec/i386/skl_dct_sse.o \
- ffmpeg/libavcodec/i386/skl_dct_sse2.o ffmpeg/libavcodec/i386/skl_fdct_mmx.o \
- ffmpeg/libavcodec/i386/skl_fdct_sse2.o
+ imgFilters/simple_idct_mmx.o imgFilters/Skal/skl_dct_sse.o \
+ imgFilters/Skal/skl_fdct_mmx.o

lib:
$(MAKE) -C baseclasses

[cc979]

thanks but still same error

[Liisachan]

Hi,
I think the embedded sub in MKV is out of sync (by about 0.5 sec?) in mplayerc_20080308 mplayerc_20080414 from Guliverkli2, if the internal MKV splitter is used, and if Queue is enabled in ffdshow.
mpc 20080127 build works ok, so this could be a pb of MPC.

(Edit: I mean, if MPC's internal sub renderer is used in VMR9. And output color space is RGB32 btw)

Tested ffdshow tryouts official beta4 and the newest nightly (ffdshow_rev1953_20080504_xxl.exe)

Test clip where softsubs and hardsubs should have the same timing
http://ffdshow.faireal.net/tmp/Hard_vs_Softsubs.mkv 48kb

http://ffdshow.faireal.net/tmp/hardsub_softsub.jpg

[clsid]

Confirmed. The only change to MPC in that period that is related to the above is this one:
http://guliverkli2.svn.sourceforge.net/viewvc/guliverkli2?view=rev&revision=44

I don't see any obvious bugs in that revision. Maybe someone else is able to spot it.

[Liisachan]

Thanks for confirmation. if anything, this change (http://guliverkli2.svn.sourceforge.net/viewvc/guliverkli2/src/subpic/ISubPic.cpp?r1=44&r2=43&pathrev=44) is fishy, but since MPC works ok if queue is not enabled in ffdshow, perhaps this is not a pb of MPC, but rather related to the hack used by ffdshow queue...

[thuan]

This is for EVR custom and MPCHC but it might be related to sub desynchronizing issue http://mpc-hc.svn.sourceforge.net/viewvc/mpc-hc?view=rev&revision=460.

[clsid]

The regular version of MPC does not have that code.

I will mess around with rev44 to see if I can fix the problem. But it may take a few days as I've many other things to do.

[haruhiko_yamagata]

Hi,
I think the embedded sub in MKV is out of sync (by about 0.5 sec?) in mplayerc_20080308 mplayerc_20080414 from Guliverkli2, if the internal MKV splitter is used, and if Queue is enabled in ffdshow.
mpc 20080127 build works ok, so this could be a pb of MPC.

I can't reproduce. Can this be related to video card driver? Mine is nVidia 7900GS.

[fastplayer]

Can't reproduce, too.

Setup:
ffdshow rev1945
MPC rev49
Windows XP64
Radeon X800 GTO + Catalyst 8.4

[Liisachan]

Thank you, clsid!

Additional info:
- Enable VMR9 Mixer mode and I see this problem. If uncheck that , the problem is gone. So I tend to think now this is MPC's problem.

- both internal and Haali's MatroskaSplitter have the same problem, so Splitter is not related.

I have this same problem on 2 PCs. One has nVidia quadro fx 500, the other nVidia GeForce 8400GS.

[fastplayer]

Enable VMR9 Mixer mode and I see this problem.
Yep, now subs are out of sync.
Output colorspace doesn't matter, by the way.

[haruhiko_yamagata]

Thanks, I can reproduce now.

[Keepitsimple]

When is it time for the next beta release?

Excellent work so far!

[ImAhNoBoDy]

Does ffdshow create thumbnails for avi files? Cause I notice when I encode a new avi file that ffdshow appears and goes away real fast and my thumbnail is created. I was wondering why it doesn't work for ogm or mkv files?

[clsid]

Thumbnail generation is done by a shell extension of Explorer. It is possible to activate it for any file extension using a simple registry tweak:
Windows Registry Editor Version 5.00

[HKEY_CLASSES_ROOT\.flv\ShellEx\{BB2E617C-0920-11D1-9A0B-00C04FC2D6C1}]
@="{c5a40261-cd64-4ccf-84cb-c394da41d590}"

[ImAhNoBoDy]

Hm....I'm not sure if it's Vista or not, but the thumbnails don't show for it. I know I can use the Haali reg hack for it, but I can't seem to get it to work without it. Even .flv files don't create a thumbnail.

[TheShadowRunner]

small cosmetic bug
on the "Preset autoload conditions" window, it says :
- On one of the conditoins match (OR)

small request (for Haruhiko i think)
In the Subtitles settings, could we have a feature to temporarily disable FFDshow internal subtitle support if Directvobsub is detected in the graph?

I tried to do this by setting 2 profiles.
In the first one, i have "Subtitles" enabled.
In the second one, I disabled "Subtitles" and enabled the following condition:
"on a DirectShow filter presence" = DirectVobSub

Then enabled "Automatic preset loading" but it just doesn't work, Subtitles are always enabled regardless if DirectVobSub is present in the graph or not.. (the 2nd profile is never automatically loaded)

See you,

TSR

[MatMaul]

Confirmed. The only change to MPC in that period that is related to the above is this one:
http://guliverkli2.svn.sourceforge.net/viewvc/guliverkli2?view=rev&revision=44

I don't see any obvious bugs in that revision. Maybe someone else is able to spot it.
the problem does not come frome that rev, mplayerc_20080127 (something like rev36) has the same bug.

[monohouse]

I was wondering, in the ffdshow audio decoder, when input sample format is not equal output sample format and also at the same time there is a few mixer operations inbetween, is the mixing performed before the sample format conversion or after ?

and what can help indicate that results are bit-identical ?

I am using madflac to decode and selected uncompressed in ffdshow audio decoder to use the mixer to convert 5.1 into 2.0, but I am unsure as to what happens in the ffdshow decoder to know what is best, to loose some data in the extra channels or go through sample format covertion, knowing the signal flow would help me decide what is better

[Inventive Software]

I'm not sure what you're after. Sample conversion is, AFAIK, never applied. Down-mixing's done at the input sample rate, there is no conversion.

[monohouse]

then maybe I missed something, how is it that in the Output menu when I choose different sample formats the sound changes and you can hear it

[haruhiko_yamagata]

small cosmetic bug
on the "Preset autoload conditions" window, it says :
- On one of the conditoins match (OR)
Thank you. I have fixed just now.

small request (for Haruhiko i think)
In the Subtitles settings, could we have a feature to temporarily disable FFDshow internal subtitle support if Directvobsub is detected in the graph?

I tried to do this by setting 2 profiles.
In the first one, i have "Subtitles" enabled.
In the second one, I disabled "Subtitles" and enabled the following condition:
"on a DirectShow filter presence" = DirectVobSub

Then enabled "Automatic preset loading" but it just doesn't work, Subtitles are always enabled regardless if DirectVobSub is present in the graph or not.. (the 2nd profile is never automatically loaded)

See you,

TSR"on a DirectShow filter presence" is incomplete. The preset automatic loading is applied at the early point of DirectShow filter connection. ffdshow is not connected to downstream when the preset settings are loaded.

[saint-francis]

Why is it that when I install ffdshow avisynth doesn't work for me any more?

I get this using DGDecode_mpeg2source

LoadPlugin: unable to get plugin list $Plugins$, loading "C:\etc\etc\DGDecode,dll"

and this using directshowsource

Evaluate: system exception - access violation (path\to\script.avs)

[Leak]

Why is it that when I install ffdshow avisynth doesn't work for me any more?
Did you install the ffdshow AviSynth plugin which lets you use ffdshow's filters in AviSynth? Maybe it's falling over on your machine for some reason.

While it's a C plugin called ffavisynth.dll that doesn't (or at least shouldn't) get loaded automatically, ffdshow also installs an ffavisynth.avsi file into AviSynth's plugin directory what uses Load_Stdcall_Plugin to load ffavisynth.dll.

Try renaming the avsi file, and if that doesn't work try moving both files out of AviSynth's plugins folder. I have to say though that I don't have a problem with both here...

[saint-francis]

Did you install the ffdshow AviSynth plugin which lets you use ffdshow's filters in AviSynth?

Yes I did


Try renaming the avsi file, and if that doesn't work try moving both files out of AviSynth's plugins folder. I have to say though that I don't have a problem with both here...

I did that and I still had some serious problems. I started getting all kinds of errors saying that I was missing all kinds of .ddl's like MSVCP71.dll. I uninstalled FFDShow... still problems. I reinstalled....still problems. I ended up installing an Acronis back up to fix this all.
BTW the previous installation of FFDShow was on a fresh installation of XP SP3.

[jmartinr]

Just installed latest ffdshow build: ffdshow_rev1958_20080514_clsid.exe and playback of h264 is broken (blocks and errors in MPC). Reverting to 1943 re-enabled normal playback. Test files are at: http://www.nkbv.nl/sportklimmen/media/wedstrijdvideos/4237.html

[gruppo_sd]

With the last release of CLSID unable to play H264, works fine with previous.
Win xp sp3 gforce 8600gt-175.16whql

[clsid]

It is a bug in the multi-threading code. Set the number of threads to 1 and it will play ok.

I will revert the changes and upload a new build.

[TheShadowRunner]

Thank you. I have fixed just now.
"on a DirectShow filter presence" is incomplete. The preset automatic loading is applied at the early point of DirectShow filter connection. ffdshow is not connected to downstream when the preset settings are loaded.

Thanks for explaining, I get it. ;)
Is there any solution you can think of to temporarily & automatically disable FFDShow internal Subtitles treatment if Directvobsub comes next in the graph (after ffdshow)?
An option to do so in the Subtitles settings page would be nice if not ^^
See you,

TSR

[fastplayer]

DivX H.264 Decoder Beta 1 (http://labs.divx.com/node/6443)

http://labs.divx.com/files/pictures/DH264_Beta1_Chart5.png

That's some really nice scaling they're using for the bar chart. Makes ffdshow look as if it's over 50% slower than the commercial solutions... :rolleyes:

[sheppaul]

DivX H.264 Decoder Beta 1 (http://labs.divx.com/node/6443)

http://labs.divx.com/files/pictures/DH264_Beta1_Chart5.png

That's some really nice scaling they're using for the bar chart. Makes ffdshow look as if it's over 50% slower than the commercial solutions... :rolleyes:

Oops! It confuses me. :eek:

[leeperry]

and it's a single CPU test.

one can guess the gap is not that big with dualcore ;)

[Dark Shikari]

and it's a single CPU test.

one can guess the gap is not that big with dualcore ;)The gap is much larger with dual-core because FFDshow isn't fully multithreaded ;)

[leeperry]

anyhow, speed means nothing.

I really don't like how CoreAVC decodes.....like it's lacking high frequency details.

nothing beats ffdshow, quality wise :)

[Dark Shikari]

anyhow, speed means nothing.

I really don't like how CoreAVC decodes.....like it's lacking high frequency details.

nothing beats ffdshow, quality wise :)As far as I know CoreAVC's output is bit-accurate as it should be.

DivX's decoder not only matches FFDshow in output: it actually makes some of the same mistakes FFDshow does too.

[Inventive Software]

There may be some license issues creeping in there I feel. :)

[leeperry]

As far as I know CoreAVC's output is bit-accurate as it should be.

well I don't like the way CoreAVC deblocks I think.

and I'm not the only one who prefers ffdshow for x264 ?!

well the DivX empire was built on a cracked beta ASF .dll, why not stealing from ffdshow now :D

[Dark Shikari]

There may be some license issues creeping in there I feel. :)You better not start accusing them of code-stealing, because ffdshow/ffmpeg is not exactly innocent of this either ;)
well I don't like the way CoreAVC deblocks I think.Again, deblocking is specified by the standard. It won't vary between decoders.

[Inventive Software]

I wasn't accusing them of anything, but if it's base-lined off ffmpeg's decoder, then all credit to them for improving it so damn much!

[Shinigami-Sama]

anyhow, speed means nothing.

I really don't like how CoreAVC decodes.....like it's lacking high frequency details.

nothing beats ffdshow, quality wise :)

you're just picky

coreand ffsahow output the same picture, anything else how ever you're filtering it and what your drivers do to it

[leeperry]

you're just picky

coreand ffsahow output the same picture, anything else how ever you're filtering it and what your drivers do to it

oh yes, I'm SO picky :D

and I'm not the only one who prefers ffdshow, far from it.

just a random example :
http://www.avsforum.com/avs-vb/archive/index.php/t-920229.html

Yeah, they should but they aint. As can be read in many places, CoreAVC while fastest seems to have lower quality. It takes some shortcuts to be so fast, and PQ suffers.

I have to admit, I don't use crappy LCD based displays.
only CRT and DLP projectors with high resolution glass optics.

I find that CoreAVC messes around with the picture grain :eek:

[3ngel]

I want to signal a color bug with a YV16 input (latest Elecard Mpeg2 Decoder).

This is the result

http://img216.imageshack.us/img216/7957/snapshot20080516232337pgz3.png

Hope this will be fixed.

Thanks

[Thunderbolt8]

anyhow, speed means nothing.

I really don't like how CoreAVC decodes.....like it's lacking high frequency details.

nothing beats ffdshow, quality wise :)
exactly, quality is more important. CPUs get faster anyway, so the focus on quality is why im still with ffdshow.

[cc979]

i got a problem compiling ffdshow-tryout svn 1964

it would not do a make clean

$ make clean
makefile.inc:75: *** missing `endif'. Stop.

had to remove a # from line 20 for it to start to compile but

i still get this error

make -C imgFilters
make[1]: Entering directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/imgFilters'
gcc -c -mno-cygwin -mdll -fno-rtti -mthreads -pipe -D_WINGDI_ -DUCLIBCPP -D_GLIBCPP_HAVE_MBSTATE_T -D_WIN32_IE=0x0500 -DHAVE_MMX -DARCH_IS_IA32 -DARCH_IS_32BIT -mmmx -msse -mfpmath=sse -w -DNDEBUG -UDEBUG -DFFDEBUG=0 -I. -I.. -Iuclibc++ -Ibaseclasses -I../baseclasses -IimgFilters -I../imgFilters -Implayer -I../mplayer -Isettings -I../settings -Isettings/filters -I../settings/filters -Icodecs -I../codecs -Isubtitles -I../subtitles -Iconvert -I../convert -Idialog -I../dialog -IaudioFilters -I../audioFilters -Icygwin -I../cygwin -Iffmpeg -I../ffmpeg -Iacm -I../acm -Ifilters -I../filters -Imuxers -I../muxers -I/dx/Include -L/dx/MingLib -ldx9 -O3 -fomit-frame-pointer -march=i686 -MMD -o ffdshow_imgfilters_all.o ffdshow_imgfilters_all.cpp
gcc.exe: -ldx9: linker input file unused because linking not done
make[1]: *** No rule to make target `../ffmpeg/libavcodec/i386/skl_dct_sse.o', needed by `.all'. Stop.
make[1]: Leaving directory `/home/User/svn/ffdshow-tryout/ffdshow-tryout/src/imgFilters'
make: *** [lib] Error 2


any clues, i am doing something wrong now ?

cheers

*updated
i managed to get a sucessful compile by do this

src\imgFilters\makefile

had to change line 7 & 8 to

SRCS_NASM=skal/skl_dct_sse.asm \
skal/skl_fdct_mmx.asm \


and

src\makefile

had to change line 61 & 62 and blank line 63

imgFilters/simple_idct_mmx.o imgFilters/Skal/skl_dct_sse.o \
imgFilters/Skal/skl_fdct_mmx.o \

and moved at line 211 to the top
ifeq ($(64BIT),yes)
TARGET_OS=i686-pc-mingw64
else
TARGET_OS=i686-pc-mingw32
endif



and

src\makefile_C.inc

line 11 into
TARGET_OS=i686-pc-mingw64

line 13
TARGET_OS=i686-pc-mingw32

[madshi]

To the ffdshow developers:

I think there is a need for discussion about the correct channel mask for 5.1 and 6.1 channels.

http://www.microsoft.com/whdc/device/audio/multichaud.mspx
http://msdn.microsoft.com/en-us/library/aa474707.aspx

It seems to me that ffdshow often uses WAVE_FORMAT_PCM, so there's no channel mask at all. But which channel mask are you using if you use WAVE_FORMAT_EXTENSIBLE? I think the default for 5.1 used to be 0x3f, right? According to newest MS headers this channel mask seems to be obsolete and 0x60f seems to be recommended now. See ksmedia.h here:

// DirectSound Speaker Config
#if (NTDDI_VERSION >= NTDDI_WINXP)
#define KSAUDIO_SPEAKER_DIRECTOUT 0
#endif
#define KSAUDIO_SPEAKER_MONO (SPEAKER_FRONT_CENTER)
#define KSAUDIO_SPEAKER_STEREO (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT)
#define KSAUDIO_SPEAKER_QUAD (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_BACK_LEFT | SPEAKER_BACK_RIGHT)
#define KSAUDIO_SPEAKER_SURROUND (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_FRONT_CENTER | SPEAKER_BACK_CENTER)
#define KSAUDIO_SPEAKER_5POINT1 (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY | \
SPEAKER_BACK_LEFT | SPEAKER_BACK_RIGHT)
#define KSAUDIO_SPEAKER_7POINT1 (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY | \
SPEAKER_BACK_LEFT | SPEAKER_BACK_RIGHT | \
SPEAKER_FRONT_LEFT_OF_CENTER | SPEAKER_FRONT_RIGHT_OF_CENTER)

#if ( (NTDDI_VERSION >= NTDDI_WINXPSP2) && (NTDDI_VERSION < NTDDI_WS03) ) || (NTDDI_VERSION >= NTDDI_WS03SP1)

#define KSAUDIO_SPEAKER_5POINT1_SURROUND (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY | \
SPEAKER_SIDE_LEFT | SPEAKER_SIDE_RIGHT)
#define KSAUDIO_SPEAKER_7POINT1_SURROUND (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT | \
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY | \
SPEAKER_BACK_LEFT | SPEAKER_BACK_RIGHT | \
SPEAKER_SIDE_LEFT | SPEAKER_SIDE_RIGHT)
// The following are obsolete 5.1 and 7.1 settings (they lack side speakers). Note this means
// that the default 5.1 and 7.1 settings (KSAUDIO_SPEAKER_5POINT1 and KSAUDIO_SPEAKER_7POINT1 are
// similarly obsolete but are unchanged for compatibility reasons).
#define KSAUDIO_SPEAKER_5POINT1_BACK KSAUDIO_SPEAKER_5POINT1
#define KSAUDIO_SPEAKER_7POINT1_WIDE KSAUDIO_SPEAKER_7POINT1

#endif // XP SP2 and later (chronologically)
What do you think?

[haruhiko_yamagata]

What do you think?
Microsoft is responsible for this mess :devil:.

For 5.1ch, 0x3f should be fine. It depends on sound driver's implementation.
Please let me know if anyone has any problems with your card.

[madshi]

Microsoft is responsible for this mess :devil:.

For 5.1ch, 0x3f should be fine. It depends on sound driver's implementation.
Please let me know if anyone has any problems with your card.
The new GeForce 8200 mainboard with integrated HDMI port which supports 7.1 PCM transport doesn't like 0x3f. When playing back 5.1 with 0x3f, the surround channels are placed in the back channels instead of the surround channels.

Now I think this is a driver bug, since 0x3f should be special cased for backwards compatability. However, according to the latest Microsoft headers, 0x3f is actually outdated and 0x60f would be the more correct channel mask to use.

Thoughts?

[leeperry]

exactly the same problem with WMV-HD files on my system!

M-Audio Audiophile USB in Kernel Streaming on XP SP2.

I have to insert the mixer ffdshow filter, set to STEREO to get the front channels.

I love how there is an explanation to everything in this world :D

PS : problem occurs with the Taxi3 1080p sample on the official WMV HD site.

[haruhiko_yamagata]

I want to signal a color bug with a YV16 input (latest Elecard Mpeg2 Decoder).
What is YV16? Y-U-V or Y-V-U?
YUVTools (http://www.sunrayimage.com/index.html):YV16 (YUV422, Y-U-V, progressive, planar)
FOURCC.org (http://www.fourcc.org/yuv.php):8 bit Y plane followed by 8 bit 2x1 subsampled V and U planes.

[haruhiko_yamagata]

The new GeForce 8200 mainboard with integrated HDMI port which supports 7.1 PCM transport doesn't like 0x3f. When playing back 5.1 with 0x3f, the surround channels are placed in the back channels instead of the surround channels.You have 7.1ch sound system and you don't want sound from back speakers when you play 5.1ch file?

Where are your surround speakers placed? What is preferred as default?

Either way, it is possible to make mixer more configurable.

[3ngel]

What is YV16? Y-U-V or Y-V-U?
YUVTools (http://www.sunrayimage.com/index.html):
FOURCC.org (http://www.fourcc.org/yuv.php):

I don't know what sequence does it refers.

I got the value from FFDshow Raw Processor, with an Elecard Decoded Mpeg2 Stream in it.

So i suppose, that because as far as i know Elecard is the only decoder that actually output YV16, can FFShow adeguate to it?

Personally i've never heard of YVU sequence (and if you do a YVU Google search it redirects to a YUV search)

EDIT:
BTW from wikipedia:
16 bit Y'UV to RGB conversion formulae

r = min((9535 * (y - 16) + 13074 * (v - 128)) >> 13,255)
g = min((9535 * (y - 16) - 6660 * (v - 128) - 3203 * (u - 128)) >> 13,255)
b = min((9535 * (y - 16) + 16531 * (u - 128)) >> 13,255)

I don't know if actually ffdshow uses this formula, but this can be pheraps considered "standard"?

EDIT2:
Oh i forgot to tell.
With the Leadtools "YUV Converter" DirectShow filter, the same YV16 is correctly decoded to RGB, so i think it's a matter of FFDShow.

[tebasuna51]

The new GeForce 8200 mainboard with integrated HDMI port which supports 7.1 PCM transport doesn't like 0x3f. When playing back 5.1 with 0x3f, the surround channels are placed in the back channels instead of the surround channels.

Thoughts?

I think is the correct behaviour.

The mask 0x30 correspond to Back channels and not to Side channels (don't mistake with Surround, please speak properly).

I'm sure a 7.1 system must have a way to send the surround channels, from a 5.1 audio, to Back and Side channels.

[madshi]

You have 7.1ch sound system and you don't want sound from back speakers when you play 5.1ch file?

Where are your surround speakers placed? What is preferred as default?

Either way, it is possible to make mixer more configurable.
It's not my system. A madFlac user reported this problem to me. As a result I did some research about the channel masks and found that Microsoft has recently (with XPSP2) changed its mind and wants us to use 0x60f now instead of 0x3f for typical 5.1 content. See the links and header extract I posted earlier...

I think is the correct behaviour.

The mask 0x30 correspond to Back channels and not to Side channels (don't mistake with Surround, please speak properly).
You're right, I should have chosen the words more carefully. Anyway. So what is the consequence of all this? We should change 0x3f to 0x60f, right? Or did I misunderstand your comment?

[juanchu]

Hi, and thank you so much for this incredible project and for all the job that you are doing!

Please, sorry for my poor english, but I'll try to explain al well as I can.

I use FFDShow for decoding, but too for doing a downmix to 2 channels ("uncompressed" activated). Well, I've seen that, if I load a file which has multiple audio tracks, FFDShow will be loaded only the first time. So, if I change the audio track, FFDShow won't be loaded, and I won't be able to do the downmixing of all file's audio tracks properly.

So. if the first audio is loaded in this way:

-Splitter --> Audio Decoder --> FFDShow Audio Decoder --> Default Directsound Device

when I change the audio track, this is what happens:

-Splitter --> Audio Decoder --> Default Directsound Device

Moreover, if I come back to the first audio which was loaded, FFDShow isn't used, although first time it was used to do the downmixing properly.

I've seen that if audio decoder changes when audio track is changed, this happens.

Could you take a look to this, and see if this has a solution?

Thank you so much!

[Leak]

So. if the first audio is loaded in this way:

-Splitter --> Audio Decoder --> FFDShow Audio Decoder --> Default Directsound Device

when I change the audio track, this is what happens:

-Splitter --> Audio Decoder --> Default Directsound Device

Moreover, if I come back to the first audio which was loaded, FFDShow isn't used, although first time it was used to do the downmixing properly.
Have you tried changing "Multiple ffdshow instances" on the "DirectShow control" configuration page of the ffdshow audio decoder to something else? Setting it to "No limitations" should do the trick...

[tebasuna51]

You're right, I should have chosen the words more carefully. Anyway. So what is the consequence of all this? We should change 0x3f to 0x60f, right? Or did I misunderstand your comment?

If you want the Surround channels from a 5.1 played in Side channels, in 7.1 systems, of course you must use the ChannelMask 0x60f.

For me you can use this channel mask, only I don't know if this can cause troubles in old standard 5.1 systems.

But a 7.1 system must offer the option to distribute the Surround channels, from a 5.1, in Side and Back channels not matter if the channel mask is 0x3f or 0x60f (backward compatibility).

[nautilus7]

If you want the Surround channels from a 5.1 played in Side channels, in 7.1 systems, of course you must use the ChannelMask 0x60f.
Isn't this the right thing? :confused:

[juanchu]

Have you tried changing "Multiple ffdshow instances" on the "DirectShow control" configuration page of the ffdshow audio decoder to something else? Setting it to "No limitations" should do the trick...

Unfortunately, this doesn't work, Leak.

I've activated the option that you told me, but the only thing I've seen is that in Windows Media Player a lot of FFDShow instances are loaded, but only one in Media Player Classic, the same only one that always had been loaded. Changing audio track encoded with diferent codecs does the same in WMP and MPCHC, and aren't heared properly. In WMP, when I change audio track, all FFDShow icons disappear, so it seems FFDShow instances are unloaded :confused:

I've used a MKV file with MP3, AAC and Ogg Vorbis audio tracks. I've used too a MKA file generated by me with three audio tracks: one in FLAC (8 channels), one in TTA (8 channels) and one in WavPack (8 channels). If I load the original files, one by one, independently that are or not inside a MKA, there are no problem, but if I use them in an only file (the MKA I mentioned), FFDShow won't work in audio changing. The only exception is that audio tracks have the same encoding; then, FFDShow continues loaded and are heared correctly, with the downmixing done properly.

I don't know if this info may help, but I'll say it: I'm using latest Haali Media Splitter, and I'm using too latest build of FFDShow tryouts.

Thank you so much for your help!

[_xxl]

I have successfully cross-compiled libavcodec from win32 to win64 using MinGW GCC 4.4.0.
This had to be changed:
Index: idct_sse2_xvid.c
===================================================================
--- idct_sse2_xvid.c (revision 1965)
+++ idct_sse2_xvid.c (working copy)
@@ -338,7 +338,7 @@
"movdqa %%xmm6, 4*16("dct") \n\t" \
"movdqa "SREG2", 7*16("dct") \n\t"

-inline void ff_idct_xvid_sse2(short *block)
+void ff_idct_xvid_sse2(short *block)
{
asm volatile(
"movq "MANGLE(m127)", %%mm0 \n\t"
ffdshow-tryouts 1966 x64 build:
Download (http://www.dump.ro/download.php?file=ffdshow_rev1965_20080520_x64.7z&id=MDE5OGNiMWRiZTA3MGZhM2RhZTY2YTBkMjUyZmYxMTA=&download=true)

[chros]

I've used a MKV file with MP3, AAC and Ogg Vorbis audio tracks. I've used too a MKA file generated by me with three audio tracks: one in FLAC (8 channels), one in TTA (8 channels) and one in WavPack (8 channels). If I load the original files, one by one, independently that are or not inside a MKA, there are no problem, but if I use them in an only file (the MKA I mentioned), FFDShow won't work in audio changing. The only exception is that audio tracks have the same encoding; then, FFDShow continues loaded and are heared correctly, with the downmixing done properly.

I don't know if this info may help, but I'll say it: I'm using latest Haali Media Splitter, and I'm using too latest build of FFDShow tryouts.
Are you sure that the internal mkv splitter is disabled in MPCHC/MPC ? (so you're using Haali's splitter ?)

I have tried with mkv+x264+aac+ac3 and I have set the corresponding internal audio decoders in MPCHC, and set "uncompressed" to "all supported" in ffdshow: some strange thing's happening, but I can switch the audios back-and-forth.

But if I disable the the built-in audio decoders and ffdshow is responsible for the audio decoders, everything is fine with Haali's splitter but NOT with the built-in splitter (same effect when I said "strange") !!!
See this post: http://forum.doom9.org/showthread.php?p=1124285#post1124285

So, my suggestion is that try to find the corresponding audio decoders within ffdshow (if they exist in it) ...

[juanchu]

OK, chros, thank you so much.

I'm totally sure that I'm using Haali Media Splitter (icon is loaded in sistem tray).

This is what has happened when I've activated FFDShow's filters of codecs with >2 channels support (Ogg Vorbis, FLAC, TTA and WavPack), and whose decoders I'm using haven't downmixing support, the ones that are importante to me for being heared correctly:

-If I activate Ogg Vorbis support, the decoding is correctly, even when I change audio track.

-If I activate FLAC support, the decoding is correctly, even when I change audio track.

-If I activate TTA support, doesn't work, it's heared horribly. However, is I use True Audio Decoder and FFDShow's downmixing, everything is OK, even when I change audio track, because True Audio Decoder's output is a PCM 8 Channels, and FFDShow will be forced to be loaded, because mi soundcard doesn't support that tipe of audio input (however, FFDShow's PCM 2 channels output works great).

-So, the problem comes with WavPack. I don't see that FFDShow has support for this, and with this codec happens what I've said: if it's loaded if first place, it works the downmixing, but if it's loaded in second, third, etc. place, downmixing won't be done, and I will only hear the first two channels (it doesn't happen same thing as TTA, because CoreWavPack's output is WAVE_FORMAT_EXTENSIBLE).

So, with this configuration, the problem is partially resolved, but problems will continue when secondary audio tracks are encoded with codecs with >2 channels support, FFDShow isn't used to decode them, only for doing the downmixing, and the decoders used to decode these tracks haven't downmixing support (I'm forced to use FFDShow with FLAC (I normally use madFlac decoder) and Ogg Vorbis (I normally use CoreVorbis decoder)).

So, as I said, problem is partially resolved, but if I get or do a file with secondary audio tracks in WavPack with >2 channels, I will be forced to make a MKV file whose default audio track is the one that isn't downmixed when I do the track's changing.

Well, I hope that this problem is seen by developers, and fixed. In ZoomPlayer, with changing audio track with this configuration, I got sometimes memory errors, so I think it's FFDShow's problem.

Thank you so much for your help!

[clsid]

The problem that you describe is a "known" problem. It has been discussed in the past in the Haali Media Splitter topic. I don't remember the details, but it has something to do with dynamically changing the DirectShow filter graph. That doesn't always go well when different decoders are used. It's not a specific ffdshow problem.

[juanchu]

Thank you for this interesting info, clsid!

Well, it seems that I'll have to wait until WavPack is supported in FFDShow. Then, this problem will "dissapear". Until then, I will use MKVmerge to generate a MKV or MKA whose main audio track is the WavPack's one.

Thank you!

[chros]

Thank you so much for your help!
I'm glad that I could help.

Ask haruhiko_yamagata for supporting wavpack, he's "responsible" for the audio part :) He's very smart and very kind !!!

And welcome here at the doom9 community ! :)

[leeperry]

I was wondering, if I resize 1280*544 to 1280*768 and keeping AR, does ffdshow only add black borders or does it actually resize the picture ?

because if I increase the luma sharpen, it does make a difference........after all :)

otherwise is there a way to simply add borders up to 1280*768 w/o resize ?

[clsid]

FFmpeg does not have good WavPack decoding code last time I checked, and I am against adding functionality that doesn't work properly.

[juanchu]

Obviusly, clsid, is preferible waiting for until WavPack's support is better in FFMPEG (lossy, lossless and hybrid modes). So, I will continue using CoreWavPack, and will generate my own MKV with WavPack's track as the default one if I get one with this particularity (WavPack's track as secondary one, and >2 channels). It's not the best solution, but doesn't matter.

Thank you for all info that you're giving to me! I'm learning a lot of things!

[yangxi]

New GCC Quick Reference Guide for Optimized x86 Code Generation
http://developer.amd.com/assets/AMD_GCC_Quick_Reference_Guide080509.pdf

Does it any improve for PHENOM cpu?..

[Manao]

haruhiko_yamagata : YV16 is Y/V/U 422 planar. I422 is Y/U/V 422 planar. It's the same difference as YV12 / I420 (and, by inference, YV24/I444)

[ericab]

anyone know where i can get a pre-compiled ffdshow build 1967 ?

[cc979]

svn1967
http://www.speedyshare.com/721267853.html

[ericab]

svn1967
http://www.speedyshare.com/721267853.html

thanks cc979.

[chros]

FFmpeg does not have good WavPack decoding code last time I checked, and I am against adding functionality that doesn't work properly.
But I meant that we can see at the video config: eg. supporting xvid -> xvid and vc-1 -> wmv9 (external decoders ?)

Sorry if I'm misunderstanding something ...

[nautilus7]

I get glitches decoding this (http://www.sendspace.com/file/5421g7) H.264 video sample (it's from Chicken little blu-ray disc). The problem is only in the very first seconds. Both ffdshow and mpc-hc's internal decoder behave the same, while other decoders such CoreAVC and Cyberlink work fine. Is it an ffdshow problem or an ffmpeg one? Thanks.

[haruhiko_yamagata]

haruhiko_yamagata : YV16 is Y/V/U 422 planar. I422 is Y/U/V 422 planar. It's the same difference as YV12 / I420 (and, by inference, YV24/I444)
Thank you.
In ffdshow,
YV12 is FF_CSP_420P and
I420 is FF_CSP_420P|FF_CSP_FLAGS_YUV_ORDER.

So
YV16 should be FF_CSP_422P and
I422 should be FF_CSP_422P|FF_CSP_FLAGS_YUV_ORDER.

Unfortunately, in current ffdshow FF_CSP422P is I422.

I tried to change this, but it was too much work. I have added FF_CSP_YV16. It's working. I'll commit soon.

[lamer_de]

I know this is prbably not high on the priority list, but it would be really nice to have a slightly improved ssa/ass subtitle support. A rather important feature for me would be the ability to display 2 lines of text at once. Currently, whenever there are 2 lines displayed simultaneously, only the one that starts earlier gets shown. This is basically the case everytime a sign is displayed (which happens quite a lot, be it with fansubs or subtitles for "regular" movies).

What would also be nice to have is the support for the \fr (rotation) tag, which is also quite common for styled signs. I have no clue how much coding this would involve, but seeing that mplayer supports these things, I guess there might be code that is reusable(?)

[haruhiko_yamagata]

I know this is prbably not high on the priority list, but it would be really nice to have a slightly improved ssa/ass subtitle support. A rather important feature for me would be the ability to display 2 lines of text at once. Currently, whenever there are 2 lines displayed simultaneously, only the one that starts earlier gets shown. This is basically the case everytime a sign is displayed (which happens quite a lot, be it with fansubs or subtitles for "regular" movies). Please check "Subtitles->Text->Overlap subtitles with same time". Collisions are not supported though.

What would also be nice to have is the support for the \fr (rotation) tag, which is also quite common for styled signs. I have no clue how much coding this would involve, but seeing that mplayer supports these things, I guess there might be code that is reusable(?)
Yes, there is re-usable code. But I'm sorry, it's priority is low.

[Mr VacBob]

svn1967
http://www.speedyshare.com/721267853.html

It's not important, but the GPL shouldn't be used as a click-through license here. It specifically allows you to use ffdshow/whatever even if you don't agree to it.

edit: also, the proper fix for r1960 would be adding 'dst->mb_xy = src->mb_xy;' to copy_mb_to_context in h264.c. If that doesn't work, recalculate it as in the removed line in decode_mb_cabac.

[cc979]

It's not important, but the GPL shouldn't be used as a click-through license here. It specifically allows you to use ffdshow/whatever even if you don't agree to it.


Explain, do you mean - many ffdshow-tryout compiles avaible for testing, why post about that one ?

[Mr VacBob]

It was the last one linked - that applies to anything with the same installer.

[cc979]

i did not design it, i just compiled it from the sources - best take it up with the project leaders

[Mercury_22]

I have successfully cross-compiled libavcodec from win32 to win64 using MinGW GCC 4.4.0.
This had to be changed:
Index: idct_sse2_xvid.c
===================================================================
--- idct_sse2_xvid.c (revision 1965)
+++ idct_sse2_xvid.c (working copy)
@@ -338,7 +338,7 @@
"movdqa %%xmm6, 4*16("dct") \n\t" \
"movdqa "SREG2", 7*16("dct") \n\t"

-inline void ff_idct_xvid_sse2(short *block)
+void ff_idct_xvid_sse2(short *block)
{
asm volatile(
"movq "MANGLE(m127)", %%mm0 \n\t"
ffdshow-tryouts 1966 x64 build:
Download (http://www.dump.ro/download.php?file=ffdshow_rev1965_20080520_x64.7z&id=MDE5OGNiMWRiZTA3MGZhM2RhZTY2YTBkMjUyZmYxMTA=&download=true)

Hi ! link it's not working ! when I try to download it's an 798 bytes (ffdshow_rev1965_20080520_x64_[from_www.dump.ro].7z ) file which I can't open :confused:

[LoRd_MuldeR]

Hi ! link it's not working ! when I try to download it's an 798 bytes (ffdshow_rev1965_20080520_x64_[from_www.dump.ro].7z ) file which I can't open :confused:

Open the downloaded file in a text editor. You'll see:
<br />
<b>Warning</b>: filesize() [<a href='function.filesize'>function.filesize</a>]: stat failed for upload//ffdshow_rev1965_20080520_x64 [from www.dump.ro].7z in <b>/var/www/localhost/htdocs/vhosts/www.dump.ro/download.php</b> on line <b>48</b><br />
<br />
<b>Warning</b>: Cannot modify header information - headers already sent by (output started at /var/www/localhost/htdocs/vhosts/www.dump.ro/download.php:48) in <b>/var/www/localhost/htdocs/vhosts/www.dump.ro/download.php</b> on line <b>48</b><br />
<br />
<b>Warning</b>: readfile(upload//ffdshow_rev1965_20080520_x64 [from www.dump.ro].7z) [<a href='function.readfile'>function.readfile</a>]: failed to open stream: No such file or directory in <b>/var/www/localhost/htdocs/vhosts/www.dump.ro/download.php</b> on line <b>49</b><br />

Obviously a server problem :rolleyes:

[Shinigami-Sama]

go go phphail

looks like output buffering wasn't active/working/used when you tried to DL it

[_xxl]

Please wait new ffdshow-tryouts 32 and 64 bit will be available from SF.

[Mercury_22]

Please wait new ffdshow-tryouts 32 and 64 bit will be available from SF.

I've just installed ffdshow_rev1972_20080524_x64 Thanks for the great work! :thanks:

However I have a small problem with ffdshow_rev1972_20080524_x64's Audio decoder : on a vob file with 2 audio ac3streams 2Channel and 5+1Channels I can't hear anything all though in ffdshow's Audio decoder Volume "page" I can see levels moving for L, C, R, BackL, BackR and LFE ! Any Idea ?:confused:
(With MPC-HCx64's internal AC3 I can hear the 2 channels stream)

[clsid]

Please try an older x64 build to see if that gives the same problems.

[_xxl]

No just MinGW GCC x64 is broken, msvc x64 works.
I''ll update soon.

[draggoon01]

are there plans to release a stable build any time soon?

[clsid]

No just MinGW GCC x64 is broken, msvc x64 works.
I''ll update soon.
Ok, so the problem is unrelated to your code changes from rev1972? Just some GCC bug?

[_xxl]

Don't know yet, before 1972 liba52 couldn't be compiled by MinGW GCC 4.4.0 for x64.
Will update soon.

[arfster]

HI folks. Quick suggestion: how about an auto mode for bt601/709? The method Haali renderer uses makes sense, based on horizontal res (if above 1024 use bt709, if below bt601).

[3ngel]

It seems that the YV16 problem is corrected in the last build.
Thank you very much.

[Atak_Snajpera]

Can we expect e-ac3 support in near future?

[clsid]

From what I read on the FFmpeg mailing list, it is finally going to be commited to FFmpeg SVN in the very near future. After that we will attempt to add it to ffdshow as well. xxl recently already tried to add AC3/EAC3/DTS support to ffdshow's libavcodec, but that didn't work properly yet. So it isn't going to to be a trivial update.

[iron2000]

HI folks. Quick suggestion: how about an auto mode for bt601/709? The method Haali renderer uses makes sense, based on horizontal res (if above 1024 use bt709, if below bt601).

I'm doing this now with auto-load profiles as previously suggested by leeperry.

[nnever2000]

are there plans to release a stable build any time soon?

Good question, I'm still with last stable: ffdshow_rev610_20061201_clsid.exe

[_xxl]

That is an old buggy build, please update!

[compunett]

Indeed Justin Ruggles who work on the EAC3 support is doing the last adjustment before committing. I don't know how much the intricate things of ffmpeg and directshow can infer on this merge, or even if the difficulty is tied to that, but I guess the author can probably help you out.

[nnever2000]

That is an old buggy build, please update!

There is no other stable release!

Can the tryouts be considered stable?

Thanks

[Eragon4ever]

The beta builds are labeld "stable" because they are tested for a while before release. Latest beta ist beat4 (rev1723).

But I consider even the nightlys as stable - never had any problems.

[nnever2000]

The beta builds are labeld "stable" because they are tested for a while before release. Latest beta ist beat4 (rev1723).

But I consider even the nightlys as stable - never had any problems.

Ok I will replace the 610 version with 1723.
Thanks for help

[Atak_Snajpera]

Ok I will replace the 610 version with 1723.
Thanks for help

Don't pay attention on 'unstable' label. Use latest revision.

[Shinigami-Sama]

Don't pay attention on 'unstable' label. Use latest revision.

+1
that 'stable' release stuff is BS and shouldn't be applied like a blanket or you end up like debian

this is an example of a package that its latest builds are stable, often more so than the older builds!

[nnever2000]

Ok, ok, just updated with the last

ffdshow_rev1980_20080602_clsid.exe

for now, no problem at all.

Thanks, the h264 decoder seem much better.

[Atak_Snajpera]

I have tiny request. Could you clsid and xxl add revision number in registry? This would help me to detect installed FFDShow revision on users pc.

[clsid]

The installer already does that :)

HKLM\Software\GNU\ffdshow, dword: revision

[Atak_Snajpera]

OK. Thanks a lot. I was looking in Current_user :)

[nnever2000]

Ok, ok, just updated with the last

ffdshow_rev1980_20080602_clsid.exe

for now, no problem at all.

Thanks, the h264 decoder seem much better.

Only a little bug: on the vfw decoder the DIVX3 combo is blank and not "libvcodec" or "disabled"

[clsid]

It works fine for me.

[fastplayer]

It's blank for me too. Both on XP32 and XP64 with rev1980.

[leeperry]

anyone knows of a way to do realtime IVTC in ffdshow ?

I've tried with TIVTC in the avisynth section, but I get 5 fps(I have a 3.5Ghz C2D) and Reclock still says 29.97 fps on a mpeg2 .ts file

and if I only check "3:2 pulldown" in the avisynth section, I get deinterlacing artefacts.

The only way I've found so far is to output to OVERLAY and check "pulldown detection" in my ATi drivers.

then it seems pretty OK in 48Hz......yet Reclock still shows 29.97fps

is there a way to get a really good IVTC algorithm in ffdshow, and then output 23.976 fps to Reclock...and be able to use HR ?

I don't feel like reencoding the damn thing to x264 :(

or should I add a "pulldown" flag in the .ts file ?!

[LigH]

Possibly already mentioned - but because a user did a few tests and reported interesting details (http://forum.gleitz.info/showthread.php?p=368641):

ffdshow crashes using its H.264 VfW codec, at the frame with the number of consecutive B-frames (no B-frames: at Frame 0).

An out-of-bounds memory access (access violation) occurred in module 'ff_x264'...
...while compressing frame 5 from 03e40020 to 03430020 using codec "ffdshow Video Codec" (VideoSequenceCompressor.cpp:618)...
...while running thread "Processing" (thread.cpp:150).

The last working build may be from November 2007 (1625).

[Dark Shikari]

Possibly already mentioned - but because a user did a few tests and reported interesting details (http://forum.gleitz.info/showthread.php?p=368641):

ffdshow crashes using its H.264 VfW codec, at the frame with the number of consecutive B-frames (no B-frames: at Frame 0).In other words, it crashes the instant it starts encoding.

It would be useful if someone could pinpoint the specific x264 revision at which this occurred.

If they're using MSVC to compile FFDShow-tryouts, that could be part of the problem, because MSVC doesn't align the stack correctly, resulting in a crash.

[_xxl]

I always use MinGW GCC to compile ff_x264.dll.
4.2.3 for 32 bit and 4.4.0 for x64.

[Kurtnoise]

A small thing is missing in the current code I think...

Index: src/simd.h
===================================================================
--- src/simd.h (revision 1986)
+++ src/simd.h (working copy)
@@ -119,7 +119,7 @@

static __forceinline void movlhps(__m128i &dst,const __m128i &src) {dst=_mm_castps_si128(_mm_movelh_ps(_mm_castsi128_ps(dst),_mm_castsi128_ps(src)));}
#else
- static __forceinline __m128i _mm_castps_si128(__m128 &src) {return (__m128i&)src;}
+ static __forceinline __m128i _mm_castps_si128(__m128i &src) {return (__m128i&)src;}
static __forceinline void movlpd(__m128i &dst,const void *src) {(__m128d&)dst=_mm_loadl_pd((__m128d&)dst,(double*)src);}
static __forceinline void movhpd(__m128i &dst,const void *src) {(__m128d&)dst=_mm_loadh_pd((__m128d&)dst,(double*)src);}
static __forceinline void movlpd(void *dst,const __m128i &src) {_mm_storel_pd((double*)dst,(const __m128d&)src);}


otherwise it returns several errors during compilation.

[clsid]

Isn't stack alignment a general problem on Windows, even with GCC?

Using "__attribute__((force_align_arg_pointer))", like we do in libavcodec, could perhaps solve this problem. Or does x264 already use that?

[Dark Shikari]

Isn't stack alignment a general problem on Windows, even with GCC?

Using "__attribute__((force_align_arg_pointer))", like we do in libavcodec, could perhaps solve this problem. Or does x264 already use that?x264 deals with three types of alignment:

1. Stack alignment of variables, done using DECLARE_ALIGNED (an internal macro aliasing the appropriate gcc/msvc alignment option). This is done to up to 16-byte alignment levels.

2. Heap alignment, done by doing a malloc and then offsetting the pointer to the appropriate aligned starting location. This is done up to 64-byte alignment levels.

3. Stack alignment, used to simplify and speed up SSE2 deblocking assembly. This makes the assumption that not only can the compiler obey the order to align the stack, but that it conserves alignment between function calls.

[Jeremy Duncan]

ffdshow_source()
global idx1 = 10 # global hint by IanB
MT("""
idx1 = idx1 + 1
# different threads for top and bottom half of frame must have different idx (trick by Foxishadis)
backward_vec = MVAnalyse(blksize=16, isb = true, chroma=false, pel=1, searchparam=1, idx=idx1)
forward_vec = MVAnalyse(blksize=16, isb = false, chroma=false, pel=1, searchparam=1, idx=idx1)
last.MVFlowFps(backward_vec, forward_vec, num=2*FramerateNumerator(ffdshow_source()), \
den=FramerateDenominator(ffdshow_source()), mask=0, idx=idx1)
""",2) # number of threads.

This is my script in ffdshow avisynth.
However, it uses a framerate of 23.976 from ffdshow_source.
So, by doubling the frame rate, I get 48fps on my ntsc mpeg-2 movies that have pulldown.

What I need is to be able to set the fps in ffdshow_source()
Example: ffdshow_source(fps=29.970).

Maybe have a pal, or ntsc option.
ntsc having pulldown for a 29.970 fps.

I'm hoping this can be done. :)

[_xxl]

Confirmed, ff_x264 crashes with SSE2 cpu's and Win32.

[nnever2000]

Just a question: when I update FFDSHOW I have to uninstall previous version and then install or I can install directly over prevoius installation?
I ask because ffdshow installer ask me to remember settings so I think that I can update without uninstalling.

[Px]

>I can update without uninstalling
Yes

[clsid]

I ask because ffdshow installer ask me to remember settings so I think that I can update without uninstalling.I doesn't show such a message. Are you sure you are using a recent version?

[nnever2000]

I doesn't show such a message. Are you sure you are using a recent version?

I'm sorry. I see it in the matroska splitter installer (I installed both in short time).

For the future I will try to connect my brain before posting :p

[georgevalkov]

I use ffdshow encoder to produce RAW H264 streams: Store frames to external file [RAW frames].
I noticed that if two RAW streams are encoded with the same settings e. g. [file1.264] and [file2.264]; then one of the streams [file2.264] can be appended to the other [file1.264] with a HEX editor; then mp4box.exe can be used to create a MP4 container for the merged stream; the [result.mp4] will play both in MPlayer and DirectShow using the ffdshow decoder. No transition is noticed.

No transition or artifacts are noticed durring the frame by frame analysis of [result.mp4]. Audio and Video remain in sync.

Question:
Is it a good or a bad practice to combine raw H264 streams in binary? Or is there a recommended way to append frames from different encoding sessions?



I also use this manner to resume encoding after a power failure. First I check the number of the last encoded frame in stream [file1.264] and then continue the encoding from the next frame to [file2.264].

No transition or artifacts are noticed durring the frame by frame analysis of [result.mp4]. Audio and Video remain in sync. The only difference here is a warning displayed in MPlayer:
[h264 @ 00D4FB70]error while decoding MB 31 30, bytestream (td)
[h264 @ 00D4FB70]concealing 282 DC, 282 AC, 282 MV errors


Thank You for any advice!

[georgevalkov]

Confirmed, ff_x264 crashes with SSE2 cpu's and Win32.

I use the ffdshow encoder to render raw H264 frames. My computer is CPU: AMD Athlon XP 1700+; Windows 2003-SP2. Works fine.
ffdshow-rev1971_20080523_clsid_sse_icl10.exe

I tried to run an encoding job on my friend's computer: AMD64 Sempron 3100+; Windows XP 64 bit, and Virtual Dub crushed.
ffdshow-rev1943_20080416_clsid_sse_icl10.exe

Are we talking about the same problem?
Any known workarounds so far?

[Dark Shikari]

I use the ffdshow encoder to render raw H264 frames. My computer is CPU: AMD Athlon XP 1700+; Windows 2003-SP2. Works fine.Athlon XP doesn't support SSE2 ;)

[akupenguin]

Is it a good or a bad practice to combine raw H264 streams in binary? Or is there a recommended way to append frames from different encoding sessions?
If you abort an encode with ^C, then you can simply resume encoding at the next frame and concatenate the files. But in the case of a power failure that's not necessarily enough, as there may be a fraction of a frame at the end. Open the first file in a hex editor, delete the last startcode and everything after it, then count frames and result encoding.

[georgevalkov]

Athlon XP doesn't support SSE2 ;)

I guess for once I'm being lucky on that. :cool: By the way, on my friends AMD64 Sempron 3100+ (that supports SSE2), the crush occurs at the first frame, and the RAW [output.264] is an empty file (0 bytes).



If you abort an encode with ^C, then you can simply resume encoding at the next frame and concatenate the files. But in the case of a power failure that's not necessarily enough, as there may be a fraction of a frame at the end. Open the first file in a hex editor, delete the last startcode and everything after it, then count frames and result encoding.

Yes, after a power failure there is a fraction of an unfinished frame in the end. I guess that's why MPlayer warns me. Luckily it ignores the garbage and continues playing.

Please tell me how to find the start code of a frame? How does it look like in the HEX editor? :thanks:

AM 2:14 2008-6-09 update:
00 00 01 -- Is this the marker? Truncating the file from the beginning of this removed the warning in MPlayer.
0FB5D670 3C A2 3D 3C EE F8 24 14 5D AF 02 85 0E 0D F8 7A
0FB5D680 81 00 00 00 01 01 9E 15 0F 3A 24 B2 FF EC 43 06

[akupenguin]

00 00 01 -- Is this the marker?
yes .

[_xxl]

I tried to run an encoding job on my friend's computer: AMD64 Sempron 3100+; Windows XP 64 bit, and Virtual Dub crushed.
Can you test my 1975 x64 build please?
That sse2 crash should be only in Win32?

[3ngel]

There is a strange problem with FFDshow and Sonic HDDVD Decoder v4.3.0 (VC-1 Video only inside MKV)

With VMR it does work ok

http://img66.imageshack.us/img66/1082/clipboard01lh3.png

But when connecting it to FFDShow we have

http://img66.imageshack.us/img66/4732/clipboard02ai8.png

What's the matter?

Thanks

[leeperry]

anyone knows why randomly Reclock says "no video stream found" when using AVS scripts :(

I've tried with LimitedSharpen and LimitedSharpenFaster.

they both look really good, but randomly it forbids Reclock from getting the frame rate info :(

If I disable the avisynth section of ffdshow, then everything's fine

[Mercury_22]

I can't play this mpeg4 (DX50) (http://www.dump.ro/download.php?file=Testmpeg4.avi&id=MzBmNzhiNzhlYjZiOGQ3ZGU5OGJiZjkzODMxMTk2MmI=&download=true) (OR Try this server (http://rapidshare.com/files/121074851/Mpeg4Test.avi.html)) file with ffdshow (rev 1975-1993 tested) and WMP but with DivX and WMP I can ! If I use ffdshow (rev 1975-1993 tested) and MPC-HC it's working only with VMR9 !
Any idea what's wrong ? :helpful:

My OS Vista Ultimate x64 SP1

[georgevalkov]

Can you test my 1975 x64 build please?
That sse2 crash should be only in Win32?

Hello _xxl! I'll try my best to help You. Tell me, if You need anything else! Please note that I prefer not to abuse with my Administrator rights on my neighbour friend's computer. That means: I will install and run programs, but I prefer not to grab the local screen, so I will use Telnet to run or debug programs.


copy /Y VirtualDubMod.jobs "\\192.168.192.8\c$\Program Files (x86)\Virtual Dub Mod"
copy saferun.exe \\192.168.192.8\c$\WINDOWS
copy ffdshow_rev1975_20080526_x64.exe \\192.168.192.8\d$
copy test.mpg \\192.168.192.8\d$

telnet 192.168.192.8

cd /d d:\
ffdshow_rev1975_20080526_x64.exe /silent
cd /d "c:\Program Files (x86)\Virtual Dub Mod"
saferun "VirtualDubMod.exe /r /x" idle

The entire video material was encoded successfully. :thanks:
MPEG-2; PAL 720x576; progressive 25 fps; 29 seconds.
VirtualDubMod is version 1.5.10.2 build 2542/release
saferun was used to execute VirtualDubMod with idle CPU priority.


Finally I ran MPlayer, to verify the output. (seems ok)
MPlayer 1.0rc2-4.2.1 (C) 2000-2007 MPlayer Team
CPU: AMD Athlon(tm) (Family: 6, Model: 8, Stepping: 1)
CPUflags: MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 1 SSE2: 0
Compiled with runtime CPU detection.

Playing \\192.168.192.8\d$\test.264.
H264-ES file format detected.
FPS seems to be: 25.000000
[gl] using extended formats. Use -vo gl:nomanyfmts if playback fails.
==========================================================================
Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family
Selected video codec: [ffh264] vfm: ffmpeg (FFmpeg H.264)
==========================================================================
Audio: no sound
Starting playback...
VDec: vo config request - 720 x 576 (preferred colorspace: Planar YV12)
VDec: using Planar YV12 as output csp (no 0)
Movie-Aspect is 1.36:1 - prescaling to correct movie aspect.
VO: [gl] 720x576 => 786x576 Planar YV12

[obieobieobie]

I can't play this mpeg4 (DX50) (http://www.dump.ro/download.php?file=Testmpeg4.avi&id=MzBmNzhiNzhlYjZiOGQ3ZGU5OGJiZjkzODMxMTk2MmI=&download=true) (OR Try this server (http://rapidshare.com/files/121074851/Mpeg4Test.avi.html)) file with ffdshow (rev 1975-1993 tested) and WMP but with DivX and WMP I can ! If I use ffdshow (rev 1975-1993 tested) and MPC-HC it's working only with VMR9 !
Any idea what's wrong ? :helpful:

My OS Vista Ultimate x64 SP1

I can play it just fine in almost every renderer I can choose in Zoomplayer 5 plus haali renderer. The only renderer it won't work with (no video, only audio) is the VMR9 renderless mode, but I think that is because of Zoomplayer and not the file. This was using ffdshow rev 1980. It works in VLC 0.8.6f too.

WinXP SP3 here, though.

[georgevalkov]

Can you test my 1975 x64 build please?
That sse2 crash should be only in Win32?

I remember, the last time I was on the local screen of that AMD64 computer, that both of these versions had no support for H264 encoding:
ffdshow_rev1943_20080416_clsid_x64.exe
ffdshow_rev1972_20080524_x64.exe

At that time this 32 bit version had support for H264 encodding, but crushed at the first frame.
ffdshow-rev1943_20080416_clsid_sse_icl10.exe

Today I installed this version silently from Telnet, it passed my encoding test and produced a valid H264 stream:
ffdshow_rev1975_20080526_x64.exe

[Mercury_22]

I can play it just fine in almost every renderer I can choose in Zoomplayer 5 plus haali renderer. The only renderer it won't work with (no video, only audio) is the VMR9 renderless mode, but I think that is because of Zoomplayer and not the file. This was using ffdshow rev 1980. It works in VLC 0.8.6f too.

WinXP SP3 here, though.

I don't know about Zoomplayer but VLC 0.8.6f it's NOT using ffdshow, only internal filters AFAIK !
Strange about Zoomplayer not playing in VMR9 renderless mode because MPC-HC iit's playing in VMR9 renderless mode but NO EVR ! Did you try WMP ?

Vista x64 SP1 here !

[georgevalkov]

There is a strange problem with FFDshow and Sonic HDDVD Decoder v4.3.0 (VC-1 Video only inside MKV)
But when connecting it to FFDShow we have
What's the matter?

First open ffdshow raw video filter's settings, and make sure that the proper video format is enabled under codecs. You can try setting the decoder field to "all supported".

Second, some filters will not connect to a pin that offers Aspect ratio or Interlace format. Try adding some intermediate that supports video format conversion. But as long as my tests show, ffdshow has no problems with this case.

Could it be that the decoder wants to negotiate an encrypted path to a secure video render and avoids connecting to filters like ffdshow? Try using the ffdshow video decoder to decompress the source.

I don't have this filter, and cannot test it for you. Good Luck!

[Mercury_22]

Hello _xxl! I'll try my best to help You. Tell me, if You need anything else! Please note that I prefer not to abuse with my Administrator rights on my neighbour friend's computer. That means: I will install and run programs, but I prefer not to grab the local screen, so I will use Telnet to run or debug programs.


copy /Y VirtualDubMod.jobs "\\192.168.192.8\c$\Program Files (x86)\Virtual Dub Mod"
copy saferun.exe \\192.168.192.8\c$\WINDOWS
copy ffdshow_rev1975_20080526_x64.exe \\192.168.192.8\d$
copy test.mpg \\192.168.192.8\d$

telnet 192.168.192.8

cd /d d:\
ffdshow_rev1975_20080526_x64.exe /silent
cd /d "c:\Program Files (x86)\Virtual Dub Mod"
saferun "VirtualDubMod.exe /r /x" idle

The entire video material was encoded successfully. :thanks:
MPEG-2; PAL 720x576; progressive 25 fps; 29 seconds.
VirtualDubMod is version 1.5.10.2 build 2542/release
saferun was used to execute VirtualDubMod with idle CPU priority.


Finally I ran MPlayer, to verify the output. (seems ok)
MPlayer 1.0rc2-4.2.1 (C) 2000-2007 MPlayer Team
CPU: AMD Athlon(tm) (Family: 6, Model: 8, Stepping: 1)
CPUflags: MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 1 SSE2: 0
Compiled with runtime CPU detection.

Playing \\192.168.192.8\d$\test.264.
H264-ES file format detected.
FPS seems to be: 25.000000
[gl] using extended formats. Use -vo gl:nomanyfmts if playback fails.
==========================================================================
Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family
Selected video codec: [ffh264] vfm: ffmpeg (FFmpeg H.264)
==========================================================================
Audio: no sound
Starting playback...
VDec: vo config request - 720 x 576 (preferred colorspace: Planar YV12)
VDec: using Planar YV12 as output csp (no 0)
Movie-Aspect is 1.36:1 - prescaling to correct movie aspect.
VO: [gl] 720x576 => 786x576 Planar YV12

Hi I'm just curious how did you manage tu use ffdshow's 64-bit version in Virtual Dub Mod which AFAIK it's 32-bit ? :helpful:
It's that a 64-bit encoding chain ?

Please post the links to 64-bit version of Virtual Dub Mod Thanks ! :thanks:

[obieobieobie]

I don't know about Zoomplayer but VLC 0.8.6f it's NOT using ffdshow, only internal filters AFAIK !
Strange about Zoomplayer not playing in VMR9 renderless mode because MPC-HC iit's playing in VMR9 renderless mode but NO EVR ! Did you try WMP ?

Vista x64 SP1 here !

Uhm, I just added VLC because it has it's own codecs just to show that it can't be the file's fault. It's most likely a renderer problem on your end.

The renderless problem is on my end, but I don't care since the haali renderer suits my needs perfectly.

btw, unless I am misunderstanding you, you do know that you have to configure ffdshow to enable decoding and when configured it works in any directshow-based player like zoomplayer and wmp, right?

[georgevalkov]

Hi I'm just curious how did you manage to use ffdshow's 64-bit version in Virtual Dub Mod which AFAIK it's 32-bit ? :helpful:
It's that a 64-bit encoding chain ?

Please post the links to 64-bit version of Virtual Dub Mod Thanks ! :thanks:

Well, I guess that the 64 bit version of ffdshow installer offers backwards compatibility support for 32 bit applications running on 64 bit computers. Ask _xxl, he is the author.

The important thing for me is that installing ffdshow_rev1975_20080526_x64.exe fixed the problem: the ffdshow encoder no longer crushes on the first frame, when encoding to a H264 stream. The encoding process with the 32 bit version VirtualDubMod went nice and easy after installing the 64 bit version of ffdshow. Cheers. :cool:

[iron2000]

Seems like the latest build 1993 has broken profile autoload?

[Mercury_22]

Uhm, I just added VLC because it has it's own codecs just to show that it can't be the file's fault. It's most likely a renderer problem on your end.

The renderless problem is on my end, but I don't care since the haali renderer suits my needs perfectly.

btw, unless I am misunderstanding you, you do know that you have to configure ffdshow to enable decoding and when configured it works in any directshow-based player like zoomplayer and wmp, right?

You misunderstood me, My bad , my question was why Divx can decode with any renderer / player and ffdshow NOT :helpful: ?

Well, I guess that the 64 bit version of ffdshow installer offers backwards compatibility support for 32 bit applications running on 64 bit computers. Ask _xxl, he is the author.

The important thing for me is that installing ffdshow_rev1975_20080526_x64.exe fixed the problem: the ffdshow encoder no longer crushes on the first frame, when encoding to a H264 stream. The encoding process with the 32 bit version VirtualDubMod went nice and easy after installing the 64 bit version of ffdshow. Cheers. :cool:

I wonder how it's that possible ? Can anyone explain to me please :helpful: ?
XXl asked me before you to test ffdshow_rev1975_20080526_x64.exe for encoding (64-bit chain), but I still can't manage to make VirtualDub-AMD64 to use ffdshow_rev1975_20080526_x64 to encode because unlike in the 32-bit chain where VirtualDub (32-bit) has ffdshow (32-bit) listed in the compression menu VirtualDub-AMD64 doesn't have ffdshow (64-bit) listed in the compression menu ? :helpful:

[leeperry]

anyone knows why randomly Reclock says "no video stream found" when using AVS scripts :(

I've tried with LimitedSharpen and LimitedSharpenFaster.

they both look really good, but randomly it forbids Reclock from getting the frame rate info :(

If I disable the avisynth section of ffdshow, then everything's fine

to get back on that matter, if I add the "levels" filter before the AVS script, then Reclock always gets the frame rate information.

bug ?! :(

also, the LimitedSharpenFaster AVS script doesn't work if the width is not a multiple of 4, so 352*240 videos can't be processed.

but if I add the "Resize" filter in 1280*768 before the AVS script, it outputs 1024*768 for 1.33 AR files, but for 352*240 it outputs 1126*766 even if I force 1.33 AR :(

that's too bad you can't choose a mandatory width in the ffdshow resize :(

[clsid]

The x64 builds can only be used in 64-bit applications. The x86 builds can only be used in 32-bit applications.

[clsid]

Seems like the latest build 1993 has broken profile autoload?
Please double check that, because there have been no code changes related to that functionality.

[georgevalkov]

I wonder how it's that possible ? Can anyone explain to me please :helpful: ?
XXl asked me before you to test ffdshow_rev1975_20080526_x64.exe for encoding (64-bit chain), but I still can't manage to make VirtualDub-AMD64 to use ffdshow_rev1975_20080526_x64 to encode because unlike in the 32-bit chain where VirtualDub (32-bit) has ffdshow (32-bit) listed in the compression menu VirtualDub-AMD64 doesn't have ffdshow (64-bit) listed in the compression menu ? :helpful:

64-bit applications cannot link to 32-bit dynamic link libraries (DLLs). AX filters are DLLs. So if you use ffdshow in a 64-bit application, then it is using the 64-bit version of ffdshow. And if You use ffdshow in a 32-bit application on a 64-bit machine, then it is using the 32-bit version of ffdshow, which probably comes with the 64-bit installer for compatibility with 32-bit applications.

[Mercury_22]

64-bit applications cannot link to 32-bit dynamic link libraries (DLLs). AX filters are DLLs. So if you use ffdshow in a 64-bit application, then it is using the 64-bit version of ffdshow. And if You use ffdshow in a 32-bit application on a 64-bit machine, then it is using the 32-bit version of ffdshow, which probably comes with the 64-bit installer for compatibility with 32-bit applications.

Exactly my point ! Then how did you manage to use a 64-bit version of ffdshow with Virtual Dub Mod (a 32-bit app AFAIK) ? :helpful:

Didn't you in fact tested the 32-bit version of ffdshow and not the ffdshow_rev1975_20080526_x64 which XXL asked you to test ?

[Mercury_22]

Does ffdshow64 actually has a 64-bit VFW component ?

[leeperry]

I'd like to have 4 profiles, to overcome the problem that resize can't be done before the avisynth filter :

1080p
720p
SD 4:3
SD 16/9

I'm having problems to set the 2 last ones

I've tried if "x <1000" and "aspect=1.33" and check "on all conditions match", but it's always loading the SD 16/9 profile(using 1993)

any idea ?

TIA,

[iron2000]

Please double check that, because there have been no code changes related to that functionality.

Ok, found the cause.
Its the recent changes I made to my profiles, I corrected it and all is smooth again.:)

[clsid]

Does ffdshow64 actually has a 64-bit VFW component ?Yes. If it works? I dunno, haven't got a 64-bit OS.

Profiles are selected using the information that is available before decoding starts. At least afaik. So it can only use aspect ratio info from the container, not from the bitstream itself.

[obieobieobie]

You misunderstood me, My bad , my question was why Divx can decode with any renderer / player and ffdshow NOT :helpful: ?


My guess is some kind of color conversion issue. Check what ffdshow is configured to output.

[Mercury_22]

Yes. If it works? I dunno, haven't got a 64-bit OS.

Profiles are selected using the information that is available before decoding starts. At least afaik. So it can only use aspect ratio info from the container, not from the bitstream itself.

But it doesn't seem to work or can anyone tell me what I'm doing wrong ? :helpful: More details here

http://forums.virtualdub.org/index.php?act=ST&f=4&t=15774&

My guess is some kind of color conversion issue. Check what ffdshow is configured to output.

ffdshow is configured correct ! At your side does it work with WMP ?

[LoRd_MuldeR]

I always use MinGW GCC to compile ff_x264.dll.
4.2.3 for 32 bit and 4.4.0 for x64.

Which revision of x264 do you compile? And are there any special modifications needed to get the DLL working?
Apparently latest x264 doesn't compile as DLL, neither with MSVC nor with MinGW GCC ...

We also can't get the ff_x264.dll from latest ffdshow-tryouts to work with Avidemux for some reason :confused:
http://www.avidemux.org/admForum/viewtopic.php?pid=27541#p27541

Any help would be appreciated :)

[obieobieobie]

ffdshow is configured correct ! At your side does it work with WMP ?
Yes, I am watching the clip with WMP11 right now and it lists ffdshow as being the decoder.

[_xxl]

Which revision of x264 do you compile? And are there any special modifications needed to get the DLL working?
Apparently latest x264 doesn't compile as DLL, neither with MSVC nor with MinGW GCC ...
Any help would be appreciated :)

Just make a diff and you will see the diferenced from x264 and ff_x264. Also for latest ff_x264 you need to compile ffdshow.ax. I'll test later if I have time. Thanks!

[Mercury_22]

Yes, I am watching the clip with WMP11 right now and it lists ffdshow as being the decoder.

By any chance do you use Haali Media Splitter or ffdshow it's direct connected ?

[LoRd_MuldeR]

Just make a diff and you will see the diferenced from x264 and ff_x264. Also for latest ff_x264 you need to compile ffdshow.ax. I'll test later if I have time. Thanks!

Do you want me to make a diff between the the binaries or the sources?

I can tell you that the latest working libx264.dll that shipped with Avidemux was 438 KB in size, while the latest ff_x264.dll from ffdshow-tryouts is 707 KB in size :confused:

[_xxl]

I have compiled latest ff_x264. Tested with winxp and Amd XP and X2, both work fine. SSE2 is disabled, because MinGW or MSVC can't properly compile x264. Libavcodec had that same issue, but we added some custom code that fixed that crash. Libmpeg2 when SSE2 is used it crashes, so I disabled it.

[Dark Shikari]

I have compiled latest ff_x264. Tested with winxp and Amd XP and X2, both work fine. SSE2 is disabled, because MinGW or MSVC can't properly compile x264. Libavcodec had that same issue, but we added some custom code that fixed that crash. Libmpeg2 when SSE2 is used it crashes, so I disabled it.Why not just disable the specific problematic functions that your compiler can't handle? Slashing overall performance by 20-40% on a wide variety of popular CPUs is not exactly a good method of solving the problem...

[_xxl]

I know, but safety comes before speed.

[clsid]

Can't the problem be solved for MinGW GCC by adding the appropriate alignment attributes (from GCC 4.2+) to the crashing functions? Supporting MSVC is not important.
How do the regular x264 builds for Windows handle this problem?

[Dark Shikari]

How do the regular x264 builds for Windows handle this problem?There is no such problem in the regular builds; they work on everything from GCC 2.95 all the way up to 4.4 with no issues (as far as I've heard). This is why I have always suspected there must be some user error going on here because I have never seen any of these problems in any of the following:

1. Windows 32-bit executables
2. *nix 32-bit executables
3. *nix 64-bit executables
4. *nix 32-bit libraries, without PIC
5. *nix 32-bit libraries, with PIC
4. *nix 64-bit libraries, without PIC
5. *nix 64-bit libraries, with PIC

[_xxl]

Fix crash on PC's with SSE2 support from x264 vfw
http://x264vfw.svn.sourceforge.net/viewvc/x264vfw/trunk/codec.c?r1=7&r2=6&pathrev=7
or from ffdshow's libavcodec encoder:
http://ffdshow-tryout.svn.sourceforge.net/viewvc/ffdshow-tryout/trunk/src/ffmpeg/libavcodec/dsputil.c?r1=1214&r2=1213&pathrev=1214

[clsid]

That's libavcodec. But x264 isn't part of that. So similar fix should be added to TvideoCodecX264.cpp. Although that normally gets compiled with MSVC, so some kind of wrapper might be needed.

Are there many people (percentage wise) that encode with ffdshow? I vote for completely removing encoding support from ffdshow.

[_xxl]

I vote for completely removing encoding support from ffdshow.
All encoding support, ffdshow only for decoding?
Vfw is anyway outdated.

@clsid start a new svn branch with no encoding support?

[Dark Shikari]

All encoding support, ffdshow only for decoding?
Vfw is anyway outdated.

@clsid start a new svn branch with no encoding support?I would disagree; I find FFDshow's VfW encoders extremely useful; not the ones like x264, but rather the ones like FFVHUFF or FFV1, for which otherwise there is no VfW or DirectShow encoding interface.

[clsid]

There also is a DirectShow version of the encoder in ffdshow. I suggested removing encoding in the past, mainly since it is old, a bit buggy, and not maintained. There are plenty of alternatives for it imho. But there were a few people against it because they still used it. But on the other hand, those that wish to use this 'outdated' functionality can always stick to current/past versions.

I am curious as to how many people use it. Having a poll would probably not give a realistic representation though.

How do FFVHUFF and FFV1 compare to HuffYUV (for which a standalone codec exists)?

[LoRd_MuldeR]

Why kill the entire encoding support, just because one of the encoders currently has a problem?
You can simply disable the problematic decoder until the bug has been fixed. This way people can still use the working encoders :)

Also if x264 cannot be compiled as a DLL anymore, this is a serious problem! It would exclude all applications that use libx264.dll from future updates :(

[Shinigami-Sama]

I've used it about 4 times
mostly just to play with it

maybe just cut out the encoding part and leave it as a separate part of the project so anyone that wants to use it or fix it can?

then again I guess they could do that with the old version's source as well..

[Dark Shikari]

How do FFVHUFF and FFV1 compare to HuffYUV (for which a standalone codec exists)?FFVHUFF is both faster and better compression than HuffYUV, plus it has YV12 support.

FFV1 has no standalone comparable codec; while slower than Lagarith or HuffYUV it completely smokes them in compression ratio.

There are other useful codecs there too, such as MJPEG. VfW isn't inherently evil, especially with intra-frame-only codecs...

[clsid]

Why kill the entire encoding support, just because one of the encoders currently has a problem?That was never the reason for the suggestion. Everything currently works, with SSE2 disabled. The idea was to remove stuff that nobody uses or should use (because of better alternatives). But Dark Shikari has given some strong arguments for just keeping things the way they are :)

Although, in the case of H.264 encoding, there are plenty of alternatives for ffdshow.

Having SSE2 disabled for libmpeg2 is bad. So if no solution can be found for that, I suggest reverting to the older fully working code.

[Dark Shikari]

That was never the reason for the suggestion. Everything currently works, with SSE2 disabled. The idea was to remove stuff that nobody uses or should use (because of better alternatives). But Dark Shikari has given some strong arguments for just keeping things the way they are :)I would agree with removing x264 though, because it would be a great way to encourage people to stop using x264-in-AVI... :p

[_xxl]

First we should try to fix that bug and if no other solution is found remove encoders support.

[_xxl]

Having SSE2 disabled for libmpeg2 is bad. So if no solution can be found for that, I suggest reverting to the older fully working code.
That wasn't fully working:
1. No MinGW GCC with SSE2 support.
2. No x64 just plain C and couldn't be compiled by MinGW 64.
3. Wasn't updated to latest.

[Episode]

Ooh, almost 2000 revisions already. Keep up the good work guys!

[obieobieobie]

By any chance do you use Haali Media Splitter or ffdshow it's direct connected ?

Yes, I use haali but only for MKV and MP4, so when in WMP11, the file is played back using the default AVI splitter, VMR9 renderer and ffdshow decoder and ac3filter for the audio.

[georgevalkov]

My vote is to keep the encoding capabilities in ffdshow, and also to keep the H264 encoder. If the developers are too busy to update it, leave as it is or disable SSE2 for it, perhaps make it configurable if possible? but please don't remove it!


First we should try to fix that bug and if no other solution is found remove encoders support.

@_xxl
Count me on the test team. I'll help You with whatever I can to keep this feature. I have Visual Studio 2008 and WinDBG in hand. You can find me on this address:
the(underscore)boss(underscore)gfc(at)m s n(dot)com


I would agree with removing x264 though, because it would be a great way to encourage people to stop using x264-in-AVI... :p

@Dark Shikari
Please have a look at the attached images! The H264 stream encoded by ffdshow has two major ways to get a compatible container.
1. First is to connect ffdshow filter's output pin to another DirectShow filter like Haali Matroska Muxer that will add a compatible container MKV or MP4. Any DirectShow application encoding to MP4 or MKV can take advantage of this.
2. The second is to output the frames to a RAW stream file.264 and then add a compatible container for it. This can be done by MP4box.exe; mkvmerge.exe; meGUI or another tool.

I have been using both of the ways for a long time; encoded hundreds of H264/MP4 files, most of which by the second method, because I like using VirtualDubMod.

My capture program records in MPEG-2 format. I do understand that I can also use avisynth script+meGUI+x264.exe. Here are the cons: The MPEG-2 file has to be indexed first. After avisynth, every frame is a key-frame, the application (meGUI) will not be able to fast seek to the original key-frames. Decoding the mid-frames is extremely slow. So if I need to cut the advertising, I will waste a lot of time in seeking. After that the life is good. x264.exe will produce a bit smaller files than ffdshow, but will consume more CPU time.

This is why I use VirtualDubMod+ffdshow+MP4 container. And the dummy AVI that is created (is full of dropped frames when the output pin is disabled), well that's just a dummy AVI, so I delete it.

Another advantage of ffdshow is the built-in filtering. I can attach ffdshow filtering on the playback chain; then configure the desired filtering for the source file during a live playback. Finally copy the filtering configuration to the encoder's filtering page. This makes ffdshow the most powerful encoder. This makes it the desired encoder for many people. On the other hand, the logical choice for video format is H264, because it produces the best compression and high quality picture.

On my computer there are only Haali Matroska Splitter and ffdshow. No other codecs. I can play any format in any file and I can encode to any format I wish. MPEG-2 to H264/MP4 transcoding with filtering in a single pass. Its is the best all in one solution. Please don't remove this functions!



[Shell Extensions for RAW H264 steam files :: adds MP4 container]
264.reg.txt
1. Edit the text file to fix the path to MP4box.exe
2. rename to .reg (registry file) and add to the registry.

[Shell Extensions for RAW audio steam files :: encodes to AAC in MP4 container]
raw.reg.txt
1. Edit the text file to fix the path to faac.exe
2. rename to .reg (registry file) and add to the registry.

:thanks: Thanks for taking time to read my long posting! I understand that ffdshow consumes a large piece from Your time. That's simply because it is a great project. :)
George Valkov

[Mercury_22]

Yes, I use haali but only for MKV and MP4, so when in WMP11, the file is played back using the default AVI splitter, VMR9 renderer and ffdshow decoder and ac3filter for the audio.

I wasn't talking about VMR9, which it's working for me too
I can't play this mpeg4 (DX50) (http://www.dump.ro/download.php?file=Testmpeg4.avi&id=MzBmNzhiNzhlYjZiOGQ3ZGU5OGJiZjkzODMxMTk2MmI=&download=true) (OR Try this server (http://rapidshare.com/files/121074851/Mpeg4Test.avi.html)) ........it's working only with VMR9 !
Any idea what's wrong ? :helpful:

My OS Vista Ultimate x64 SP1

Let's say I wasn't clear so once again : VMR9 it's working, EVR it's NOT working with ffdshow but EVR it's working with DivX for the above file !

Clear now ? Can you test this please ? :thanks:

[LoRd_MuldeR]

My vote is to keep the encoding capabilities in ffdshow, and also to keep the H264 encoder. If the developers are too busy to update it, leave as it is or disable SSE2 for it, perhaps make it configurable if possible? but please don't remove it!

That's my position too. Better have an outdated/limited feature that might be updated at a later time than not having that feature at all.

Also it seems the SSE2 problem for the DLL has been found and a fix has been suggested:
http://forum.doom9.org/showpost.php?p=1148167&postcount=10

As far as I understand, the problem was that code outside of the DLL does not align the stack to 16...

[CiNcH]

Hey guys,

does anybody have the input format ffdshow expects in case of H.264 and VC-1? If you feed it with VC-1 and/or H.264 Elementary Stream, ffdshow outputs some weird colors and may crash.
Haali seems to serve ffdshow right by adding some special header.

Here is what RadScorpion said about it:

00 00 00 01 are special start codes that are used under the
ISO/IEC 14496-10 - Annex B (AVC/H.264) standard in several
situations like PES packet transmission. After the start code a single
NAL unit follows. The size of the NAL unit is then the difference
between two successive start codes.

MP4/AVC files use different way of storing the
NAL units. There is an avcC box that tells e.g. how many bytes
are used to signal NAL unit size. MP4 also does not
store SPS and PPS in the data itself but stores them separately.

[_xxl]

Also it seems the SSE2 problem for the DLL has been found and a fix has been suggested:
http://forum.doom9.org/showpost.php?...7&postcount=10
As far as I understand, the problem was that code outside of the DLL does not align the stack to 16...

I work to fix it, for libmpeg2 seems to work but more testing is required.
X264 still crashes, ffdshow is compiled by MSVC not GCC, so that fix can't be used outside x264.

[LoRd_MuldeR]

X264 still crashes, ffdshow is compiled by MSVC not GCC, so that fix can't be used outside x264.

Maybe a wrapper can be used?

[georgevalkov]

@CiNcH
This is what graphedit indicates on the property pages.
I am on my first steps in DirectShow programming, so my answer may not be correct, but I think that the source filter has to set this parameters in the Major Type+Sub Type+(maybe also Format), so that ffdshow knows that it should use the H264 decoder. If it tries to use another decoder, it is pretty normal to crush, since the decoder is fed with invalid data.

Major Type: Video

Sub Type:
{31435641-0000-0010-8000-00AA00389B71}
Format: avc1 704x576, 24 bits,
Aspect Ratio: 4x3
Interlace format: Frames
rcSrc=(0,0,0,0)
rcDst=(0,0,0,0)
CtrlFlags=00000000
^ The source filter is Haali Matroska splitter; H264 stream in MP4 container; video is PAL 704x576 25 fps interlaced.



@Mercury_22

Let's say I wasn't clear so once again : VMR9 it's working, EVR it's NOT working with ffdshow but EVR it's working with DivX for the above file !

Clear now ? Can you test this please ? :thanks:

I made a few tests for You. Results:
Plays with any of this renders: [EVR; VMR9; Default Video Render (overlay); Default Video Render (GDI); Haali Video Renderer]
I used revision 1971, May 23 2008, icl 10, x86, unicode, clsid
on Windows 2003 Server x86 with SP2 and .NET framework 3.5. I think that EVR on Win2003 came with this version of .NET. For Vista you don't need to install the .NET framework. I don't have a 64 bit machine to test it. I don't have DivX or any other codecs, because ffdshow alone is pretty much enough to decode anything.

PS: just to add a proof that it works, see the two screen-shots:
Jack-and-Jill-sample-EVR-render.png (http://www.mediafire.com/?tlwmrjmjdc3)
Jack-and-Jill-sample-info.png (http://www.mediafire.com/?dmgzx8qmdut)

[Mercury_22]

@CiNcH
This is what graphedit indicates on the property pages.
I am on my first steps in DirectShow programming, so my answer may not be correct, but I think that the source filter has to set this parameters in the Major Type+Sub Type+(maybe also Format), so that ffdshow knows that it should use the H264 decoder. If it tries to use another decoder, it is pretty normal to crush, since the decoder is fed with invalid data.

Major Type: Video

Sub Type:
{31435641-0000-0010-8000-00AA00389B71}
Format: avc1 704x576, 24 bits,
Aspect Ratio: 4x3
Interlace format: Frames
rcSrc=(0,0,0,0)
rcDst=(0,0,0,0)
CtrlFlags=00000000
^ The source filter is Haali Matroska splitter; H264 stream in MP4 container; video is PAL 704x576 25 fps interlaced.



@Mercury_22


I made a few tests for You. Results:
Plays with any of this renders: [EVR; VMR9; Default Video Render (overlay); Default Video Render (GDI); Haali Video Renderer]
I used revision 1971, May 23 2008, icl 10, x86, unicode, clsid
on Windows 2003 Server x86 with SP2 and .NET framework 3.5. I think that EVR on Win2003 came with this version of .NET. For Vista you don't need to install the .NET framework. I don't have a 64 bit machine to test it. I don't have DivX or any other codecs, because ffdshow alone is pretty much enough to decode anything.

PS: just to add a proof that it works, see the two screen-shots:
Jack-and-Jill-sample-EVR-render.png (http://www.mediafire.com/?tlwmrjmjdc3)
Jack-and-Jill-sample-info.png (http://www.mediafire.com/?dmgzx8qmdut)

A big Thanks :thanks: but for me it's not working -> must be my Video Intel G35 :(

[clsid]

A new build (rev. 2000) is now available at the sourceforge project page. The recent issues with libmpeg2/x264 updates should be pretty much resolved now, I hope.

Everyone please test the following:
* MPEG-2 decoding with libmpeg2, specially if performance is on par with older builds.
* Theora decoding, idem for performance.
* Theora encoding.
* x264 encoding.

If all is well, then a new beta will is likely to be released sometime at the end of the month.

[LoRd_MuldeR]

A new build (rev. 2000) is now available at the sourceforge project page. The recent issues with libmpeg2/x264 updates should be pretty much resolved now, I hope.

I'd like to test that file, but apparently none of the SourceForge mirrors can provide it :confused:


[EDIT]

Never mind, works now.

Everyone please test the following:
* MPEG-2 decoding with libmpeg2, specially if performance is on par with older builds.
* Theora decoding, idem for performance.
* Theora encoding.
* x264 encoding.

Yes, they all work fine here on my SSE2 capable processor :)

[_xxl]

Everyone please test the following:
* MPEG-2 decoding with libmpeg2, specially if performance is on par with older builds.
* Theora decoding, idem for performance.
* Theora encoding.
* x264 encoding.

Libmpeg2 will have SSE2 mc added soon.
Theora, libmpeg2 and x264 should only be compiled by MinGW GCC 4.2.x.

[georgevalkov]

A new build (rev. 2000) is now available at the sourceforge project page. The recent issues with libmpeg2/x264 updates should be pretty much resolved now, I hope.

Everyone please test the following:
* MPEG-2 decoding with libmpeg2, specially if performance is on par with older builds.
* Theora decoding, idem for performance.
* Theora encoding.
* x264 encoding.

If all is well, then a new beta will is likely to be released sometime at the end of the month.


PC1:AMD 32-bit [SSE:yes] [SSE2:no ] Athlon XP 1700+ overclocked to 1.89 GHz
results on PC1 are measured by process explorer and are accurate.
timing relays on CPU:total time.

PC2:AMD 64-bit [SSE:yes] [SSE2:yes] Sempron 3100+
PC2 was in use during this benchmarks; timing results may not be accurate;
timing relays on the log from VirtualDubMod.jobs file [end time - start time]

[MPEG-2 progressive file playback: PAL 720x576, 20 Mbit/s, length: 265 s]
s. ms
86.406 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10, YV12 [faster]
88.171 PC1 ffdshow_rev2000_20080612_clsid, YV12
117.812 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10, YUY2 [faster]
118.828 PC1 ffdshow_rev2000_20080612_clsid, YUY2
N/A PC2 is remote

[Theora progressive file playback: PAL 704x576, length: 60 s, PC1-2000-theora.avi]
s. ms
34.812 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10, YV12 [faster]
33.296 PC1 ffdshow_rev2000_20080612_clsid, YV12
41.484 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10, YUY2 [faster]
40.359 PC1 ffdshow_rev2000_20080612_clsid, YUY2
N/A PC2 is remote


[VirtualDubMod encode 1500 frames (1 minute) PAL 720x576 to 704x576 H264,
crop 16px from right to 704x576; one pass - quality 52; force RGB24]
s. ms
232.000 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10
result.264: 17 804 575 bytes
228.625 PC1 ffdshow_rev2000_20080612_clsid
result.264: 17 481 900 bytes [best compression; faster]
363.078 PC2 ffdshow_rev2000_20080612_clsid, computer is in use!
result.264: 17 633 037 bytes
245.594 PC2 ffdshow_rev1975_20080526_x64_xxl, computer is in use!
result.264: 17 947 672 bytes

[VirtualDubMod encode 1500 frames (1 minute) PAL 720x576 to 704x576 Theora,
crop 16px from right to 704x576; one pass - quality 52; force RGB24]
s. ms
165.781 PC1 ffdshow-rev1971_20080523_clsid_sse_icl10
result.avi: 22 530 048 bytes [faster]
187.375 PC1 ffdshow_rev2000_20080612_clsid
result.avi: 21 864 448 bytes
N/A PC2 ffdshow_rev2000_20080612_clsid, computer is heavy use! /gaming/
result.avi: 21 895 168 bytes [best compression]
N/A PC2 ffdshow_rev1975_20080526_x64_xxl, computer is in use!
result.avi: 21 895 168 bytes [best compression]

[Liisachan]

wow the millenium build :cool: ~ ffdshow_rev2000_20080612_clsid.exe

[clsid]

Summarizing georgevalkov results:
MPEG-2/Theora decoding performance is pretty similar. Small differences might be due to ICL10 vs Normal build.
Theora is slower. That is probably due to algorithmic changes, since output sizes are also different.
x264 is faster. That is likely due to code improvements.

So altogether it is good news.

[leeperry]

is there any filter in ffdshow than can improve the motion blur ?
or should I mess around with AVS scripts again ? :D

the LSF avs script looks so good, I'm really impressed :eek:

[Episode]

@clsid, could you please compile icl10 version of rev 2000?

[Momber]

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 using a standard 5.1 setup (3/0/2 + LFE) and have been able to reproduce this flaw using rev1971_20080523_clsid_sse_icl10 and rev1990_20080608_xxl.

Should I try a different build?
Is this a known problem or should I open a bug tracker for it?

Thanks,
S.

[iron2000]

Wow ffdshow-tryouts now breaking into 2000, feels like some big achievement to me.
Good job guys!

[MythCreator]

Just a question: How to compile it by use VS2008 under Windows Server 2008?

[Px]

I vote for completely removing encoding support from ffdshow.
I disagree, I've using ffvfw huffyuv for compressing raw video before processing in vurtualdubmod. Space saving in compare to single huffyuv is near 10 Gigs on 60 GB file....

[leeperry]

is there a way to convert BT.709 to 601 in ffdshow ? so I can output YUY2 to HR ? its BT709 coeffs are screwed up :(

I've tried the "picture properties" filter but it's not accurate enough.

and I'm still getting the frame rate randomly in Reclock when the Avisynth section of ffdshow is enabled.

sometimes Reclock says "no video stream found", I close the video/open it again and it's fine

if I disable the Avisynth script in ffdshow, it never happens.

[georgevalkov]

Yesterday I had desktop assess to the AMD64/XP64+SP1 computer so I made a few tests.

H264 encoding of a football match (length 1h38m), using VirtualDubMod (32 bit) and ffdshow_rev2004_20080613_clsid_sse_icl10 (32 bit) went okay.
I noticed that when encoding a football match in one pass quality mode, the AQ Strength should be set to 100. That way most of the details on the grass terrain are preserved and the file size is 20% smaller compared to the default value for AQ Strength=30, where most of the terrain's details are lost.


Then I installed the 64 bit version of VirtualDub 1.8 + ffdshow 64 bit (I tried both versions):
ffdshow_rev1975_20080526_x64 :: by _xxl
ffdshow_rev1975_20080525_clsid_x64
both don't have VFW component. Otherwise the ffvdub VirtualDub filter is in the list and I can add and configure it, I didn't test it though, because I wasn't prepared.

Then I installed the 64 bit versions of graphedt64.exe:
I was able to add ffdshow's DirectShow encoder to the graph; I noticed that some of the encoders are not available in the 64 bit build, including the H264. All the other (ffdshow filters Video and Audio decoders and procesors) returned an error and cannot be added to the graph:
The filter could not be created. Resources used by this filter may already be in use.
No such interface supported (Return code: 0x80004002)
(see the screen-shot:ffdshow64_rev1975 DS-filters not working.png) (http://www.mediafire.com/?uocngaiy0uw)

Anyway I don't care much for the 64 bit version. The most important thing is that the current 32 bit version is fixed and can encode H264 on processors that support SSE2.

[Momber]

The LFE Crossover bug is also present in rev. 2004 icl BTW. Just tried it.

[clsid]

The older 64-bit versions didn't include x264. Next build will have it.

Edit: found bug in installer that broke VFW on x64, now fixed.

[Dark Shikari]

The older 64-bit versions didn't include x264. Next build will have it.How? x264 doesn't support Windows 64-bit.

[clsid]

I thought that xxl somehow managed to make it compile on x64 at some point, but I may be wrong. Current revision doesn't compile.

[Dark Shikari]

I thought that xxl somehow managed to make it compile on x64 at some point, but I may be wrong. Current revision doesn't compile.x264 hasn't worked on 64-bit since sometime in 2005 (if ever), and even then it was probably just incidental.

You can get it working on 64-bit now if you munge a few of the #defines and completely disable all assembly optimizations :p

[LoRd_MuldeR]

Why does the assembler code in x264 depend on a specific OS? Since assembler code uses CPU instructions it's specific to a certain CPU architecture, that's for sure. But where does the OS interfere with the assembler code? As long as the assembler code runs on a suitable CPU it shouldn't have to care about the OS. Well, except for system calls maybe. So what is the difference between Win64 and Linux64 that makes x264 not work? And is there a chance it will ever be fixed? (and no, I'm currently not willing to pay akupenguin for that ^^ I just want to understand...)

[Dark Shikari]

Why does the assembler code in x264 depend on a specific OS? Since assembler code uses CPU instructions it's specific to a certain CPU architecture, that's for sure. But where does the OS interfere with the assembler code? As long as the assembler code runs on a suitable CPU it shouldn't have to care about the OS. Well, except for system calls maybe. So what is the difference between Win64 and Linux64 that makes x264 not work? And is there a chance it will ever be fixed? (and no, I'm currently not willing to pay akupenguin for that ^^ I just want to understand...)Because every operating system except Windows uses a certain 64-bit calling convention... while Windows uses a totally different one. And Akupenguin doesn't want to maintain the Windows 64-bit version of the calling code.

Fortunately, making it work on Windows 64-bit is much easier than with most programs because of x264's calling convention abstraction code, but it still requires someone with detailed knowledge of assembly and the Windows calling convention, along with a few hours spent poring over the current code and adding %ifdefs in some corner cases.

[Px]

Maybe because it's not assembler code, but intrinsic functions...

[Dark Shikari]

Maybe because it's not assembler code, but intrinsic functions...x264 doesn't currently use any assembly intrinsics. I've considered it, but the main problem is just that GCC sucks, and it generally doesn't compile intrinsics very well.

[LoRd_MuldeR]

Because every operating system except Windows uses a certain 64-bit calling convention... while Windows uses a totally different one. And Akupenguin doesn't want to maintain the Windows 64-bit version of the calling code.

I see. But calling conventions should only make a problem if the called function expects a different convention than the caller, right? So as long as all functions in x264 (including the assembly) agree on the same calling conventions, it should work fine. So why do we need to care about the OS' calling convention, as long as we don't want to call any system functions?

[Dark Shikari]

I see. But calling conventions should only make a problem if the called function expects a different convention than the caller, right? So as long as all functions in x264 (including the assembly) agree on the same calling conventions, it should work fine. So why do we need to care about the OS' calling convention, as long as we don't want to call any system functions?The Windows 64-bit calling convention is extremely weird and does some pretty bizarre things, so I'm not sure how possible that is.

The other reason is because you'd probably have to reprogram the compiler to use a certain calling convention for all externally called functions, and the *nix calling convention for all internal-only functions. I'm not sure if there's an easy way to do this with __fastcall or not.

[LoRd_MuldeR]

The other reason is because you'd probably have to reprogram the compiler to use a certain calling convention for all externally called functions, and the *nix calling convention for all internal-only functions. I'm not sure if there's an easy way to do this with __fastcall or not.

hmmm, I see. Thanks for info.

[squid_80]

x264 hasn't worked on 64-bit since sometime in 2005 (if ever), and even then it was probably just incidental.
It was working up until at least August last year. It was not at all incidental, I put a fair bit of effort into it and if you look back through the revisions you'll probably faintly recognize the start of the assembly abstraction (Even though your blog says "Written from scratch by Loren Merritt...").

LoRd_MuldeR: You can tell the C compilier what convention to use (fastcall, stdcall etc.) but you can't tell it how the ABI works; that is the problem here.
P.S. Windows will occassionally butt in between function calls and have it's way with the stack. I haven't figured out how or why, I just know the "red zone" is another luxury Linux has that Windows doesn't.

[Dark Shikari]

It was working up until at least August last year. It was not at all incidental, I put a fair bit of effort into it and if you look back through the revisions you'll probably faintly recognize the start of the assembly abstraction (Even though your blog says "Written from scratch by Loren Merritt...").Hmm, I got my info from Loren, so either he wasn't very clear or I missed it.

Though his commit message where he removed all Windows 64-bit-related code would disagree with you.

I suspect for more detailed info you're going to have to ping him to this thread...

[squid_80]

November 2007: http://mirror05.x264.nl/Cef/force.php?file=./x264_x64_r682.7z
I don't mind having the code removed; I was never willing to consistently maintain it, and now I can tell people who PM me about updating it to buzz off because my name's not on there.

[LoRd_MuldeR]

LoRd_MuldeR: You can tell the C compilier what convention to use (fastcall, stdcall etc.) but you can't tell it how the ABI works; that is the problem here.

Yes, I understand. But when calling your own (internal) functions you don't need to care about the OS' ABI. So my though was: Use whatever calling convention satisfies the OS for calling system functions and use your preferred calling convention for calling the internal functions. This way the calling convention used to call your (internal) assembly functions could always be the same, no matter which OS you are running on. So you wouldn't have to adjust the assembly for every OS (*nix vs. Win64). But from what I understand now, the compiler doesn't support that method...

[Dark Shikari]

Yes, I understand. But when calling your own (internal) functions you don't need to care about the OS' ABI.But what if your internal calling convention relies on certain registers being conserved during context switches--registers that the OS doesn't conserve?

[squid_80]

Yes, the C compiler uses the ABI for the target OS. If you're calling assembly functions from assembly functions, you can pretty much use whatever convention you want (as long as stack alignment is observed).

(Quick 'n' Dirty way to fix the "xmm6-xmm15 are non-volatile" problem on win64: write 2 assembly functions to save/restore them, wrap them around the asm function calls (like emms) and don't use floating point in between. It won't unwind properly but who gives a crap.)

The only registers win64 doesn't preserve over context switches are MMX registers in kernel mode.

[akupenguin]

But what if your internal calling convention relies on certain registers being conserved during context switches--registers that the OS doesn't conserve?
The OS has to preserve all registers, because context switches can occur anywhere in the code, not just at function boundaries. A register that isn't preserved on switches can't ever be used.

It was working up until at least August last year.
Sure there were patches floating around, but they weren't submitted to svn, and I don't care enough about win64 to pull patches from someone elses repository without input from the authors.

Win64 support was added in r336 (2005-10-25). The first subsequent revision to break it was r353 (2005-11-02) which used 10 xmmregs without saving any. Win64 was never committed again.

[leeperry]

I know there's little chance I'll get a reply but anyway :D

I understand the "deband" filter of ffdshow is actually the "gradfun2db" AVS script, which I prefer to use to run it multithreaded with MT.

anyone knows what AVS script the "unsharp masking" filter of ffdshow comes from ?

Thanks,

[Inventive Software]

You could :search: in the AviSynth forums. ;)

[leeperry]

well I know I can get unsharp masking with LSF....that's not the point..

I would like to get the exact same algorithm as ffdshow, but in AVS.

set to 6 it creates some very slight EE, that I find very appealing.

[Mercury_22]

Great improvements in x64 version now with ffdshow_rev2009_20080616_clsid_x64 I can see ffdshow in VirtualDub-1.8.1-AMD64's video compression menu and I can encode using "Libavcodec" (mpeg4, DivX3...) ! :thanks: (and it's so fast... :) ! .... only if x264 would been available in x64 :p :))

But when I'm using xvid for encoding VirtualDub-1.8.1-AMD64 it's crashing with this error message :
roblem Event Name: APPCRASH
Application Name: Veedub64.exe
Application Version: 1.8.1.0
Application Timestamp: 48556ca0
Fault Module Name: xvidcore.dll
Fault Module Version: 0.0.0.0
Fault Module Timestamp: 48555728
Exception Code: c0000005
Exception Offset: 0000000000088199
OS Version: 6.0.6001.2.1.0.256.1
Locale ID: 1033
Additional Information 1: 6687
Additional Information 2: 7aa12ef2e1d3af19ae1243cad32742e1
Additional Information 3: 595c
Additional Information 4: 5a3571e18b531ebc7ebe8615900c8fc0

[clsid]

Does xvidcore work properly for decoding?

[Mercury_22]

Does xvidcore work properly for decoding?

(Yes decoding it's OK !) = wrong :confused:

EDIT : I was wrong sorry NOT working I forget to change to xvid instead of Libavcodec when I tested

[_xxl]

I would like to know if any MinGW GCC x64 compiled:
libmpeg2
theora
liba52
libtds
libmad
faad
tremor
realaac
libavcodec
libmplayer
xvidcore
x264
are working or not?
Please test with VirtualDub64 and 64bit media player.
Thanks!

[Mercury_22]

I would like to know if any MinGW GCC x64 compiled:
libmpeg2
theora
liba52
libtds
libmad
faad
tremor
realaac
libavcodec
libmplayer
xvidcore
x264
are working or not?
Please test with VirtualDub64 and 64bit media player.
Thanks!

Encoding Using VirtualDub-1.8.1-AMD64 :

- libavcodec (mpeg4, divx 3...), theora are working
- xvidcore NOT working
- the others how should I test them since they are NOT in the ffdshow64 video encoder configuration (VFW configuration) ??? :mad: :helpful:

Decoding using MPC-HC x64 & WMP x64 :
- libavcodec ( all ), theora, liba52, libmad, mp3lib, libfaad2, realaac, tremor, are working
- libmpeg2, xvidcore NOT working

[_xxl]

libmpeg2, xvidcore NOT working
Are those crashing?
Did you tried theora decoder/encoder?
x64 libmpeg2, x264, xvidcore, libmplayer, liba52, libdts, faad, realaac, tremor will be compiled from plain C only.
Only libavcodec seems to work, but not 100% sure.
Please test all libavcodec codecs encoding and decoding.
So Win64 is not supported?

[Mercury_22]

Are those crashing?
Did you tried theora decoder/encoder?
x64 libmpeg2, x264, xvidcore, libmplayer, liba52, libdts, faad, realaac, tremor will be compiled from plain C only.
Only libavcodec seems to work, but not 100% sure.
Please test all libavcodec codecs encoding and decoding.
So Win64 is not supported?

NONE IS CRASHING (except xvid when encoding but it's mre probable to be VirtualDub-1.8.1-AMD64's fault look here (http://forums.virtualdub.org/index.php?act=ST&f=4&t=15774&)) just not decoding or encoding
Yes I've tried theora decoder/encoder and it's working !
Tested all ( AFAIK ) libavcodec codecs decoding, for all encoding later today

Also libdts later today because i don't have a sample ! Can anyone give me a link to a DTS sample ? Please

Update : libdts not working in X64 AND APPARENTLY NOT WORKING IN X86 TOO ! Can anyone confirm this (x86) !
I'm using ffdshow_rev2008_20080615_xxl for x86 and ffdshow_rev2009_20080616_clsid_x64 for x64

Update 2 tested "all libavcodec codecs encoding " and no problems !

[fastplayer]

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.
Can somebody confirm this?

[_xxl]

Please check that all libs are compiled by MinGW GCC x64, not MSVC.

[Mercury_22]

Please check that all libs are compiled by MinGW GCC x64, not MSVC.

I'm using ffdshow_rev2009_20080616_clsid_x64 So.. :confused: ask clsid because ther's nothing in the Release Notes (http://sourceforge.net/project/shownotes.php?group_id=173941&release_id=569930) :logfile:

Maybe someone can compile a new version and tell us what he used to compile all libs ! :script: :helpful:

Update : libdts not working in X64 AND APPARENTLY NOT WORKING IN X86 TOO ! Can anyone confirm this (x86) !

I'm using ffdshow_rev2008_20080615_xxl for x86 and ffdshow_rev2009_20080616_clsid_x64 for x64

[clsid]

I've used GCC for: libavcodec, libmplayer, libmpeg2, libtheora, xvidcore, tomsmocomp

[_xxl]

I will make new x86 and x64 builds today. Please test them.

[Mercury_22]

I will make new x86 and x64 builds today. Please test them.

Ok but can anyone confirm the DTS problem on x86 ? :logfile:

[_xxl]

No, it works with my samples, x86.

[Mercury_22]

No, it works with my samples, x86.

Can you upload a sample please!

[fastplayer]

DTS workin fine with rev2000.
Sample: http://www.mediafire.com/?39cbbpnb0gz

[Mercury_22]

DTS workin fine with rev2000.
Sample: http://www.mediafire.com/?39cbbpnb0gz

Your sample it's working for me too on both x86 and x64 :thanks:
Can you test my sample audio only DTS 5.1 wav (http://www.diatonis.com/downloads_dts_ac3.html) for me it's working only with AC3Filter or PowerDVD ( audio cd )

[_xxl]

Wav in dts was not working before, this is a know issue, doesn't work with mpc, only with latest ac3filter.

[fastplayer]

Can you test my sample audio only DTS 5.1 wav (http://rapidshare.com/files/123095978/testdts.wav.html) for me it's working only with AC3Filter or PowerDVD ( audio cd )
All I get is noise. VLC plays it back though.
I'm not sure if the wave parser correctly detects DTS... Did this sample work with previous ffdshow versions?

Edit: XXL cleared it up! :)

[Mercury_22]

All I get is noise. VLC plays it back though.
I'm not sure if the wave parser correctly detects DTS... Did this sample work with previous ffdshow versions?

I don't know i don't have previous ffdshow version and i'm not going to install old versions

Also i've update my link with new samples Also wavs

Wav in dts was not working before, this is a know issue, doesn't work with mpc, only with latest ac3filter.

Tested with WMP and PowerDVD (with powerdvd it's working as audio cd/dvd) also and not working so ffdshow it's not working not mpc :eek:

[Mercury_22]

ffdshow_rev2014_20080617_clsid_x64 tested and :
- xvid CAN decode and encode without any problem s! :) :thanks:
- libmpeg2 still NOT working :mad:

[clsid]

Xvid will probably be a bit slow, since the assembly optimizations have been disabled for x64 builds of xvidcore.

[Dark Shikari]

Xvid will probably be a bit slow, since the assembly optimizations have been disabled for x64 builds of xvidcore."A bit" is probably a severe understatement.

I'm not sure why one would want to include such an encoder without assembly; if anything you're just going to mislead people into thinking its slower than it actually is rather than advise them to download the standalone 32-bit version. Unless maybe you put a gigantic notice that says "THESE ENCODERS ARE REALLY SLOW AND HAVE NO ASM SO DON'T USE THEM"--but if so, what's the point of even including them?

[_xxl]

Only theora and libavcodec has been compiled for x64 with assembly, the rest is just C.
So there isn't no Win64 support for x264, xvidcore and libmpeg2.
Is there a plan to support Win64?

[Dark Shikari]

Only theora and libavcodec has been compiled for x64 with assembly, the rest is plain C.
So there isn't no Win64 support from x264, xvidcore and libmpeg2.
Is there a plan to support Win64?I would strongly object to including x264 without asm support, given that the speed would be absolutely abominable and it would reflect badly on the x264 project.

There is no plan to support Win64 unless someone steps up to maintain it.

[clsid]

I will remove it from my future installers. If I decide to make any that is, because I am getting really fed up with all the x64 crap that I don't even use myself.

[georgevalkov]

I will remove it from my future installers. If I decide to make any that is, because I am getting really fed up with all the x64 crap that I don't even use myself.

I agree with that. After all of the changes that You and _xxl made recently, I'd recommend you to have a rest and eventually make plans for the feature. :cool:

[Mercury_22]

With XXl's ffdshow_rev2014_20080617_x64 version libmpeg2 it's working now :thanks: and I even have H.264 in the encoding menu ..... to bad that it's not working do to lack of win x64 support in x264 :mad:
Anyway Thanks XXL and all other ffdshow's developers for your hard work ! :thanks:

[video_magic]

.....
Anyway Thanks XXL and all other ffdshow's developers for your hard work ! :thanks:

Amen! Thanks so much from Me also :)

[Anima123]

Big thanks for all who devoted in this great project.

[tiresias]

A couple of questions about the ffdshow mpeg2 decoder ...

When you use the Microsoft DV decoder, you can set its properties to "half resolution" which produces a half-size decoded image and reduces the load on the CPU. Is there a similar property that can be changed in the ffdshow mpeg2 decoder, so that high definition mpeg clips could be displayed in half resolution and with a reduced load on the CPU which would hopefully help stutter-free playback on lower powered PCs? If there is not such a setting, would this be a good addition?

High definition at full resolution is larger than most PC screen resolutions, so in most cases it would have to be displayed at less than it's full res on screen. In that case it seems a good idea to get hopefully stutter-free playback by decoding at half of full resolution.

The second question is do the ffdshow decoders provide access to properties via a DirectShow programming interface to the filter (instead of the properties dialog box).

[Mercury_22]

After more testing I found that liba52 it's not working in XXL 's ffdshow_rev2014_20080617_x64 and libmpeg2 it's not working in clsid's ffdshow_rev2014_20080617_x64 ! :confused:
Help please ! :helpful:

[georgevalkov]

A couple of questions about the ffdshow mpeg2 decoder ...

When you use the Microsoft DV decoder, you can set its properties to "half resolution" which produces a half-size decoded image and reduces the load on the CPU. Is there a similar property that can be changed in the ffdshow mpeg2 decoder, so that high definition mpeg clips could be displayed in half resolution and with a reduced load on the CPU which would hopefully help stutter-free playback on lower powered PCs? If there is not such a setting, would this be a good addition?

My compurer's CPU is Athlon XP1700+ overclocked to 1.89 GHz.

CPU usage with ffdshow for interlaced 20 Mbit/s MPEG-2 720x576 PAL 25 fps playback is 33% for both libavcodec and libmpeg2.

If you are lucky enough, a decoder that supports DXVA_ModeMPEG2 will be able to utilize the video hardware and reduce the CPU load. ATi All in Wonder 9600XT come with a DVD decoder called: CyberLink Video/SP decoder. CPU usage for the same media is less than 15%. Or if I force BOB deinterlacing and connect to VMR9, it will play at 50 fps and consume about 30% of CPU.

MPlayer consumes about 30% of CPU.
^ Opening video decoder: [mpegpes] MPEG 1/2 Video passthrough

full-HDTV [DivXTestCDv2.0.b/H_VIDEO_CODECS/VID-1E.mpg] 1920x1088p consumes less than 19% with the CyberLink decoder and DXVA enabled. Currently this is the only format, for which I don't use ffdshow.


High definition at full resolution is larger than most PC screen resolutions, so in most cases it would have to be displayed at less than it's full res on screen. In that case it seems a good idea to get hopefully stutter-free playback by decoding at half of full resolution.

When I want to play an HDTV move, I change the screen resolution to match the movie's and enjoy it.

The second question is do the ffdshow decoders provide access to properties via a DirectShow programming interface to the filter (instead of the properties dialog box).
I think that the answer is Yes, but I don't know anything else.
Good luck!

[clsid]

When you use the Microsoft DV decoder, you can set its properties to "half resolution" which produces a half-size decoded image and reduces the load on the CPU. Is there a similar property that can be changed in the ffdshow mpeg2 decoder, so that high definition mpeg clips could be displayed in half resolution and with a reduced load on the CPU which would hopefully help stutter-free playback on lower powered PCs? If there is not such a setting, would this be a good addition?That is not possible and this functionality will not be added.

The second question is do the ffdshow decoders provide access to properties via a DirectShow programming interface to the filter (instead of the properties dialog box).There is a very basic API for controlling some parts of ffdshow. Look at the source code or ask Albain (he implemented it). But for the most part of ffdshow, there is no way to control it programmatically. And there are no plans to add such functionality.

[Reimar]

[Concerning playing MPEG-2 at half resolution]

That is not possible and this functionality will not be added.


It is not in any way important to me, but I am curious why it is impossible? libavcodec itself does offer the functionality, at least for some codecs.

[clsid]

For libavcodec's mpeg-2 decoder there already is a "use speedup tricks" option in ffdshow. I don't know what that does precisely.


Generally speaking, at this moment, it is very unlikely that new functionality will be implemented in ffdshow, unless someone steps up and offers to do it. The current devs simply don't have the time, skills, or motivation to add new stuff. The only new functionality that is planned (afaik) is adding the AC3/E-AC3/AAC decoders from libavcodec. xxl is going to do another attempt to get them working with ffdshow.

[Leak]

It is not in any way important to me, but I am curious why it is impossible? libavcodec itself does offer the functionality, at least for some codecs.
Well, since all non-I-frames in MPEG2 fully depend on neighbouring decoded frames you can't just "decode it to half the size" and expect something watchable to come out of it.

And since you have to decode it at full resolution scaling it down afterwards only increases CPU load - and you can already do that using ffdshow's Resize & Aspect filter, anyway.

[Kurosu]

Well, since all non-I-frames in MPEG2 fully depend on neighbouring decoded frames you can't just "decode it to half the size" and expect something watchable to come out of it.

And since you have to decode it at full resolution scaling it down afterwards only increases CPU load - and you can already do that using ffdshow's Resize & Aspect filter, anyway.

This is somewhat funny: would you also explain to David Huffman, were he alive, what variable length coding is?

Distortion is indeed a side effect, but it does improve decoding speed. The mplayer man page for instance, and google, have explanations about the lowres decoding option found in lavc.

[Leak]

This is somewhat funny: would you also explain to David Huffman, were he alive, what variable length coding is?
Errr... so you're implying that David Huffman is to Variable Length Coding as Reimar is to MPEG2? :confused: (Sorry, I really genuinely don't understand that comment of yours...)

[LoRd_MuldeR]

Errr... so you're implying that David Huffman is to Variable Length Coding as Reimar is to MPEG2? :confused: (Sorry, I really genuinely don't understand that comment of yours...)

I think what he actually meant is:
You don't need to explain how MPEG-2 decoding works to Reimar, because as an MPlayer/Codec developer he should be very familiar with it...

[Leak]

I think what he actually meant is:
You don't need to explain how MPEG-2 decoding works to Reimar, because as an MPlayer/Codec developer he should be very familiar with it...
Ehehe... ummm, since he asked that question above I didn't consider that he could be an mplayer dev... :)

Thanks for clearing that up...

[Mr VacBob]

For libavcodec's mpeg-2 decoder there already is a "use speedup tricks" option in ffdshow. I don't know what that does precisely.

It (CODEC_FLAG2_FAST) uses a simpler dequantization step which results in more zero DCT coefficients. It's not that much faster, really.

What he asked for is controlled by AVCodecContext.lowres, which certainly works for MPEG2. I think it might already be exposed in the UI too.

[nxmehta]

Hi all,

There's a bug in ffdshow (reported at http://sourceforge.net/tracker/index.php?func=detail&aid=1924063&group_id=173941&atid=867360 ) that makes some of the software I'm running unusable. I was just wondering if this bug was fixed in any of the nightly builds? If not, is it something that might be fixed in the future?

Thanks.

[leeperry]

is it me, or is it only possible to resize horizontally in ffdshow ?!

[jmartinr]

is it me, or is it only possible to resize horizontally in ffdshow ?!

It's you :p. Horizontal and vertical resize works fine here. Maybe the problem is in your renderer.

[leeperry]

well you can keep the exact same width and only resize vertically ?

now how do you do that ?

I've got a profile to resize to 1024x768, yet sometimes it resizes to 1152x768 with SD content.....I want freaking 1024*xxx :(

[cyberbeing]

It works here as well. Use "Specify horizontal and vertical size" > only change the vertical size > make sure Aspect ratio is set to "no aspect ratio correction".

I'm assuming you're using YV12 and YUY2 with VMR/EVR? With those color spaces VMR/EVR will sometimes use odd internal resolutions (which seems to be a built-in restriction) yet it will still display at your desired resolution.

[clsid]

Maybe those are anamorphic encodes? Check your PAR settings. Because it should work.

[JarrettH]

think we'll have a new beta soon? :devil:

[clsid]

At the end of the month if no regressions are reported.

[LoRd_MuldeR]

At the end of the month if no regressions are reported.
Apparently the "Cubic Blending" deinterlacer doesn't deinterlace at all or it fails horrible to remove combing artifacts (which I think isn't possible with that method)

[Yong]

Guys, is it possible to make "custom" OSD display as list?
if the video resolution is small, OSD will got truncated if display all in one line like this
http://xs128.xs.to/xs128/08255/clipboard01263.png

i hope this feature request isn't too hard for devs :p

[mike7]

Guys, is it possible to make "custom" OSD display as list?

use \n for new line, ex.: %idimar\nRemaining time: %rtime

[Reimar]

Ehehe... ummm, since he asked that question above I didn't consider that he could be an mplayer dev... :)

Thanks for clearing that up...

Well, you started from the assumption that decoding at half resolution is not possible in a useful way. But the reality is, that even quarter resolution decoding is possible with (IMO) acceptable quality losses simply by setting a libavcodec flag. Half resolution in my tests reduces decoding CPU usage by 60%, quarter resolution by another 20% (compared to half res) for high bitrate HD material.
So my question was about the part I did not know: why it would not be possible to export this functionality via ffdshow (seems I was mistaken to assume you knew about this libavcodec feature).

[leeperry]

Apparently the "Cubic Blending" deinterlacer doesn't deinterlace at all or it fails horrible to remove combing artifacts (which I think isn't possible with that method)

yeah the deinterlacing section of ffdshow is very poor.

the "cubic blending" doesn't do anything whatsoever, and the "linear blending" is very blurry.

the best deinterlace I've found so far is the PS "blending" script in MPC...but it only works w/ EVR or VMR9

Seb.26 told me that it's using a very simple script, but because noone is willing to add new features to ffdshow......it's not gonna happen anytime soon.

ffdshow is so great, but it's got a few annoying bugs that noone will ever fix I guess :(

for instance....when using Avisynth scripts, very often Reclock says "no video stream found", I stop/press play again, and it's fine...

[clsid]

but because noone is willing to add new features to ffdshow......it's not gonna happen anytime soon.Comments like this ain't gonna help in upping the motivation of people. Mumbles something about free, open-source, spare time. Currently we only have the manpower to maintain ffdshow. For actual development we simply need a truckload of programmers with no personal life.


I assume that those deinterlacing problems exist for a long time already, and they are no recent regressions. So they won't hold back a beta. But anyway, I am no C++ programmer, and I don't really care about deinterlacing functionality.

[leeperry]

For actual development we simply need a truckload of programmers with no personal life.
Yeah, where are those guys ?! :D

I'm willing to betatest until the end of days.

well ffdshow is a fantastic piece of software, but it would need some polish on several points(buggy winamp2 plugin support, buggy AVISYNTH script support, buggy deinterlacing, etc etc)

but well there's no alternative to ffdshow anyway.....so I'll just STFU and enjoy what I get(for free at that, as you stated above) ;)

:thanks: all for your fantastic work so far!

[Yong]

use \n for new line, ex.: %idimar\nRemaining time: %rtime
i did tried with \n but i was using blackslash instead :eek:
what a stupid me :D
thanks btw.

This should make the OSD look more tidy:)

[LoRd_MuldeR]

yeah the deinterlacing section of ffdshow is very poor.

That's not true. Try "KernelBob" or "TomsMoComp" for example. :rolleyes:

Maybe there are some "better" Deinterlacers out there, but ffdshow includes some very good deinterlacers that work at realtime.
Having deinterlacers in ffdshow that don't work in realtime are kind of pointless...

My tip: If none of the "built-in" deinterlacers makes you happy:
1. Use ffdshow's Avisynth option to load your favorite deinterlacer, e.g. Yadif or TDeint. That's what I do for DVB-S playback.
2. ffdshow can use DScaler plugins, which offers various additional deinterlacers

the "cubic blending" doesn't do anything whatsoever, and the "linear blending" is very blurry.

Yes, "Cubic Blending" is apparently broken. But that's all.

"Linear Blending" is working as it's expected to work. If you don't like that (very simple) method, don't use it :p

ffdshow is so great, but it's got a few annoying bugs that noone will ever fix I guess :(

namely?

for instance....when using Avisynth scripts, very often Reclock says "no video stream found", I stop/press play again, and it's fine...

Why do you think it's the job of the ffdshow developers to workaround bugs in Reclock?


But anyway, I am no C++ programmer, and I don't really care about deinterlacing functionality.
Well, for me it's very important, because I use ffdshow for DVB-S playback. So deinterlace is essential!
And in my opinion ffdshow offers a great number of good deinterlacing methods :)

All I wanted to report is: One single deinterlacer filter appears broken, while all the others work fine.
If you can't fix "Cubic Interpolation", simply disable it for the time being. I won't miss it...

[clsid]

I'll add it to the bug list.

[leeperry]

Yes, "Cubic Blending" is apparently broken. But that's all.

"Linear Blending" is working as it's expected to work. If you don't like that (very simple) method, don't use it :p

well I only use deinterlacers for poorly encoded videos with combing artefacts, or interlaced WMV's.

and for this purpose, blending is my only option I think ?! maybe you have some better suggestion ?

in ffdshow, bicubic blending doesn't do jack and linear blending is blurry to hell.

the blending PS script of MPC is not blurry at all, yet it does the job very nicely.....it's in fact a very simple blending script....but it doesn't work with HR, only VMR9 and EVR :(

Why do you think it's the job of the ffdshow developers to workaround bugs in Reclock?

if I don't use the Avisynth capabilities of ffdshow, it NEVER happens.
if I enable it, half of the time it happens....BUT if I press stop/play, everything's fine again :confused:

several friends of mine have the exact same problem...

[leeperry]

I'll add it to the bug list.

any chance that you add the terrible winamp2 plugin support too, please ?

using the Izotope OzoneMP plugin is freezing to death each time you open a new file :(

there's some deadlock, mostly ffdshow kills the DLL before it closed down :(

[AnnaFan777]

Hi, thanks for the great ffdshow,

I'd like to suggest on " resize if "

resize if certain dimension or aspect ratio (or codec like DV) is met:

for example:
resize if -> 720x480 (or the codec is DV) then resize to 640x480
rezise if -> 320x240 (VCD) then resize to 640x480

thanks for the consideration..

[AnnaFan777]

Hi

I use ffdshow to watch DV and divx, some are progressive, some interlaced..

I was thinking "Deinterlace IF" option:

activate deinterlace if interlace pattern (comb) is dectected for ?? sec
or disable deinterlace if no comb is dectected for ?? sec

PS: I know it's just a few clicks away, but some are lazy,
and many users have no idea about interlace, they have de-int on,
so they always watch blurry progressive videos.


thanks for the consideration..

[clsid]

The resize stuff can mostly be accomplished by using the Profiles/Presets functionality of ffdshow.

You should post your deinterlace feature request on the SourceForge project page. Because otherwise tomorrow everyone will have forgotten about it.

[Mercury_22]

In ffdshow_rev2020_20080622_clsid_x64 everything it's OK Finally a working x64 version :thanks:

P.S. NO XviD encoding but who needs it ? :p

[Inventive Software]

I have a feature request for the remote API (external key support, basically). Is there a way (or future possibility) within that menu to add another shortcut key or combo for a Bob deinterlacing filter, as well as the Deinterlace filter? To visualise this, imagine there being 2 (or more) sections in the Deinterlacing tab: one for bob filters (LeakKernelDeint, DGBob, Kernel Bob etc), the other for the rest of them (and you could add one for the interpolation filters, for example). Then in the remote API have a key that switches between chosen filters for each kind to accomodate different video content (for example Bob for sports like football and F1, Blend for TV shows, etc). Simpler suggestion: just truncate the Deinterlacing filters and have Bob filters on their own tab, but of course don't allow them both to operate at the same time.

[LoRd_MuldeR]

Hi

I use ffdshow to watch DV and divx, some are progressive, some interlaced..

I was thinking "Deinterlace IF" option:

activate deinterlace if interlace pattern (comb) is dectected for ?? sec
or disable deinterlace if no comb is dectected for ?? sec

PS: I know it's just a few clicks away, but some are lazy,
and many users have no idea about interlace, they have de-int on,
so they always watch blurry progressive videos.


thanks for the consideration..

I second that request :)

It would be really useful for DVB playback too, because it switches between film and interlaced all the time :rolleyes:

BTW: I know that TDeint (full = false) as well as DScaler already have combing detection code...

[rickardk]

Anything changed with input stage off ffdshow audio recently?
I used to be able to have ffdshow audio connected after madFlac by setting format: uncompressed to "all supported".

But this does not work now I think.

[clsid]

Works fine for me. Tested with latest revision.

[Inventive Software]

libavcodec MPEG-1 decoder's broken with MPG files, it displays the wrong aspect ratio in MPC. Setting to libmpeg2 fixes the issue.

[clsid]

Works fine for me. Are those files perhaps anamorphic?

[Inventive Software]

I think so, the SAR reported is 5 digits / 5 digits, so I guess it's Anamorphic.

[clsid]

Added to bug list.

[thuan]

I'm using ffdshow_rev2026_20080625_clsid and keyboard hotkeys don't work. Can anyone confirm?

[Inventive Software]

You're using Ctrl+Alt when you press them, right?

[thuan]

Yes, of course. I have used these hot keys for a long time.

[Inventive Software]

Confirmed, it's broken with 2026. It was working with an older revision, but I think it was a couple hundred versions old.

[thuan]

Not really, it still worked fine for a couple of build back, or even the 2020 build, but I have no time to check on it now.

[LoRd_MuldeR]

I'm using ffdshow_rev2026_20080625_clsid and keyboard hotkeys don't work. Can anyone confirm?

Works fine, using the rev2020 ICL build...

[clsid]

It probably got broken by 2024. I'll revert that later.

[iron2000]

Not critical but-
Is it possible to have the system time in the OSD without the "System time:" heading?
And also make it 12hr instead of 24hr?

[clsid]

Is it possible to have the system time in the OSD without the "System time:" heading?Yes, if you create a user defined OSD string.
And also make it 12hr instead of 24hr?Nope.

[Thunderbolt8]

I noticed there seems to be again a problem with .ass subs (inside a mkv file). for a long time switching to a 2nd subtitle inside a muxed file resulted in some sort of strange characters and .srt like fonttype & white colour. then for a shorter time it worked fine with multiple (.ass) subtitle inside one .mkv file, and now theres no difference which of the multiple subs I choose inside that .mkv, only the first default one is displayed all the time.

[clsid]

Subtitle related stuff hasn't been changed recently. So any inconsistent behavior you are getting is probably splitter/file related. But anyhow, the subtitle functionality of ffdshow is still incomplete and there is nobody available to fix it. Haruhiko seems to be busy with other things.

[rack04]

Is anyone else experiencing broken AAC playback with ffdshow_rev2026_20080625_clsid.exe? Even if I set ffdshow audio decoder to prefer in MPC HC if won't attach to the filter.

Edit. I uninstalled ArcSoft and all is well. What was weird is that it even effected the DXVA operation of MPC HC.

[Thunderbolt8]

Subtitle related stuff hasn't been changed recently. So any inconsistent behavior you are getting is probably splitter/file related. But anyhow, the subtitle functionality of ffdshow is still incomplete and there is nobody available to fix it. Haruhiko seems to be busy with other things.
hm I tried it with older haali versions, but its all the same there :S

[georgevalkov]

Is anyone else experiencing broken AAC playback with ffdshow_rev2026_20080625_clsid.exe? Even if I set ffdshow audio decoder to prefer in MPC HC if won't attach to the filter.

Edit. I uninstalled ArcSoft and all is well. What was weird is that it even effected the DXVA operation of MPC HC.

Confirmed. I made a test under a Virtual PC, It seems that even when ffdshow's priority is set to the highest, ArcSoft's audio decoder will still have higher priority and is loaded instead of ffdshow. And because ArcSoft's audio decoder is broken for me too, the playback stops before it begins. The most easy way around is to unregister it:

go to start, run:
regsvr32 /u "%programfiles%\Common Files\ArcSoft\MPEG Engine\ASAudioHD.ax"

And enjoy ffdshow again :-)

[JarrettH]

Can anyone tell me about the dynamic contrast adjustment setting (Levels > Automatic)? What does it do :)

I mean, it appears better, but also has the effect of hiding shadow detail.

[K1ngp1ng]

request concerning ac3 decoder:
- drc should be off by default
- an option to disable dialogue normalization
- the dialnorm value should be shown in the info field

volume tab: level meters are too slow and inaccurate

then i wouldn't have to use ac3filter anymore...

when ffdshow does the decoding (e.g. flac 24bit) the output is 16bit - and after disabling 16bit output it is 32bit? (all processing tabs (volume, mixer,...) are disabled)

[clsid]

volume tab: level meters are too slow and inaccurateThat is because of performance reasons. The refresh rate was higher long ago, but that caused an unacceptable amount of CPU usage.

[K1ngp1ng]

i see. just thought if it could work like in ac3filter it would be perfect.
there is also a field to set the "interface update rate (ms)".
ac3filter only supports 5.1 that's why i would like to get rid of it.
also when a channel goes up to 0dbFS it doesn't go up that high in ffdshow or is that due to the low update rate?

[clsid]

There are dozens of suggestions and ideas that could be implemented into ffdshow. The only problem is that there are no people that have the time, interest, skills, or motivation to do it.

[cca]

I was testing ffdshow 64 bit, and discovered that the color conversion routines are tragically slow, I cannot play 720p h264 if I dare set YUV2 as output. YV12 woks just fine. I am correct to assume this is known issue, caused but unoptimized code for 64bit? The 32bit version can even do YV12 to RGB32 with decent speed. The system I use for testing is the one in my signature.

[clsid]

Make sure that you are using a recent build, because older ones are slower. If you are already using a recent build, then you are out of luck.

[cca]

Make sure that you are using a recent build, because older ones are slower. If you are already using a recent build, then you are out of luck.

Using your rev2026 build :) That's the only problem I noticed that could impact performance. It's nothing so major though, I just feed YV12 to EVR Custom instead of YUV2, and plays just fine. Just wanted to be sure it's nothing wrong my setup.

[Liisachan]

Yay for official beta5 :D !!!

[ShannonE]

Has anyone had any luck with the x64 builds? I've tried them all and can't seem to get any files to playback.

[cca]

Has anyone had any luck with the x64 builds? I've tried them all and can't seem to get any files to playback.

They work for me, at least for common files with XviD or H264 video. Please make sure that ffdshow is set for YV12 output because color conversion is still too slow.

[ShannonE]

They work for me, at least for common files with XviD or H264 video. Please make sure that ffdshow is set for YV12 output because color conversion is still too slow.
I've worked it out. It's because I was trying to load ffdshow64 with x86 MPC. I only want ffdshow64 so Vista x64 explorer.exe will give me damn AVI preview thumbnails.

[tiresias]

Re the half-resolution decoding of MPEG2 mentioned earlier in this thread.

What he asked for is controlled by AVCodecContext.lowres, which certainly works for MPEG2. I think it might already be exposed in the UI too.

I found it mentioned here:

http://www.irisa.fr/texmex/people/dufouil/ffmpegdoxy/structAVCodecContext.html#f260210a39ad4781d8d47ddac1541d04

int AVCodecContext::lowres

low resolution decoding. 1-> 1/2 size, 2->1/4 size

* encoding: unused
* decoding: set by user

Definition at line 1680 of file avcodec.h.

Reimar said

But the reality is, that even quarter resolution decoding is possible with (IMO) acceptable quality losses simply by setting a libavcodec flag. Half resolution in my tests reduces decoding CPU usage by 60%, quarter resolution by another 20% (compared to half res) for high bitrate HD material.
So my question was about the part I did not know: why it would not be possible to export this functionality via ffdshow?

Can I assume that half (or even quarter) resolution decoding of MPEG2 is possible via AVCodecContext::lowres, and that it produces a significant CPU usage reduction?

Also can I assume that AVCodecContext::lowres is not currently exposed in the GUI? I've looked through the filter properties for the MPEG decoder but can't find anything that looks like it (apologies if I'm looking in the wrong place, I'm a ffdshow newbie!).

Is it a significant amount of effort to expose AVCodecContext::lowres in the ffdshow gui?

[clsid]

Is it a significant amount of effort to expose AVCodecContext::lowres in the ffdshow gui?Probably not. If someone submits a proper patch, I'll gladly commit it.

[tiresias]

Probably not. If someone submits a proper patch, I'll gladly commit it.

Any takers?

I think this could be a useful feature for getting stutter-free playback of high definition MPEG on lower powered PCs.

[rickardk]

I'm using CoreAVC, Microsofts VC-1 decoder and Cyberlinks MPEG2 decoder to handle Blu-ray and HD DVDs put in mkvs.

To work around the video level bug (Vista Media Center with nvidia 8600) I use ffdshow to convert to RGB32.
Also ffdshow is used by Media Control to match refresh rate/frame rate.


So I configure ffdshow with Raw video set to "all supported".

ffdshow connects as expected to the output of above mentioned decoders (often YUY2). ffdshow handles the RGB conversion and outputs RGB32, flipped. I don't know why it says flipped.


But then I get another instance of ffdshow connected after the RGB conversion is done. So my next instance of ffdshow gets RGB32 as input and outputs the same.

If I restrict ffdshow to only one instance the video does not start. So it seems like at least 2 instances of ffdshow must be loaded.

But if I open the same video in graph edit I get 6 instances of ffdshow connected after each other.



Is this expected behaviour?

[clsid]

You should be getting only one instance with the default setting "only one - check all filters in graph".

[rickardk]

Yes thats right. But when just allowing one instance the video will not play.
For some reason (bug?) a minimum of two instances must be loaded when using ffdshow for this task (RGB conversion) under Vista x86 Ultimate (both with and without SP1).

What does the RGB32, flipped ouput mean?
The first ffdshow instance outputs RGB32, flipped. The second one outputs RGB32.

[clsid]

I assume flipped means that it is upside down. But why that happens I don't know. Someone else has to answer that.

[rickardk]

But it is not upside down. And I really don't understand why a second instance of ffdshow is needed to get playback working. Strange...

[Nightshiver]

I just have a quick question that I didn't think warrented the creation of a new thread, but which is the latest version of ffdshow? I went to the clsid builds on sourceforge and there, the latest general build is "ffdshow_prebeta5_rev2031_20080702_clsid", but when I went to clsid's SVN changelog, it says that there is a revision 2036. Just a tad bit confused, because I think I've been running a very old version of ffdshow. The one I currently use is "revision 1723 2007-12-24".

[clsid]

SVN is our version management system for the source code. Revision 2036 is currently the last version of the source code. There is no build for that yet. The latest build on the sourceforge site is 2031. Higher revision number means a newer version.

[Nightshiver]

So the prebeta5_rev2031 is the latest?

[Eragon4ever]

Actually ffdshow_beta5_rev2033_20080705_clsid.exe is the latest. It's listed under "Official ffdshow beta releases" not "SVN builds by clsid".

[Nightshiver]

Thanks :)

[thuan]

I'm using beta5 and somehow the ffdshow internal subfilter is acting funny when rendering styled sub. Whenever there's a line on top of the image, there's a weird line of numbers collides with the bottom line like this:
http://img73.imageshack.us/img73/7083/grab008170uw8.th.jpg (http://img73.imageshack.us/my.php?image=grab008170uw8.jpg)

[Anima123]

I'm using CrystalPlayer Pro 1.98 for playback. If rmvb files have aac (tagged as raac or racp) audio track, CrystalPlayer's internal real media splitter failed to forward the audio stream to ffdshow aac decoder, while normal splitter can. I have no idea if it's a bug on CrystalPlayer's internal splitter.

What I want to ask is if there some workaround that ffdshow can do to fix it?

Many thanks.

[clsid]

That is a splitter issue. Not related to ffdshow.

[dandy]

Hi

I'm new here - so first of all: thanks for the ffdshow tryouts, your efforts are greatly appreciated!

I have a problem with audio playback using ffdshow (all latest builds). Playing a mp2/mp3 file with mpc or graphedit there is - after using the STOP-Button (stop of playback and reset media file to beginning) - more often than not a clicking sound when playback is started again. I presume that these players use DirectShow API-functions IMediaControl.Stop and IMediaSeeking.SetPositions. According to MS-documentation every seek should flush all buffers of the filtergraph - but does ffdshoiw do it?

I have tested it with a variety of mp2/mp3 files, mp3lib or madlib, disabling all other filters in ffdshow, various output options etc. The result is always the same.

Can someone confirm?

Thanks
dandy

[yesgrey]

In ffdshow we can change the order of some of the filters.
I have tryed but I cannot move any filter after the RGB conversion option. Could this be done? I would like to use the avisynth filter after ffdshow's RGB conversion, and currently it's not possible...

[LoRd_MuldeR]

In ffdshow we can change the order of some of the filters.
I have tryed but I cannot move any filter after the RGB conversion option. Could this be done? I would like to use the avisynth filter after ffdshow's RGB conversion, and currently it's not possible...

I think that is because the "RGB conversion" page is just a sub-page of "Output", which has to be the last page/filter for obvious reasons.

Why not put "ConvertToRGB()" at the beginning of your Avisynth script?

[bb10]

I'm having a problem with ffdshow tryout beta 5 (on Windows Vista Ultimate 32bit) and my Pentium 4 with HyperThreading.

Because of the HyperThreading, the installer sets the amount of threads to "2", which is correct. However, this slows down decoding speed by ~20% as opposed to having the amount of threads set to "1". I have used the timecodec on a 3 minute 720p H.264 clip using threads set to "1" and "2":

1 Thread - User: 143s, kernel: 0s, total: 143s, real: 144s, fps: 50.7, dfps: 50.3
2 Threads - User: 11s, kernel: 3s, total: 14s, real: 180s, fps: 500.3, dfps: 40.3

Can anyone with HyperThreading confirm this?

[yesgrey]

Why not put "ConvertToRGB()" at the beginning of your Avisynth script?

Because the Avisynth "ConvertToRGB()" does not have all the functionality that the RGB Conversion in ffdshow has; and I think using a "ConvertToRGB()" at the beggining and a "ConvertToYUY2" at the end of the script for using it it's just throwing away cpu power and adding two desnecessary conversions...

[LoRd_MuldeR]

What functionality are you missing in ConvertToRGB() ???

And why do you need ConvertToYUY2() at the end of your script when you want ffdshow to put out RGB at end anyway?
That would be: YUY2 -> RGB -> YUY2 -> RGB

[iron2000]

A short question:
Is there anyway to make the OSD stay at a set position and not shift when playing videos of different resolutions?

[Leak]

A short question:
Is there anyway to make the OSD stay at a set position and not shift when playing videos of different resolutions?
Unless you use ffdshow to resize all videos to the same size (preferably the resolution of your screen) there's no way to do this, as the OSD text is added directly into the video and thus it's out of ffdshow's control how the video is going to get stretched and positioned.

[clsid]

Because of the HyperThreading, the installer sets the amount of threads to "2", which is correct. However, this slows down decoding speed by ~20% as opposed to having the amount of threads set to "1".The installer is actually supposed to set the number of threads to the amount of physical CPUs/cores. HT sucks, as you have already discovered.

I shall look into it.

[yesgrey]

What functionality are you missing in ConvertToRGB() ???

And why do you need ConvertToYUY2() at the end of your script when you want ffdshow to put out RGB at end anyway?
That would be: YUY2 -> RGB -> YUY2 -> RGB

The possibility of creating a custom levels YUV->RGB conversion in the "Contrast" group.

Read this (http://forum.doom9.org/showthread.php?p=1159507#post1159507) thread.

That would be: YUY2 -> RGB -> ddcc -> RGB-> YUY2 -> RGB

If we could apply the avisynth filter after the RGB conversion, that would be: YUY2 -> RGB -> ddcc -> RGB

[CHEF-KOCH]

Hi, i am using the beta 5 on Vista x64 sp1. All work's good, but i have a screen.

http://img376.imageshack.us/img376/330/unbenanntvh7.th.jpg (http://img376.imageshack.us/my.php?image=unbenanntvh7.jpg)

i use avisynth 2.5.8 Rc2 witsh is from here http://www.free-codecs.com/download/AviSynth.htm

the beta 5 is incompatible with this version of avisynth?

thx for answers

[Leak]

the beta 5 is incompatible with this version of avisynth?
That message is from the AviSynth MT.dll plugin, not ffdshow - the regular 2.5.x AviSynth builds won't work with it, as the support code for it is only in 2.6.x or tsp's patched 2.5.x builds.

So you need to either install a compatible version of AviSynth or stop using the MT filter...

np: Nine Inch Nails - Capital G (Year Zero)

[CHEF-KOCH]

That message is from the AviSynth MT.dll plugin, not ffdshow - the regular 2.5.x AviSynth builds won't work with it, as the support code for it is only in 2.6.x or tsp's patched 2.5.x builds.

So you need to either install a compatible version of AviSynth or stop using the MT filter...

np: Nine Inch Nails - Capital G (Year Zero)


Ok, thanks .. i use mt.dll v 0.7.0.0 and it work :)

[vadeluma]

I use ffdshow for playing recordings and livetv in vistamce. After a new install i'm not able to use avisynth anymore. When i check the avisynthtab in ffdshow my files won't play and vistamce is telling me some audio and videocodecs are missing. When i uncheck the tab all files play well. So what did i forget? Someone?. I know they played before with avisynth and all codecs are there.

[Leak]

I use ffdshow for playing recordings and livetv in vistamce. After a new install i'm not able to use avisynth anymore. When i check the avisynthtab in ffdshow my files won't play and vistamce is telling me some audio and videocodecs are missing. When i uncheck the tab all files play well. So what did i forget? Someone?. I know they played before with avisynth and all codecs are there.
Well, ummm - did you install AviSynth? And can you open a simple AVS file only consisting of Version() in VirtualDub or Media Player?

np: Nine Inch Nails - The Good Soldier (Year Zero)

[Sharktooth]

bug - ttxt subtitles are always rendered in italic with small size and with times (new) roman font. independently from the subtitle settings.
config: HAALI MS (subs enabled), MPC (internal source filters disabled), ffdshow rev 2031. but this bug is really there from a while....

[vadeluma]

Well, ummm - did you install AviSynth? And can you open a simple AVS file only consisting of Version() in VirtualDub or Media Player?

np: Nine Inch Nails - The Good Soldier (Year Zero)

Yes, i can. Everything works in wmp classic.

[iron2000]

Maybe this is posted already.

I found that by enabling "Allow multichannel audio processing" in Winamp2 causes stuttering audio in videos with multichannel sound(had the porblem on one with aac and one with ac3) .

[CruNcher]

I make it short :)

Remuxing this http://mirror05.x264.nl/CruNcher/force.php?file=.//Vegas_1280x528.wmv with gdsmux to .mkv and playing it with ffdshow as Video decoder, plays it 1/2 second unsync to the sound no difference if ffdshow video + ffdshow audio or ffdshow video + microsoft audio decoder is used.

[pitch.fr]

I think there is a problem with the "export all settings" option....it doesn't export the Avisynth scripts :(

If I export the /GNU/ key with regedit, then it's properly exported to a .reg

[Inventive Software]

Are the devs aware of the following H.264/AVC patch or commit for ffmpeg/mplayer? http://forum.doom9.org/showthread.php?t=139526 It'd be nice to see it implemented, if it's easily patchable against ffdshow. I'm unable to try it due to me screwing around with different things on my laptop. :)

[clsid]

That is unfinished stuff. Anyhow, xxl already tried patching our version of libavcodec, but it crashed.

[clsid]

I think there is a problem with the "export all settings" option....it doesn't export the Avisynth scripts :(

If I export the /GNU/ key with regedit, then it's properly exported to a .reg
Ask Leak if he can fix it.

[_xxl]

Are the devs aware of the following H.264/AVC patch or commit for ffmpeg/mplayer? http://forum.doom9.org/showthread.php?t=139526 It'd be nice to see it implemented, if it's easily patchable against ffdshow. I'm unable to try it due to me screwing around with different things on my laptop.
Already tried but crashing when frame threading is enabled. This weekend if I have time I''ll try again with latest code. The easy way is to compile ffmpeg.exe from ffmpeg-mt.

[Leak]

Ask Leak if he can fix it.
I thought Haruhiko fixed that some time ago?

Works for me - it writes a 1:1 copy of the ffdshow registry key.

Of course, the old "avisynthScript" value will still be b0rked when there are newlines in the script, but it also writes out a "avisynthScriptMULTI_SZ" value with the script hex-encoded.

If the export somehow acts funky, try deleting the "avisynthScript" values under "HKEY_CURRENT_USER\Software\GNU\ffdshow\<profile name>" in the registry. (After making sure that the "avisynthScriptMULTI_SZ" value exists; if not, try changing the script in the config dialog and hitting "Apply" - I'd hazard a guess that the new MULTI_SZ value isn't written unless the script is changed...)

np: Radiohead - Backdrifts (Honeymoon Is Over) (Hail To The Thief)

[JarrettH]

I have a question...

I want to use Spline and Resize to Screen Resolution. Is there a way, using MPC, to have the video enter full screen instead of staying windowed? If I Spline Resize to this and double click to enter full screen, doesn't this just resize my resized video?

:thanks:

[LoRd_MuldeR]

If you resize to your screen resolution in ffdshow (e.g. Spline resize) and then go fullscreen in MPC, the video renderer will not change (re-resize) the video.
But if you resize to a different resolution than your screen resolution, then going fullscreen will resize your video again and might sacrifice quality...

[Leak]

If you resize to your screen resolution in ffdshow (e.g. Spline resize) and then go fullscreen in MPC, the video renderer will not change (re-resize) the video.
But if you resize to a different resolution than your screen resolution, then going fullscreen will resize your video again and might sacrifice quality...
Selecting "Video Frame > Normal size" should take care of any unwanted resizing in the renderer...

np: Nathan Fake - Charlie's House (Apparat Remix) (Things To Be Frickled (Disc 1))

[pitch.fr]

Works for me - it writes a 1:1 copy of the ffdshow registry key.

Of course, the old "avisynthScript" value will still be b0rked when there are newlines in the script, but it also writes out a "avisynthScriptMULTI_SZ" value with the script hex-encoded.
ok thanks for your help.

well I dunno, seems to work now ?!

but from now on I will export with regedit I think.

problem solved on my end :D

[DrGuz]

(I was torn between starting a new thread and appending to this one ... got no response (http://ffdshow-tryout.sourceforge.net/phpBB2/viewtopic.php?p=3152#3152) on the official ffdshow tryouts support forum ... apologies if I'm doing it wrong)

I'm getting a black screen trying to process an MJPG AVI. Is there some diagnostic tool to make sure ffdshow VFW is working right? Or would someone be willing to help me troubleshoot?

SOFTWARE VERSIONS:
VirtualDub 1.8.3, VirtualDub-MPEG2 1.6.15, and Vegas 4.0
ffdshow tryouts revision 2033 Jun 27 2008 14:11:43 (msvc 2003, x86, unicode)
Windows XP SP2

Here are some things I've already tried....

The .avi file gets a clean report from GSpot. 4CC = MJPG, 704x480, 204 MB, 848 frames, 29.970 fps, PCM audio (44100/16/stereo). It was created by VirtualDub capture on a Windows 98 machine, through a miro DC30 card from laserdisc.

VirtualDub shows it perfectly if I set its "Preferences > AVI > Prefer internal codecs over installed third-party codecs (MJPEG and DV)".

ffdshow "VFW configuration > Decoder > codecs > MJPEG" is set to "libavcodec".

Removing "libavcodec.dll" from the ffdshow folder causes VirtualDub to work properly, presumably because VFW fails and it falls back on its internal MJPG decoder. This proves to me that VirtualDub is finding the right files for ffdshow / libavcodec.

Activating "VFW configuration > Decoder > Bitmap overlay" with opacity 100% fails to produce an image in the VirtualDub output pane. This suggests that ffdshow itself is not running properly, rather than libavcodec being to blame.

The ultimate goal is to get this AVI into Vegas 4.0, which has no built-in MJPG support and only works with VFW, not DirectShow. Prior to installing ffdshow, Vegas would only import the audio track; after installing, it now imports both audio and video tracks, but the video track is solid black.

[quacka]

Hi,

I am trying to use ffdshow tryouts filter in my application to switch audio stream. I am a VB.NET developer and have built the graph manually to include ffshow fine.

I have read that albain has added remote API into ffdshow and have opened up ffdshowRemoteAPI.cpp to have a look. However due to my lack of knowledge of C++ I do not know how to use this in my application. Anyone have a sample snippet which uses RemoteAPI they can show me?

Cheers

[Thunderbolt8]

I found some kind of image corruption when I watched the dark city blu-ray with ffdshow 2060 & libavcodec. when I choose ffdshow & wmv9 instead to decode the picture then it looked fine.

libavcodec
http://thumbnails9.imagebam.com/1054/4fcd3210533562.gif (http://www.imagebam.com/image/4fcd3210533562)

wmv9
http://thumbnails9.imagebam.com/1054/da116110533563.gif (http://www.imagebam.com/image/da116110533563)

this problems only occured at 2 places in the movie, but is still somwhat strange

[foxyshadis]

I found some kind of image corruption when I watched the dark city blu-ray with ffdshow 2060 & libavcodec. when I choose ffdshow & wmv9 instead to decode the picture then it looked fine.

libavcodec
http://thumbnails9.imagebam.com/1054/4fcd3210533562.gif (http://www.imagebam.com/image/4fcd3210533562)

wmv9
http://thumbnails9.imagebam.com/1054/da116110533563.gif (http://www.imagebam.com/image/da116110533563)

this problems only occured at 2 places in the movie, but is still somwhat strange

These decoding issues have to be reported to the ffmpeg team, not ffdshow. We can't fix them (well, very rarely).

[Sharktooth]

it seems a deblocking issue but im not so sure.

[clsid]

I have added a recent FFplay build to the first post of this topic. With that tool you can check whether a decoding problem is caused by a bug/limitation of the FFmpeg code.

[LoRd_MuldeR]

I have added a recent FFplay build to the first post of this topic. With that tool you can check whether a decoding problem is caused by a bug/limitation of the FFmpeg code.

I just noticed that your download is missing the "SDL.dll" lib...

[clsid]

Fixed.

[rebkell]

Fixed.

Download link isn't working, 'File Not Found'

[LoRd_MuldeR]

Download link isn't working, 'File Not Found'

Works here. But is it normal the FFplay will die quietly as soon as I resize the window?

[clsid]

Resizing works here. FFplay is a very very simple player. I doesn't seem to output any error messages at all.

[rebkell]

Works here. But is it normal the FFplay will die quietly as soon as I resize the window?

It's there for me now also.

[JarrettH]

Selecting "Video Frame > Normal size" should take care of any unwanted resizing in the renderer...

Hey thanks that's a good solution :D

[rkalwaitis]

Doesnt the later versions of the tryouts have snow? I can not see it? Trying to use VirtualDub. Thanks

[fastplayer]

It has been disabled a while ago due to "incompleteness".

[rkalwaitis]

does anyone know he last build with snow?

[fastplayer]

does anyone know he last build with snow?
Any build with revision < 1526.
Try this one:
http://downloads.sourceforge.net/ffdshow-tryout/ffdshow_rev1523_20071009_clsid.exe

[clsid]

The SNOW specification is still unfinished. So anything you encode now is potentially incompatible with future versions. I don't know why you want to use SNOW, but I would strongly advice against using that old ffdshow build. Better use a recent FFmpeg build.

[Seraphic-]

MPC-HC 1.1.604.0
ffdshow_beta5_rev2033_20080705_clsid
ac3filter_1_51a
MatroskaSplitter

Hi,

Have the above four software apps installed to handle playback of my video/audio. However, now when I try to play my 5.1DTS track, I get FL, FC, FR, BL, BR, but I'm getting nothing from my SUB.

I know the track has a LFE and I even tried 5.1 tracks from HERE (http://www.kellyindustries.com/sounds.html), but with them all, I'm still getting no SUB.

Used to use codec pack (http://www.free-codecs.com/download/K_Lite_Mega_Codec_Pack.htm) and always had full 5.1 sound.

Is there something I'm missing?

[Wishbringer]

Hope, it's only a sourceforge problem (hickup):
when trying to access site on sourceforge I get an "Invalid Project" error.

[clsid]

Assuming that AC3Filter is decoding that DTS audio, you need to adjust those settings, not ffdshow's.

[rkalwaitis]

thanks clsid Ill go that route

[Seraphic-]

Assuming that AC3Filter is decoding that DTS audio, you need to adjust those settings, not ffdshow's.

AC3Filter is set to 5.1 though.

As is my system and ffdshow tryouts.

Maybe it has to do with MPC-HC not working with AC3Filter.

Edit:
If I uninstall AC3Filter, I just get front left/right.
ffdshow tryouts is set to 5.1 and to decode DTS, so I don't know.

[clsid]

MPC-HC works fine with AC3Filter. You must however disable the internal AC3 decoder.

[Seraphic-]

MPC-HC works fine with AC3Filter. You must however disable the internal AC3 decoder.

You mean the DTS Transform Filter right?
If so, that fixed the issue.
Thanks for the help.

Last issue that remains is that the DTS volume is MUCH louder then it was when using the codec pack.
There must be some audio normalization somewhere that was enabled in the codec pack, but not in the current settings.
It is peaking out just about, was not like that before.

Test Track (http://www.temp.seraphicgate.com/51Test3.dts)
http://www.temp.seraphicgate.com/51.jpg

[clsid]

AC3Filter has lots of options. I think 'Normalize matrix' is the one you are looking for. Although that should normally already be enabled by default in AC3Filter.

[Seraphic-]

AC3Filter has lots of options. I think 'Normalize matrix' is the one you are looking for. Although that should normally already be enabled by default in AC3Filter.

Yeah, it is. Odd that it plays with a much higher volume.

Edit:
Just tried a two channel stereo wav file and it is playing as 5.0 (no sub), lol.
Sigh... might as well go back to the codec pack.

Edit2:
I'll install the codec pack again and mirror the settings to the other install.
See what that does.

[Seraphic-]

Quite a few AC3Filter settings were different and changing them did make a difference in some cases.

Seems like I just go back to the codec pack as it works 100% for me.

I just doesn't use ffdshow tryouts or MPC-HC, it uses ffdshow and MPC.

But was hoping to use this setup to avoid conflicts with my audio/video encoding.

[clsid]

That pack also uses tryouts. Almost everybody just uses the name 'ffdshow'.

[iron2000]

I seem to have problems with video files with Vorbis sound.
Theres kind of a "zzz"/rapid skipping sound every few seconds.
When it gets severe the video slows down also.
All other sound types seem ok.

Tried beta 5, 2060 and 2063, all had the problem.
Switch from Tremor to libavcodec, no use too.
Problem exists in MPC-HC's Vorbis internal filter also.

Updated Direct X today.
Run kX Drivers for audio.

Anything else I can try?

Ok defintely a problem with ffdshow as it plays fine with CoreVorbis.
But using CoreVorbis with ffdshow uncompressed all supported reproduces the problem.
Or is this a recent known problem?

[Rumbah]

I just ran into a problematic sample but I don't know if it's the fault of ffdshow, Avisynth or the stream itself.

If I play the original video file in Virtualdub via ffdshow VFW it works. The playback with FFPlay also works. But if I open the video via Avisynth, the ending gets corrupted (and I get a green first frame). I'm using ffdshow beta 5. I did a test encode with the video to show the corrupted results I get.

The Avisynth script that I used for encoding:
AVISource("Test.avi")

The original file:
http://www.speedshare.org/download.php?id=CFE7360B12

The encoded file with errors (after 16-20 seconds):
http://www.speedshare.org/download.php?id=1E225C7D12

[Inventive Software]

Use the latest revision, as many bugs have been fixed, and tell us if the problem still arises. The ffplay version is probably a much newer compile of libavcodec than beta5 of ffdshow, thus the decoding errors are more likely to have been fixed.

[_xxl]

But using CoreVorbis with ffdshow uncompressed all supported reproduces the problem.
Or is this a recent known problem?
Are you sure that this is a ffmpeg,vorbis/tremor bug?
ffdshow uses some code from mpc and if there's a bug be sure that will be in ffdshow also.

[iron2000]

Are you sure that this is a ffmpeg,vorbis/tremor bug?
ffdshow uses some code from mpc and if there's a bug be sure that will be in ffdshow also.

The problem happened both when using ffdshow audio decoder and the MPC internal filter.
But no problems with Corevorbis.

[thuan]

I get the same vorbis problem with 2060 and 2063 using libavcodec, but not with any older or Tremor or MPC decoder. It might be a different problem though.

[clsid]

I have updated the libavcodec Vorbis decoder. Maybe it works better now.

Two modifications are not present in the current updated code:
1) Compatibility with OggDS splitter (a very old outdated filter)
2) VorbisGain support

I don't think it is worthwhile to add support for OggDS. That comes its own decoder anyway.

Does anyone here use the VorbisGain option?

[thuan]

If it's not that much of a hassle, then add it, I have some old encodes that use it.

[iron2000]

The new 2073 still have the same problems with Vorbis on Tremor.
But with 2073, Vorbis on libavcodec produces distorted sound.

One other thing is that the problem appears too when I tried to use CoreVorbis with ffdshow (to use ffdshow's Winamp2, resample and mixer by making ffdshow take in uncompressed format).
Unchecking the 3 doesn't help too.
No problems with CoreVorbis itself.

[_xxl]

But with 2073, Vorbis on libavcodec produces distorted sound.
samples please

[iron2000]

Sorry but how do I provide samples?

Tried going back to 2026, the same Vorbis problem occurs also.

[clsid]

With samples, we mean sample files.

Add the updates I made are part of revision 2074. So 2073 still has old libavcodec vorbis.

[clsid]

If it's not that much of a hassle, then add it, I have some old encodes that use it.
It's not easy. I will attempt it later this week.

[iron2000]

Sample file as in the file that is played?
For me is any MKV file with Vorbis audio.

One is the Slayers Revolution anime from SS-Eclipse fansub.
Another is Zettai Karen Children from Deculture fansub.
Theres a lot more.

[Inventive Software]

Sorry, can't help with "fansubs" or similar downloaded material. Consult rule 6 for more information. If you don't use it in day-to-day encoding, you've no need for it. Go use CoreVorbis.

[iron2000]

Ok understood.
Thanks anyway.

[vlada]

Does anyone here use the VorbisGain option?

Yes, I do. It is a very important feature for me, when listening to music. But usually I use foobar2000 which has it's own decoder. But for a DirectShow player it might be important.

On the other hand there are is only one common DirectShow music player - WMP. Next year maybe AmaroK will become popular. But both players have their own solution to equalize volume. WMP has some proprietary solution, AmaroK applies VorbisGain (ReplayGain) during playback.

Anyway ReplayGain support is part of Vorbis specification AFAIK. So it should be implemented, I think.

[thuan]

@Inventive Software: Leaving fansub aside, a bug is a bug in libavcodec vorbis decoder. You can just make a sample yourself using the latest oggenc(2). It doesn't work anywhere AFAIK. And if you ask for a sample, I'd just say what can I upload without breaking rule 6, is it some thing I made myself?

EDIT: @clsid: Thank you, your latest build fixed this problem.

[pitch.fr]

is there a way to get TWO Avisynth filters in a row ?

Ideally I'd like to do :
1) LSF in YV12 in Avisynth
2) unsharp masking in YV12
3) ConvertToRGB32 and some RGB32 filters in Avisynth

I've tried to use both the "ffdshow Video Decoder" and the "ffdshow Raw Video Filter", but they seem to use the same settings :(

and if I do it all in ffdshow, it does YV12(source)>RGB32(LSF+ConvertToRGB32)>YV12(unsharp masking)>RGB32, which ruins the PQ and the CPU cycles...

[clsid]

Not possible. You'll have to find a way to do unsharp masking in Avisynth. I assume there is an Avisynth plugin for that?

[pitch.fr]

yes, this is identical to the unsharp masking in ffdshow(based on neuron2's code) :

VD_UnsharpMask(diameter=3,strength=__,threshold=0)

but this plugin code is vastly unoptimized and not usable in real time at all(event with MT) :(

that'd be cool if ffdshow was modular, and you could use several times the same filter :)

[yesgrey]

I think you could use multiple instances of ffdshow... one after the other. See tab Directshow control. I think I have run it one time, without wanting it, and then disabled it.

[pitch.fr]

yes, but you're never sure of what order they'll be opened in.

and they would all use the same settings.

but that's OK, unsharp masking prior to LSF looks good too.

and with my 3.3Ghz Q6600, I can run LSF and DDCC in real time to get SMPTE-C gamut conversion(the native gamut of DVD/BD) :D

Avisynth support in ffdshow is really awesome :)

and I've recently found out that doing 24/96 upsampling in AC3filter was multithreaded, 96KHz upsampling in ffdshow is a lot more CPU/jitter intensive.

bottom line is : a +3Ghz Q6600 is like the ultimate CPU to use ffdshow+avisynth :)

http://pix.nofrag.com/7/b/4/ba67a6e00fe1d143238978fa776c4tt.jpg (http://pix.nofrag.com/7/b/4/ba67a6e00fe1d143238978fa776c4.html)

http://pix.nofrag.com/7/b/9/106cf2f94eaec28a0da84a6ddc5e0tt.jpg (http://pix.nofrag.com/7/b/9/106cf2f94eaec28a0da84a6ddc5e0.html)


besides, the ICL10 versions of ffdshow seem a bit faster, thanks clsid!

[Rumbah]

I just ran into a problematic sample but I don't know if it's the fault of ffdshow, Avisynth or the stream itself.

If I play the original video file in Virtualdub via ffdshow VFW it works. The playback with FFPlay also works. But if I open the video via Avisynth, the ending gets corrupted (and I get a green first frame). I'm using ffdshow beta 5. I did a test encode with the video to show the corrupted results I get.

I tried the latest version (rev 2079) but the problem still isn't gone. If I can provide any further information that helps just tell me ;)

[clsid]

The likelihood of your problem being fixed any time soon is exactly zero. Reason: there are no active developers in the ffdshow project at this moment.

The only updates that are currently being done to ffdshow are related to libavcodec.

[Rumbah]

The likelihood of your problem being fixed any time soon is exactly zero. Reason: there are no active developers in the ffdshow project at this moment.

That's sad to hear.
Then I'll just have to use FFmpegSource as I don't get the error that way.

[iron2000]

Found out that my Vorbis problem is with the changes in the kX DSP.
Problem now solved.

[pitch.fr]

is there a way to get some good DeHalo/Dering in ffdshow ?
I've tried the Dering in the "post processing" filter, but it isn't working too well on strongly halo'ed videos.

and the Avisynth filters I've found are slooooooow as hell.

you see, the Gothika BD is just unwatchable :

http://pix.nofrag.com/2/9/a/e9d735b179b19cfc592b032fc26dbtt.jpg (http://pix.nofrag.com/2/9/a/e9d735b179b19cfc592b032fc26db.html)

[Mercury_22]

Hi !
Please help me :helpful: I have an old problem : I can't play (only black screen with sound) many mpeg4 / DX50, (encoder XviD build 37) files on Vista & EVR output
This problem apperas only with Intel video "cards" and EVR or Evr custom output (=Vista).
If I use DivX codec or any player with VMR output on vista (= NO WMP) or any other video card I have no problem !
So i'm thinking if divx it's working why ffdshow isn't ?

Here is a sample Mpeg4Test (http://rapidshare.com/files/121074851/Mpeg4Test.avi.html) or Mirror (http://www.dump.ro/download.php?file=Testmpeg4.avi&id=MzBmNzhiNzhlYjZiOGQ3ZGU5OGJiZjkzODMxMTk2MmI=&download=true)

:helpful: :thanks:

Edit: solved thanks to clsid : I've just disabled YV12 in ffdshow output settings !
Why it's ffdshow the only one with Forced by default YV12 Colorspace output ? (DivX or Xvid don't have this by default)

[Eragon4ever]

http://samples.mplayerhq.hu/V-codecs/FPS1/v5/fraps-v5-bouncing-balls.avi

This video does not play correctly in ffdshow, it does however in ffplay (thanks for updating the build, clsid). This is how it should look like:
http://img401.imageshack.us/img401/396/snap0010eu5.png

In ffdshow the on-screen text is light blue and the balls are red<->blue.

[clsid]

Yes, colors are wrong, even with an up-to-date version of libavcodec. I don't know why. Perhaps a colorspace issue. I have no idea how to fix it.

[pjo]

Hello folks. I am trying to find out why AAC5.1 audio is not played ok with MPC.

If MPC setting is changed not to use ffdshow, it plays ok.

http://rapidshare.de/files/40262409/00001cutted.m2ts.html

This is the first portion of the file. The first few seconds portion is an ending of a news show with mono AAC (probably), then the main show starts with AAC5.1.

After cutting the news portion with tsMuxeR, the main show is really AAC5.1 when played with MPC without using ffdshow.

Does ffdshow support AAC5.1 ?

pjo

[pjo]

Hello folks. I am trying to find out why AAC5.1 audio is not played ok with MPC.

If MPC setting is changed not to use ffdshow, it plays ok.

http://rapidshare.de/files/40262409/00001cutted.m2ts.html

This is the first portion of the file. The first few seconds portion is an ending of a news show with mono AAC (probably), then the main show starts with AAC5.1.

After cutting the news portion with tsMuxeR, the main show is really AAC5.1 when played with MPC without using ffdshow.

Does ffdshow support AAC5.1 ?

pjo

I was able to cut the first part of the mono news show segment of the .m2ts file with MurdocCutter. And then it has only AAC5.1 part. It plays ok with ffdshow or AAC decoder. Checking "downmix to stereo" does not matter in this case.

VLC and MPC with AAC decoder can cope with the sudden change of the audio channel format. MPC with ffdshow does not cope with this change.

pjo

[tetsuo55]

For which codecs can ffdshow be considered as Best of the pack? i know that its Divx/Xvid decoder is the fastest while maintaining the same quality.

What about the other audio/video codecs?

[Sharktooth]

there is no Best... as usual. however i use it for everything...

[tetsuo55]

well in this case Best could be used because:

These applications are all doing the same thing in the same way to reach the exact same result. all things equal it comes down to the one that uses the least CPU time.

The only problem is the output image is not always bit-for-bit identical (usually this means a bug in the decoder, better/worse post processing(which im not asking about at this time) or better/worse handeling of errors in the bitstream)

i once read a mpeg1 comparison where one of the decoders "fixed" the bitstream errors with a psychovisual repair of the broken pixels. It used a bleed effect to mask them, kinda like Philips Ambilight

[Sharktooth]

nope... "best" cant apply coz i may prefer PP options while others may prefer the speed or the multithreading or other factors/features.
"best" is ALWAYS subjective and there is NO case where the word "best" is appopriate except "best is what's best for yourself". hence the forum rule number 12...

[tetsuo55]

nope... "best" cant apply coz i may prefer PP options while others may prefer the speed or the multithreading or other factors/features.
"best" is ALWAYS subjective and there is NO case where the word "best" is appopriate except "best is what's best for yourself". hence the forum rule number 12...

Fair enough.

Question:

Which of the codecs in FFdshow meet the following requirements for me:
*Has bit-for-bit equal image quality to other decoders(Basically just displays the image as it was ment to be displayed)
*Has the fastest/equal decoding engine when compared to the competition
*Has better/equal handling of broken files than competing decoders
*Has better/equal bitstream error handling than competing decoders(psychovisual fixes are a bonus)

I believe the post-processing engine in FFdshow is seperated from the codecs, so in this case the quality of post-processing does not matter. (if you happen to know the difference in quality for PP though please mention it)

[Dark Shikari]

*Has bit-for-bit equal image quality to other decoders(Basically just displays the image as it was ment to be displayed)H.264, since nearly no other lossy format has that (MPEG-4 ASP certainly doesn't), and FFDshow's support for VC-1 isn't totally complete.

[tetsuo55]

H.264, since nearly no other lossy format has that (MPEG-4 ASP certainly doesn't), and FFDshow's support for VC-1 isn't totally complete.

Hmm i didn't know that. The only way to compare would be PSNR between source and transcoded image played back with the available decoders for that format with all PP disabled

[Dark Shikari]

Hmm i didn't know that. The only way to compare would be PSNR between source and transcoded image played back with the available decoders for that format with all PP disabledNo comparison necessary, any format that uses DCT and doesn't specify an exact DCT approximation is inherently not bit-exact.

[Sharktooth]

the h.264 standard includes bit exact decoding. if the decoder isnt able to do it then it's not compliant.
you may be almost safe assuming the major h.264 known decoders are displaying the SAME picture.

[tetsuo55]

So for h264 you only have to compare the other features.

And for all the other codecs you have to do a PSNR check on a source and compressed file using all available decoders to find out which one produces and image closest to the original

No wonder almost no-one wants to do that

---

And thanks guys, i didn't know h264 rocked this much(it was already my personal favorite)

[nm]

i once read a mpeg1 comparison where one of the decoders "fixed" the bitstream errors with a psychovisual repair of the broken pixels. It used a bleed effect to mask them, kinda like Philips Ambilight
That was probably the MPEG-1/2 decoder in FFmpeg/libavcodec (and in ffdshow, if enabled): http://compression.ru/video/codec_comparison/mpeg-2_2006_en.html
It uses linear blending to cover broken macroblocks.

[tetsuo55]

That was probably the MPEG-1/2 decoder in FFmpeg/libavcodec (and in ffdshow, if enabled): http://compression.ru/video/codec_comparison/mpeg-2_2006_en.html
It uses linear blending to cover broken macroblocks.

Indeed!, does ffdshow use this for all codecs?

[Mr VacBob]

The important ones (mpeg-*) have it implemented. I think it's not really tuned for h264, though.

[3ngel]

I would like to signal a bug in the DTS-ES decoder.

With a DTS-ES 6.1 Discrete stream, the BackSurroundCenter channel is not decoded (field in the BackC of FFDshow is blank). Pheraps the ES extension is not recognized correctly?

I have samples stream to test, and this

http://www.sr.se/laddahem/multikanal/dts/norgeES.zip

that is said to be DTS-ES 6.1 Discrete.

Hope this help

[clsid]

I suspect the problem is that there is no BackC on the Volume page. Meaning that ffdshow probably just ignores any input on that channel. BackC is present on the Mixer page. So ffdshow should able to output to it (using a custom matrix).

Now you will have to find someone to add BackC to the Volume page. And no, I won't do it.

[3ngel]

But the problem, is that in the "Mixer" panel, even if i set "same as input" as matrix, the BackC rectangle channel remains blank so that means that the channel is not read at all at the first decoding stage of dts.

[clsid]

That's what I said. Noting is done with the decoded input on that channel. And there aren't multiple decoding stages. Audio gets decoded in a single step (by libdts) and the decoded audio is then processed by the filters in ffdshow.

Of course it is also possible that your file simply gets decoded as 5.1 by libdts.

[3ngel]

Of course it is also possible that your file simply gets decoded as 5.1 by libdts.

Yes, that's what i mean. Because i know that the discrete ES channel (BackC) is stored in a separate extension in the dts frame, pheraps for some reason that frame part is ignored by libdts.

So i was asking if someone can take a look at it :)

EDIT:
Ah, so you say the ES channel is decoded but ignored by the mixer and volume? But in this case would not every decoded (for axample from a 7.1 LPCM) BackC would be ignored?

[clsid]

I can confirm that libdts only supports regular DTS.
http://www.videolan.org/developers/libdca.html
None of the extensions (extra channels, 96/192kHz sample rates) have been implemented (and unfortunately the public standard provides only minimal information on them).
So there is no solution for your problem.

[3ngel]

I see, thanks for the info :)

What a pity btw :(

I suppose i'll have to search for some alternate dts decoder eheh

[3ngel]

Here i am again :)
Now first try to DTS-ES i've used use Sonic Decoder, and it effectly does decodes 6.1 channels.

But here cames two questions:

1) Why the BackL, BackC, BackR aren't considered when doing a DPLII downmixing? Here the screen
http://img373.imageshack.us/img373/6497/clipboard01cm6.png
As you can see they are 0,0,0 thus being eliminated instead of downmixed

2) What is the difference between a DPLII matrix and a Prologic Matrix? They seem identical
http://img229.imageshack.us/img229/6148/clipboard02da8.png

UPDATE:
These are the Matrix of Ac3Filter

Surround
http://img380.imageshack.us/img380/623/dpl1kn1.png

DPLII
http://img293.imageshack.us/img293/7411/dpl2ta7.png

These are from Wikipedia
http://img380.imageshack.us/img380/5731/25727837ww4.png

For the 6.1 Channels there is a specific Matrix called
Dolby Pro Logic IIx

UPDATE2:
And here it is directly from Dolby site the DPLIIx Matrix (at last from the phase point of view)
http://img169.imageshack.us/img169/2431/129025gc2.jpg

The BackC could be inserted between the Lb Rb offsiting them a bit

[tetsuo55]

I decided to set everything to libvacodec(except h264). I did the same with the audio(except for AC3/DTS).

Reading the ffmpeg libvacodec changelogs it does seem that a lot of the supported codecs are faster or as fast as other decoders.

The only way its going to get any better if people use it and report bugs right?

I set postprocessing to :
Should i enable "automatic quality control"? its now set to preset all the way to the right.

mplayer>accurate deblocking
Should i enable "fix luminance" and "Full luma range"?

Then Decoder options:
IDCT: auto
Work around Encoder bugs: autodetect
Error resilience: very agressive
Error concealment: guess MVS + deblock


Are any of these settings hurting quality instead of helping? is there any setting i can/should increase or enable to increase it?

Ps it looks like Libvacodec supports more codecs than are selectable in ffdshow.
The ones i noticed because i personally want to use them is realmedia 3.0 up to 4.0

[clsid]

All "automatic quality control" does is lower the amount of postprocessing if CPU usage is high.

Leave all those decoder options at their defaults. There is no benefit in changing them.

FFmpeg does not have working decoders for RV30/RV40 yet. Also, many of the formats that libavcodec supports can not be played in DirectShow, due to the lack of source filters for the files that contain those formats, so it would be pointless to add them to ffdshow.

[pitch.fr]

Is there a way to get artifacts-free deinterlace in ffdshow on NTSC Video 29.97fps DVD ?

I've read quite a lot about interlaced streams, and basically the 2 fields you get are not from the same original frame.....so perfect IVTC requires to do some time based computing and put back each FULL frame together by putting them back in the right order.

and it seems that BOB based deinterlacing schemes basically divide the vertical resolution/2, so blending is not quite an option as it ruins the original sharpness ?!

which leads to my original question : is there a way to get artifacts-free deinterlace on 29.97 NTSC video DVD ? apart from doing some dirty blending that will create artifacts and divide the vertical resolution/2 ? :(

I've tried the BOB thingies with a refresh rate that's the double of the frame rate, but that didn't yield good results either.

I've also tried several Avisynth deinterlacers......they prolly work really well, but they're really not meant for real time use :rolleyes:

[Octo-puss]

I noticed there's new type of package at downloads - albain's builds. How is that different from others?

[clsid]

It isn't different.

[Octo-puss]

What's the point then? Guess I'm missing something.

[pitch.fr]

well it's ICL10, and it's newer than the latest from clsid.

ICL10 builds are significantly faster on a C2D/C2Q :)

[Octo-puss]

I assume ICL10 is some kind of compiler?... programmer's voodoo or like that? :D


guess it works just as well as other builds on non-C2Ds?