ffdshow tryouts project: Discussion & Development

[XhmikosR]

I don't like for ffdshow.ax to register after every build too. You can delete that or just exclude it from build.

[albain]

LPCM is already handled in FFDShow : it was inserted by Haruhiko for DVD audio which is LPCM 24 bits HD, the same as bluray hold.

Has anyone already used this feature ?

We are trying to make it work for bluray LPCM tracks but this does not work for now.

Any feedback would be appreciated.

Otherwise, I will need Sebastii's hands to use the dumper on PDVD9 and a LPCM bluray track

Thank you !

[tal.aloni]

Under ffdshow's video output, the tooltip for Standard under Input levels says that ffdshow automatically uses Full range for Fraps videos, even though it does not. The only way to get Full range is to select it, which is quite annoying having to change that setting back and forth, depending on what content is being decoded. Should be an easy fix, though.

fix commited at rev. 3147:
Bugfix: regarding Fraps, JPEG and MJPEG: when selecting standard input levels, and when "High quality YV12 to RGB conversion" was checked, ffdshow did not automatically used full range input levels for YV12 to RGB conversion.

[STaRGaZeR]

Thanks tal ;)

[rica]

LPCM is already handled in FFDShow : it was inserted by Haruhiko for DVD audio which is LPCM 24 bits HD, the same as bluray hold.

Has anyone already used this feature ?

We are trying to make it work for bluray LPCM tracks but this does not work for now.

Any feedback would be appreciated.

Otherwise, I will need Sebastii's hands to use the dumper on PDVD9 and a LPCM bluray track

Thank you !

I think what you are waiting for:

Here is the debug log from an original LPCM BD on Vista via PDVD9:

http://www.mediafire.com/?nyiunyjzvgy



_ _ _ _ _

[STaRGaZeR]

tal, your commit doesn't fix the issue completely. FRAPS videos now have proper levels, but the input levels options don't work anymore (only with FRAPS material, and maybe with MJPEG too, can't test).

[tal.aloni]

STaRGaZeR,
basically when dealing with FRAPS, the input level option
selects between the following 4:
+ Auto (= Full Range for fraps)
+ Standard (= Full Range for fraps)
+ Full Range
+ Custom

for Fraps, the only way to get something different than full range, is by using custom.
I can see that it doesn't work for Fraps ATM, and I'm on it.
is that what you meant?

Regards,
Tal

Edit: the above issue was fixed in rev. 3149

[albain]

I think what you are waiting for:

Thanks it is but I didn't see the traces from media structures.

I am not sure you have the right version of the dumper

Could you retry with the PM link I have sent you ?

Thanks

[STaRGaZeR]

STaRGaZeR,
basically when dealing with FRAPS, the input level option
selects between the following 4:
+ Auto (= Full Range for fraps)
+ Standard (= Full Range for fraps)
+ Full Range
+ Custom

for Fraps, the only way to get something different than full range, is by using custom.
I can see that it doesn't work for Fraps ATM, and I'm on it.
is that what you meant?

Regards,
Tal

Edit: the above issue was fixed in rev. 3149

Ah, there's something strage going on here. With Fraps v2.9.8 and r3149:

- YCbCr options all give the same output, can't say if the default is BT.601 or BT.709.

Input levels:

- Auto results in 0-255.
- Standard results in 0-255. <-- Wrong, should be 16-235. We're forcing standard levels here.
- Full range results in 0-255.
- Custom is working as it should.

- Output levels working fine

Fraps v3.0.1 and r3149:

Nothing in the RGB conversion tab works, it seems default is input 0-255 output 0-255.

"High quality YV12 to RGB" checked or unchecked, doesn't matter.

Keep in mind that anything other than Auto is for forcing those settings regardless of the codec and source used.

[tal.aloni]

- Standard results in 0-255. <-- Wrong, should be 16-235. We're forcing standard levels here.

Actually, according to the tooltip, it should be 0-255.
(standard levels can also imply standard in regard to the format)

[STaRGaZeR]

Actually, according to the tooltip, it should be 0-255.
(standard levels can also imply standard in regard to the format)

The tooltip for Standard says: "Nearly all videos use this. For JPEG, MJPEG and Fraps, ffdshow automatically uses full range".

Which is logical, since that's what Auto is for: Standard for everything except H.264 with the range flag, JPEG, MJPEG and Fraps. Then options for override the Auto decision.

Same for the YCbCr options: Auto will select what is usual, then you have the option to override it by using BT.601 or BT.709.

At least that's the way I see it. Have you corfirmed the v3.0.1 issue?

[tal.aloni]

1. I'm getting what you're saying, however, I wish to leave things as they are. I feel that the original intension was to force the user into making the right choice, and I'm ok with that. if someone really wants to display the wrong level for fraps, he can use the custom option.

2. There is indeed a problem with YCbCr specifications. I'm looking into it. (Edit: fixed in rev. 3151)

3. Can you please send a sample of v3.0.1 video so I could recreate the issue? Thanks!

Tal

[rica]

Thanks it is but I didn't see the traces from media structures.

I am not sure you have the right version of the dumper

Could you retry with the PM link I have sent you ?

Thanks

OK, dunno if it's helpful ? I mastered a new LPCM ISO but when i open it with debugger it crashes at the beginning. (PDVD9 on Vista 32 SP2)

Normally it gives 5.1 LPCM before forcing it to debugging. (That was the Vista trial, i'll add another one taken over 7/64 later and edit this post.)

http://www.mediafire.com/?2nwg0ym2jym


EDIT:

Here is the debug log created on 7/64; same thing: crash at the beginning but without debug i get 5.1 LPCM on my AVR. (Denon 1910) (PDVD9 on 7/64)

http://www.mediafire.com/?t2tmzlimkwl

BTW seems i don't need dpinpurr for 7; do i?

_ _ _ _

[STaRGaZeR]

1. I'm getting what you're saying, however, I wish to leave things as they are. I feel that the original intension was to force the user into making the right choice, and I'm ok with that. if someone really wants to display the wrong level for fraps, he can use the custom option.

2. There is indeed a problem with YCbCr specifications. I'm looking into it. (Edit: fixed in rev. 3151)

3. Can you please send a sample of v3.0.1 video so I could recreate the issue? Thanks!

Tal

Look here to see how ffdshow was before the creation of the Auto setting: http://forum.doom9.org/showthread.php?t=143598. These options forced their value in all cases. I still think that this is a bug.

YCbCr fixed. However after some testing I'd suggest that if Auto is selected and Fraps is detected, BT.709 is always used regardless of resolution. All non-lossless Fraps videos are BT.709. Please confirm this. I did it comparing with PNG captures in game. Be careful as below 1024x768 Fraps records in RGB24 and no YCbCr-->RGB conversion will be done, see below.

About the v3.0.1 thing. This is really funny :). Since v3.0 Fraps has an option to force RGB lossless recording (finally!). If you use that mode, the videos are recorded in RGB24 instead of YV12. ffdshow can decode these videos, but since they are already RGB24 no conversion is done and of course the RGB conversion options won't work. However there's a bug (maybe feature?) in Fraps: even if you tell the program to not use RGB lossless, below a certain resolution between 800x600 and 1024x768, videos will be recorded in RGB24. With lossless RGB unchecked 1024x768 results in YV12, but 800x600 results in RGB24. This is why the RGB conversion options weren't working for me with v3.0.1. Then I revert back to v2.9.8 and this thing is still here! Please record something at 640x480, you'll see how it is RGB24. You'll recognize it by looking at colorful edges and by RGB options not working :p

[tal.aloni]

However after some testing I'd suggest that if Auto is selected and Fraps is detected, BT.709 is always used regardless of resolution.

1. according to the wiki, and tooltips, standard is not 15-235, but a "semi-automatic" mode. I think it's easier for novice users, (if the reports of H.264 streams with wrong flag are true), so I'm not going to change it. (see the tooltip related to auto)

2. That's what I did, BT.709 will always be used for fraps on Auto, regardless of the resolution.

also, BT.601 will always be used for MJPEG on Auto, regardless of the resolution. (that's how it was so far)

the only thing left to do is to update the tooltip.

Thanks for testing,
Tal

[madshi]

if the reports of H.264 streams with wrong flag are true
They are definitely true, trust me on that.

[albain]

OK, dunno if it's helpful ? I mastered a new LPCM ISO but when i open it with debugger it crashes at the beginning. (PDVD9 on Vista 32 SP2)

Normally it gives 5.1 LPCM before forcing it to debugging. (That was the Vista trial, i'll add another one taken over 7/64 later and edit this post.)

http://www.mediafire.com/?2nwg0ym2jym


EDIT:

Here is the debug log created on 7/64; same thing: crash at the beginning but without debug i get 5.1 LPCM on my AVR. (Denon 1910) (PDVD9 on 7/64)

http://www.mediafire.com/?t2tmzlimkwl

BTW seems i don't need dpinpurr for 7; do i?

_ _ _ _

Great ! You got the structures in it : LPCM is sent as IEEE_FLOAT type (which is PCM 32 bits float)
Could you also make those tests on Vista for DTSHD and TrueHD with PDVD9 ?
This will tell us which structures to feed for bitstream support on Vista, which is the missing key to finish the job.
Let's continue this discussion on the other thread.
This is a very good news !:)

[STaRGaZeR]

1. according to the wiki, and tooltips, standard is not 15-235, but a "semi-automatic" mode. I think it's easier for novice users, (if the reports of H.264 streams with wrong flag are true), so I'm not going to change it. (see the tooltip related to auto)

2. That's what I did, BT.709 will always be used for fraps on Auto, regardless of the resolution.

also, BT.601 will always be used for MJPEG on Auto, regardless of the resolution. (that's how it was so far)

the only thing left to do is to update the tooltip.

Thanks for testing,
Tal

1. OK, then. MJPEG and JPEG use Full range if Standard is selected too? (EDIT: I see MJPEG uses Full range with Standard selected, so I assume JPEG does so too)

2. Thanks!

[tal.aloni]

1. yes.

You welcome,
Tal

[tal.aloni]

developers,
"High quality YV12 to RGB conversion" checkbox appears twice,
first in the "Output" section, and second in the "RGB conversion" section,

I suggest removing it from the output section, do you have any objection?

[clsid]

It is in two places for legacy reasons. Many online tutorials reference the one in Output. So I think it would be best to leave it there (for now).

[TFM_TheMask]

Am I correct that normal AC3 and DTS passthrough only works when liba52 and libdts are selected? When I select libavcodec for both then the output is not AC3 or DTS and is way to big.

I'm still using revision 3111 but in the svn revision I don't see that this is a fixed or a known issue.

[tal.aloni]

Am I correct that normal AC3 and DTS passthrough only works when liba52 and libdts are selected?

No, libavcodec works as well,
moreover, the process is exactly the same for passthrough whether libdts or libavcodec were selected.

[TFM_TheMask]

No, libavcodec works as well,
moreover, the process is exactly the same for passthrough whether libdts or libavcodec were selected.

That is strange. When I set the dts and ac3 registry setting (with my tool) to 1 (libavcodec) then it doesn't work and when I set them at 17 (libdts) and 15 (liba52) then it works. Could it be my older version (3111) then?

[tal.aloni]

TFM_TheMask,
No relevant changes have been made since 3111,
but please try the latest clsid build just to be sure,
maybe your build is broken somehow.

[TFM_TheMask]

TFM_TheMask,
No relevant changes have been made since 3111,
but please try the latest clsid build just to be sure,
maybe your build is broken somehow.

Ok, will try that.

:thanks:

[TFM_TheMask]

TFM_TheMask,
No relevant changes have been made since 3111,
but please try the latest clsid build just to be sure,
maybe your build is broken somehow.

Installed the latest 3154 revision and still the same issue. When I choose Libavcodec for DTS then the output is not DTS when DTSpassthrough.

[tal.aloni]

TFM_TheMask,
I reviewed the code, and have no logical explanation, sorry. (I already assumed you have tested a wide range of audio tracks)

[TFM_TheMask]

TFM_TheMask,
I reviewed the code, and have no logical explanation, sorry. (I already assumed you have tested a wide range of audio tracks)

Ok, thanks for looking. Will keep it on libdts then.

I also tested it in GraphStudio and I couldn't get s/pdif working for libavcodec. With libdts no problem.

[jmone]

I'm not sure if this is the correct thread as I don't know if it is FFDSHOW or another filter...but I have a problem with Subtitles stop being displayed after 10sec or so. If I deselect and reselect the subtitle it will be fine for another 10sec or so then stop displaying again. The video and audio play perfectly - the subs just stop displaying.

I'm not sure what diagnostic info to post but here is the info from the filter chain if that helps, and I'm using:
- ffdshow tryouts revision 3119 Oct 18 (edit - upgraded to 3154 and still the same)
- Haali Media Splitter 1.9.42.1
- ATI HD4550 - Lastest Catalyst - Win7x64 w/Aero Glass - JR Media Center

Thanks in advance!
Nathan

Filter Graph Info:

Filter 'Default DirectSound Device'
CLSID: {79376820-07D0-11CF-A24D-0020AFD79767}
Host: C:\Windows\SysWOW64\quartz.dll
Input Pin 'Audio Input pin (rendered)'
Connected to pin 'Output' of filter 'J. River Media Proxy Audio'
Major type MEDIATYPE_Audio Sub type MEDIASUBTYPE_PCM, Format type FORMAT_WaveFormatEx

Filter 'Enhanced Video Renderer'
CLSID: {FA10746C-9B63-4B6C-BC49-FC300EA5F256}
Host: C:\Windows\SysWOW64\evr.dll
Input Pin 'EVR Input0'
Connected to pin 'Out' of filter 'ffdshow Video Decoder'
Major type MEDIATYPE_Video Sub type MEDIASUBTYPE_YUY2, Format type FORMAT_VideoInfo2
Input Pin 'EVR Input1'

Filter 'J. River Media Proxy Audio'
CLSID: {A4207487-E7D8-406B-9882-B66370DE2492}
Host:
Input Pin 'Input'
Connected to pin 'Audio' of filter 'H:\Media4\Video\Movies\Letters from Iwo Jima (2006).mkv'
Major type MEDIATYPE_Audio Sub type MEDIASUBTYPE_PCM, Format type FORMAT_WaveFormatEx
Output Pin 'Output'
Connected to pin 'Audio Input pin (rendered)' of filter 'Default DirectSound Device'
Major type MEDIATYPE_Audio Sub type MEDIASUBTYPE_PCM, Format type FORMAT_WaveFormatEx

Filter 'ffdshow Video Decoder'
CLSID: {04FE9017-F873-410E-871E-AB91661A4EF7}
Host: C:\Program Files (x86)\ffdshow\ffdshow.ax
Input Pin 'In'
Connected to pin 'Output' of filter 'J. River Media Proxy Video'
Major type MEDIATYPE_Video Sub type Unknown GUID Name: {31435657-0000-0010-8000-00AA00389B71}, Format type FORMAT_VideoInfo2
Output Pin 'Out'
Connected to pin 'EVR Input0' of filter 'Enhanced Video Renderer'
Major type MEDIATYPE_Video Sub type MEDIASUBTYPE_YUY2, Format type FORMAT_VideoInfo2
Input Pin 'In Text'
Connected to pin 'Subtitle' of filter 'H:\Media4\Video\Movies\Letters from Iwo Jima (2006).mkv'
Major type Unknown GUID Name: {E487EB08-6B26-4BE9-9DD3-993434D313FD} Sub type Unknown GUID Name: {F7239E31-9599-4E43-8DD5-FBAF75CF37F1}, Format type Unknown GUID Name: {F7239E31-9599-4E43-8DD5-FBAF75CF37F1}
Input Pin 'In Text 2'

Filter 'J. River Media Proxy Video'
CLSID: {A4207487-E7D8-406B-9882-B66370DE2492}
Host:
Input Pin 'Input'
Connected to pin 'Video' of filter 'H:\Media4\Video\Movies\Letters from Iwo Jima (2006).mkv'
Major type MEDIATYPE_Video Sub type Unknown GUID Name: {31435657-0000-0010-8000-00AA00389B71}, Format type FORMAT_VideoInfo2
Output Pin 'Output'
Connected to pin 'In' of filter 'ffdshow Video Decoder'
Major type MEDIATYPE_Video Sub type Unknown GUID Name: {31435657-0000-0010-8000-00AA00389B71}, Format type FORMAT_VideoInfo2

Filter 'H:\Media4\Video\Movies\Letters from Iwo Jima (2006).mkv'
CLSID: {55DA30FC-F16B-49FC-BAA5-AE59FC65F82D}
Host: C:\Program Files (x86)\Haali\MatroskaSplitter\splitter.ax
Output Pin 'Video'
Connected to pin 'Input' of filter 'J. River Media Proxy Video'
Major type MEDIATYPE_Video Sub type Unknown GUID Name: {31435657-0000-0010-8000-00AA00389B71}, Format type FORMAT_VideoInfo2
Output Pin 'Audio'
Connected to pin 'Input' of filter 'J. River Media Proxy Audio'
Major type MEDIATYPE_Audio Sub type MEDIASUBTYPE_PCM, Format type FORMAT_WaveFormatEx
Output Pin 'Subtitle'
Connected to pin 'In Text' of filter 'ffdshow Video Decoder'
Major type Unknown GUID Name: {E487EB08-6B26-4BE9-9DD3-993434D313FD} Sub type Unknown GUID Name: {F7239E31-9599-4E43-8DD5-FBAF75CF37F1}, Format type Unknown GUID Name: {F7239E31-9599-4E43-8DD5-FBAF75CF37F1}

[tal.aloni]

jmone,
I'm assuming that the problem you're experiencing is constant to all videos.
I'm reluctant to believe this is a bug in ffdshow,
please try another video player / another renderer.

[jmone]

Thanks tal.aloni.

* The issue is constant on other files (which were all created the same way from original discs)

I've tried different combinations of the following but it had not improvement:
* EVR, VMR 9, VMR 7, Haali
* Haali Splitter, Simple Splitter, Direct

MPC-HC played it correctly but that was not using FFDSHOW

I've also now posted on the JRMedia Center forum, but any other ideas would be great.

Nathan

[jmone]

I think I found it. Un-Installed CCCP, manually installed the latest FFDSHOW, Haali Splitter, and VSFilter (I seemed to have an old VOBSUB prior to the VSFilter branch).
Thanks
Nathan

[tal.aloni]

Glad you found it.

p.s.
codec packs are bad for your playback health.

[jmone]

Given the advance in FFDSHOW features, is there any plans to directly support native Blu-ray / HD-DVD SUP format to prevent the need to change them to SUB/IDX?
Thanks
Nathan

[jmone]

Sorry - Noob Q - I know FFDSHOW now supports more of the Blu-ray / HD-DVD HD Audio streams but are there any limitations in decoding:
- EAC3 (DD+)
- DTS-HD
- DTS-HD Master Audio
- Dolby True HD

Thanks
Nathan

[whurlston]

DTS-HD Master Audio still cannot be decoded, only the DTS Core. DTS-HD/MA can be bitstreamed however with the bitstream branch (soon to be merged with the trunk) and a Xonar, ATI 5xxx series, or Auzentech HomeTheater HD.

[albain]

Now merged !

About BR subtitles it is the responsability of the splitter/navigation filter to make the subtitles available to the decoder

However BR subtitles are in a new format similar to DVD subs (i.e bitmaps subtitles but in JPG format for BR) not supported by FFDShow. This has to be done : I think that opensource code already exists for this subtitle format

[VipZ]

Hi, I have found a bug with ffshow AAC decoding + 32bit float output selected, I get no audio with the audio info page showing nothing. This bug appears in both x86/x64 with the new HD bitstreaming update. Was fine with pre HD bitstream rev 3149
Thanks

Pin In
Filter : ffdshow Audio Decoder - CLSID : {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}

- Connected to:

CLSID: {0A68C3B5-9164-4A54-AFAF-995B2FF0E0D4}
Filter: D:\test.mkv
Pin: Undetermined (Audio 1)

- Connection media type:

Audio: AAC 48000Hz stereo

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: Unknown GUID Name {000000FF-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 256000
cbFormat: 23

WAVEFORMATEX:
wFormatTag: 0x00ff
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 0
nBlockAlign: 0
wBitsPerSample: 0
cbSize: 5 (extra bytes)

pbFormat:
0000: ff 00 02 00 80 bb 00 00 00 00 00 00 00 00 00 00 ÿ...€»..........
0010: 05 00|11 90 56 e5 00 ...�Vå.



Pin Out

Filter : ffdshow Audio Decoder - CLSID : {0F40E1E5-4F79-4988-B1A9-CC98794E6B55}

- Connected to:

CLSID: {18C16B08-6497-420E-AD14-22D21C2CEAB7}
Filter: Audio Switcher
Pin: test.mkv / Undetermined (Audio 1)

- Connection media type:

Audio: IEEE Float 48000Hz stereo 3072kbps

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Audio {73647561-0000-0010-8000-00AA00389B71}
subtype: MEDIASUBTYPE_IEEE_FLOAT {00000003-0000-0010-8000-00AA00389B71}
formattype: FORMAT_WaveFormatEx {05589F81-C356-11CE-BF01-00AA0055595A}
bFixedSizeSamples: 1
bTemporalCompression: 0
lSampleSize: 1
cbFormat: 40

WAVEFORMATEX:
wFormatTag: 0x0003
nChannels: 2
nSamplesPerSec: 48000
nAvgBytesPerSec: 384000
nBlockAlign: 8
wBitsPerSample: 32
cbSize: 22 (extra bytes)

pbFormat:
0000: 03 00 02 00 80 bb 00 00 00 dc 05 00 08 00 20 00 ....€»...Ü.... .
0010: 16 00|00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0020: 00 00 00 00 00 00 00 00 ........

- Enumerated media type 0:

Set as the current media type

- Enumerated media type 1:

Unknown

AM_MEDIA_TYPE:
majortype: MEDIATYPE_Stream {E436EB83-524F-11CE-9F53-0020AF0BA770}
subtype: MEDIASUBTYPE_None {E436EB8E-524F-11CE-9F53-0020AF0BA770}
formattype: FORMAT_None {0F6417D6-C318-11D0-A43F-00A0C9223196}
bFixedSizeSamples: 0
bTemporalCompression: 0
lSampleSize: 307200
cbFormat: 0

[albain]

Hi, I have found a bug with ffshow AAC decoding + 32bit float output selected, I get no audio with the audio info page showing nothing. This bug appears in both x86/x64 with the new HD bitstreaming update. Was fine with pre HD bitstream rev 3149
Thanks


Hi, can you provide me the sample, or better, debugview logs ?

Thank you

Damien

EDIT : damm it, I don't have a 32 bits float compatible equipment to test
You also have the issue with other bit depths ?
Which would help me a lot is a debugview log with a working build and another one without
To get logs, go to the following registry key
HKEY_CURRENT_USER\Software\GNU\ffdshow
and add a DWORD key named "allowDPRINTF" and set to 1
You can disable the debug logs by setting it back to 0 after that

[VipZ]

@ albain, as requested, debugview logs.
I select all output formats and let ffdshow choose which to use. If I force to 24 or 32bit int the audio plays fine, so it seems its only 32bit float affected. 32bit float is also working for MP3.

rev 3156
http://www.mediafire.com/?ytmwzjyqoyd

rev 3160
http://www.mediafire.com/?gjmnajzmtzm

Thanks

[STaRGaZeR]

albain,

I've been using the lastest beta of the Haali splitter which has support for TrueHD in Matroska. However I'm having problems: if a MKV file has both AC3 (audio track 1) and TrueHD (audio track 2), when the file starts the AC3 audio plays fine. Then I change to the TrueHD track, it plays fine too. But if I go back to the AC3 track, all I get is silence. Info & CPU shows nothing, and in MPC-HC right click-->Filters-->ffdshow Audio Decoder-->the libavcodec truehd decoder is still used, this is probably why I can't hear any sound: ffdshow tries to decode AC3 with the TrueHD decoder. Maybe a bug in ffdshow's parser?

If you need a sample let me know.

[albain]

@ albain, as requested, debugview logs.
I select all output formats and let ffdshow choose which to use. If I force to 24 or 32bit int the audio plays fine, so it seems its only 32bit float affected. 32bit float is also working for MP3.

rev 3156
http://www.mediafire.com/?ytmwzjyqoyd

rev 3160
http://www.mediafire.com/?gjmnajzmtzm

Thanks

Thanks, this is fixed in revision 3161

@Stargazer : indeed, this is due to the limitation in the formats changes I added. If the codec changes more than twice, I don't take in account further changes.
This rule should be reset in case of a stream switching, but how to detect it ?
The problem is how to make the difference between a skip and a stream switching (resulting both in a newsegment call)
Maybe someone has a clue

[VipZ]

Thanks for the quick fix, as always awesome work :)

[rpm7200]

libavcodec vs libdts or libavcodec vs liba52 ? is libavcodec better?

[ikarad]

DTS-HD Master Audio still cannot be decoded, only the DTS Core. DTS-HD/MA can be bitstreamed however with the bitstream branch (soon to be merged with the trunk) and a Xonar, ATI 5xxx series, or Auzentech HomeTheater HD.

I have xfi titanium and 5.1 audio kit. Can I decode DTS-HD from blu-ray with ffdshow ?

Anybody can explain me what is hdbitstream branch ?

[STaRGaZeR]

@Stargazer : indeed, this is due to the limitation in the formats changes I added. If the codec changes more than twice, I don't take in account further changes.
This rule should be reset in case of a stream switching, but how to detect it ?
The problem is how to make the difference between a skip and a stream switching (resulting both in a newsegment call)
Maybe someone has a clue

Dunno, maybe an option to override that only-twice decision?

Also, when watching DVDs (AC3 decoded with liba52) with MPC-HC and random seeking in the movie the decoder sometimes changes to eac3. eac3 decoder works for AC3, but this should not happen. I notice this specially when I do a new seek after the change, since the eac3 decoder has the same "issue" the libav ac3 decoder has: high pitch sound for a short period of time after a seek.

[albain]

I have xfi titanium. Can I decode DTS-HD with ffdshow ?

Anybody can explain me what is hdbitstream branch ?

A development branch : I made the developements on a branch apart of the trunk (main development line). See subversion for more info

auzentech xfi ? No auzentech had the *good* idea to encrypt the stream between the software and the driver.
Unless we get to know the encryption key and algorithm (this is possible), there is nothing to do for them. I advise you to grab one of the new or future radeon card

[albain]

Dunno, maybe an option to override that only-twice decision?

Also, when watching DVDs (AC3 decoded with liba52) with MPC-HC and random seeking in the movie the decoder sometimes changes to eac3. eac3 decoder works for AC3, but this should not happen. I notice this specially when I do a new seek after the change, since the eac3 decoder has the same "issue" the libav ac3 decoder has: high pitch sound for a short period of time after a seek.

I guess this is due to a loss of sync where the ffdshow parser is looking for a syncword and finds something that looks like an EAC3 syncword whereas it is only data. This occurs often ? Because there are few chances you get into that situation

[STaRGaZeR]

Yes, you could say it happens often. Each time you seek you have that possibility. Pure luck.