Issue 64 in ffmbc: input Matrox avi problem
ff...@googlecode.com
Owner: ----
Labels: Type-Defect Priority-Medium
New issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Is not possible convert completely a video file format MATROX, because the
output file duration is smaller than the input file.
I noticed that in the input description the duration of the video is not
the real one
In the test below I have converted video and audio with the same duration.
Command line:
ffmbc -i Matrox_Mpeg2_iframe_15Mbs.avi -i pcm_s16le_1536kbs.wav -target
imx50 IMX50_From_Matrox.mxf
Output:
FFmbc version 0.6.1
Copyright (c) 2008-2011 Baptiste Coudurier and the FFmpeg developers
[avi @ 0x252dbf0] non-interleaved AVI
[mpeg2video @ 0x25305d0] mpeg_decode_postinit() failure
Input #0, avi, from 'Matrox_Mpeg2_iframe_15Mbs.avi':
Metadata:
encoder: Matrox MQSink Filter For MPEG. Format: 4. Build: 2.0.0.271
Duration: 00:18:22.84, start: 0.000000, bitrate: 103317 kb/s
Stream #0.0(und): Video: mpeg2video, yuv422p, 720x576i tff [PAR 16:15
DAR 4:3], 14999 kb/s, 25.00 fps
Input #1, wav, from 'pcm_s16le_1536kbs.wav':
Duration: 02:06:05.64, start: 0.000000, bitrate: 1536 kb/s
Stream #1.0(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
auto-rescaling to IMX resolution
[pad @ 0x2534d90] w:720 h:576 -> w:720 h:608 x:0 y:32 color:0x108080FF[yuva]
Warning, QMAT_SHIFT is larger than 21, overflows possible
Last message repeated 1 times
Output #0, mxf_d10, to 'IMX50_From_Matrox.mxf':
Metadata:
encoder: FFmbc 0.6.1
Stream #0.0(und): Video: mpeg2video, yuv422p, 720x608i tff [PAR 152:135
DAR 4:3], cbr, 50000 kb/s, 25.00 fps
Stream #0.1(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
Stream mapping:
Stream #0.0 -> #0.0
Stream #1.0 -> #0.1
Press [q] to stop encoding
[mpeg2video @ 0x25305d0] mpeg_decode_postinit() failure
frame= 1413 fps=126 q=1.0 Lsize= 431324kB time=00:00:56.44
bitrate=62595.8kbits
video:344971kB audio:10584kB global headers:0kB muxing overhead 21.310178%
I'm using Ubuntu 10.10 (64bit)
I was compiling with --enable-gpl (GCC 4.4.5)
ff...@googlecode.com
Comment #1 on issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I used your newer version (0.7-rc3) ,but the errors are duplicated
Command line:
ffmbc -i Matrox_Mpeg2_iframe_15Mbs.avi -i pcm_s16le_1536kbs.wav -target
imx50 IMX50_From_Matrox.mxf
Output:
FFmbc version 0.7-rc3
Copyright (c) 2008-2011 Baptiste Coudurier and the FFmpeg developers
[avi @ 0x287e600] non-interleaved AVI
[mpeg2video @ 0x2880680] mpeg_decode_postinit() failure
Input #0, avi, from 'Matrox_Mpeg2_iframe_15Mbs.avi':
Metadata:
encoder: Matrox MQSink Filter For MPEG. Format: 4. Build: 2.0.0.271
Duration: 00:18:22.84, start: 0.000000, bitrate: 103317 kb/s
Stream #0.0(und): Video: mpeg2video (4:2:2), yuv422p, 720x576i tff [PAR
16:15 DAR 4:3], 14999 kb/s, 25.00 fps
Input #1, wav, from 'pcm_s16le_1536kbs.wav':
Duration: 02:06:05.64, start: 0.000000, bitrate: 1536 kb/s
Stream #1.0(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
File 'IMX50_From_Matrox.mxf' already exists. Overwrite ? [y/N] y
auto-rescaling to IMX resolution
[pad @ 0x2886720] w:720 h:576 -> w:720 h:608 x:0 y:32 color:0x108080FF[yuva]
Warning, QMAT_SHIFT is larger than 21, overflows possible
Last message repeated 1 times
Output #0, mxf_d10, to 'IMX50_From_Matrox.mxf':
Metadata:
encoder: FFmbc 0.7
Stream #0.0(und): Video: mpeg2video, yuv422p, 720x608i tff [PAR 152:135
DAR 4:3], cbr, 50000 kb/s, 25.00 fps
Stream #0.1(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
Stream mapping:
Stream #0.0 -> #0.0
Stream #1.0 -> #0.1
Press [q] to stop, [?] for help
[mpeg2video @ 0x2880680] mpeg_decode_postinit() failure
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] Warning MVs not available
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
frame= 63 fps= 0 q=1.0 size= 18946kB time=00:00:02.49
bitrate=62181.8kbits/s eta=00:03:43.11
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] Warning MVs not available
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
frame= 127 fps=126 q=1.0 size= 38498kB time=00:00:05.05
bitrate=62376.5kbits/s eta=00:03:38.59
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
frame= 191 fps=126 q=1.0 size= 58050kB time=00:00:07.61
bitrate=62440.4kbits/s eta=00:03:37.21
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2880680] ac-tex damaged at 44 35
[mpeg2video @ 0x2880680] concealing 45 DC, 45 AC, 45 MV errors
frame= 264 fps=125 q=1.0 Lsize= 80317kB time=00:00:10.49
bitrate=62686.1kbits/s 00:03:39.21
video:64453kB audio:1968kB global headers:0kB muxing overhead 20.920184%
ff...@googlecode.com
Status: CannotFix
Comment #2 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
This file seems damaged. In any case I need the file to confirm the issue
ff...@googlecode.com
Comment #3 on issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I looked for another file Matrox because the first one was too big.
How can I get you the file with size 300Mb, it is the smallest file that I
have.
Command line:
ffmbc -i Matrox_Mpeg2_iframe_15Mbs.avi -i pcm_s16le_1536kbs.wav -target
imx50 IMX50_From_Matrox.mxf
Output:
FFmbc version 0.7-rc4
Copyright (c) 2008-2011 Baptiste Coudurier and the FFmpeg developers
[avi @ 0x208e600] non-interleaved AVI
[mpeg2video @ 0x2090680] mpeg_decode_postinit() failure
Input #0, avi, from 'Matrox_Mpeg2_iframe_15Mbs.avi':
Metadata:
encoder: Matrox MQSink Filter For MPEG. Format: 4. Build: 2.0.0.271
Duration: 00:02:32.24, start: 0.000000, bitrate: 15069 kb/s
Stream #0.0(und): Video: mpeg2video (4:2:2), yuv422p, 720x576i tff [PAR
16:15 DAR 4:3], 14999 kb/s, 25.00 fps
Input #1, wav, from 'pcm_s16le_1536kbs.wav':
Duration: 00:02:32.24, start: 0.000000, bitrate: 1536 kb/s
Stream #1.0(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
File 'IMX50_From_Matrox.mxf' already exists. Overwrite ? [y/N] y
Auto-rescaling to imx50 resolution
[pad @ 0x2096720] w:720 h:576 -> w:720 h:608 x:0 y:32 color:0x108080FF[yuva]
Warning, QMAT_SHIFT is larger than 21, overflows possible
Last message repeated 1 times
Output #0, mxf_d10, to 'IMX50_From_Matrox.mxf':
Metadata:
encoder: FFmbc 0.7
Stream #0.0(und): Video: mpeg2video, yuv422p, 720x608i tff [PAR 152:135
DAR 4:3], cbr, 50000 kb/s, 25.00 fps
Stream #0.1(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
Stream mapping:
Stream #0.0 -> #0.0
Stream #1.0 -> #0.1
Press [q] to stop, [?] for help
[mpeg2video @ 0x2090680] mpeg_decode_postinit() failure
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] Warning MVs not available
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] Warning MVs not available
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
frame= 59 fps= 0 q=1.0 size= 17724kB time=00:00:02.32
bitrate=62440.6kbits/s eta=00:00:32.83
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
frame= 121 fps=120 q=1.0 size= 36360kB time=00:00:04.80
bitrate=62053.6kbits/s eta=00:00:31.06
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
frame= 184 fps=121 q=1.0 size= 55912kB time=00:00:07.33
bitrate=62412.9kbits/s eta=00:00:30.00
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
[mpeg2video @ 0x2090680] ac-tex damaged at 44 35
[mpeg2video @ 0x2090680] concealing 45 DC, 45 AC, 45 MV errors
frame= 262 fps=121 q=1.0 Lsize= 79698kB time=00:00:10.41
bitrate=62712.8kbits/s 00:00:29.66
video:63965kB audio:1952kB global headers:0kB muxing overhead 20.906194%
ff...@googlecode.com
Comment #4 on issue 64 by plkmaru...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I noticed that HD Matrox intraframe wrapped in AVI are not recognized at
all:
Stream #0.0(und): Video: unknown (M705), 1920x1080p, 25fps
Regards
ff...@googlecode.com
Comment #5 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Sorry, I need the file anyway to be able to look at it.
ff...@googlecode.com
Comment #6 on issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
How can I get you the file with size 300Mb?
The Max. attachments per comment is 10.0 MB
ff...@googlecode.com
Comment #7 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Upload the file to medafire or something like that, and mail me the link.
ff...@googlecode.com
Comment #8 on issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I noticed that the version 6.1 doesn't create image artefacts.
ff...@googlecode.com
Status: Accepted
Comment #9 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I'll have a look, thanks for the sample
ff...@googlecode.com
Comment #10 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Btw, can you share a small sample with M705 fourcc ? So I can add support.
ff...@googlecode.com
Comment #11 on issue 64 by plkmaru...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Here you go: http://min.us/mbgSk2nfPY, a very small Matrox M705 file.
1920x1080, 50i
ff...@googlecode.com
Comment #12 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
SD files should be supported now. HD is different I need to investigate
more.
ff...@googlecode.com
Comment #13 on issue 64 by rebaglia...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi Baptiste, Thanks! now it's fine.
But if I use a much longer file, ffmbc doesn't read the correct duration
and it doesn't encode the entire file.
Below you can see the output of the test.
Command Line:
ffmbc -v 3 -i Test.avi -i Test.wav -target imx50 -tff -aspect 4:3 -acodec
pcm_s16le Test.mxf -map_audio_channel 1:0:0:0:1:0 -map_audio_channel
1:0:1:0:1:1 -map_audio_channel 1:0:0:0:1:2 -map_audio_channel 1:0:1:0:1:3
Output:
FFmbc version 0.7-rc5
Copyright (c) 2008-2011 Baptiste Coudurier and the FFmpeg developers
[avi @ 0x1d94600] non-interleaved AVI
[mpeg2video @ 0x1d96620] mpeg_decode_postinit() failure
Input #0, avi, from 'Test.avi':
Metadata:
encoder: Matrox MQSink Filter For MPEG. Format: 4. Build: 2.0.0.271
Duration: 00:18:22.84, start: 0.000000, bitrate: 103317 kb/s
Stream #0.0(und): Video: mpeg2video (4:2:2), yuv422p, 720x576i tff [PAR
16:15 DAR 4:3], 14999 kb/s, 25.00 fps
Input #1, wav, from 'Test.wav':
Duration: 02:06:05.64, start: 0.000000, bitrate: 1536 kb/s
Stream #1.0(und): Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
Auto-rescaling to imx50 resolution
[pad @ 0x1d96060] w:720 h:576 -> w:720 h:608 x:0 y:32 color:0x108080FF[yuva]
Warning, QMAT_SHIFT is larger than 21, overflows possible
Last message repeated 1 times
Output #0, mxf_d10, to 'Test.mxf':
Metadata:
encoder: FFmbc 0.7
Stream #0.0(und): Video: mpeg2video, yuv422p, 720x608i tff [PAR 152:135
DAR 4:3], cbr, 50000 kb/s, 25.00 fps
Stream #0.1(und): Audio: pcm_s16le, 48000 Hz, 4 channels, s16, 3072 kb/s
Stream mapping:
Stream #0.0 -> #0.0
Stream #1.0 -> #0.1 [channel: 0 -> 0]
Stream #1.0 -> #0.1 [channel: 1 -> 1]
Stream #1.0 -> #0.1 [channel: 0 -> 2]
Stream #1.0 -> #0.1 [channel: 1 -> 3]
Press [q] to stop, [?] for help
[mpeg2video @ 0x1d96620] mpeg_decode_postinit() failure
frame= 288 fps=124 q=1.0 Lsize= 87681kB time=00:00:11.47
bitrate=62582.4kbits/s 00:03:40.01
video:70312kB audio:4304kB global headers:0kB muxing overhead 17.508255%
ff...@googlecode.com
Comment #14 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Can you please upload the file somewhere ? Thanks
ff...@googlecode.com
Comment #15 on issue 64 by ilf.mou...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi,
I working for a french television channel, and we also use AVI Matrox for
our old file.
I had to modify a codec to get the correct duration of the file.
The problem came with the structure of the AVI file. Matrox didn't respect
the avi specification (odml if I remember correctly) for the file that was
larger than 2Gb...
So they start like a normal avi file, but after the 9min 13s the video
stop...
That means that you arrive at the end of the normal "structure" of the avi
file...
The bug is inside FFMPEG, on the avidec.c
I was able to correct : LAVFilter for windows to be able to read a full
file and reencode it...
But when I try to do the same on FFMBC, my C skills wasn't enough....
If you wan't a file you will need a file bigger than 90000 frames (82000
something like that) and will have the bug.
I think that my correction in LAVFilter isn't a good correction... My C
skills or old but It work for me... I wasn't able to do it in ffmbc if you
want to look at it, I will put my avidec.c (of lavfilter) in attached file.
I add the method: avi_read_mmes
Which use the right index of the file.
Hope that help.
Attachments:
avidec.c 57.6 KB
ff...@googlecode.com
Comment #16 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Thanks, I'll have a look, ping me if I forget.
ff...@googlecode.com
Comment #17 on issue 64 by ilf.mou...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi,
Were you able to look at it ?
Thanks again.
ff...@googlecode.com
Comment #18 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
No, not yet, can you please send a diff ? It would be easier. Thanks,
ff...@googlecode.com
Comment #19 on issue 64 by cook.fon...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi,
we have applied the patch supplied by ilf to the sources of FFmbc version
0.7-rc8. Now we are able to read the index correctly (we hope!) but during
the execution we get a Segmentation fault. We have a diff file if it can be
use usefull.
Do you have news about this issue?
Thanks
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
ff...@googlecode.com
Comment #20 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Nope, this require some work to apply the patch correctly
ff...@googlecode.com
Comment #21 on issue 64 by fris...@yahoo.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi all.
I work for a major swiss broadcasting company and we started a project to
convert a huge bunch of Matrox MMES AVI files into MXF, so this issue is
quite important for us.
The patch proposed by ilf (see comment #15), while setting the correct
number of frames in the index (and by consequence, the correct duration of
the video stream), seems not to solve the whole problem, as stated by cook
(see comment #19).
After some days of analysing different AVI files (Matrox and not, big and
small), studying the specs and trying, it seems to me that "big" MMES files
are different from "standard big" files in the following ways:
1. the super index "index subtype" (index_sub_type) is 16 instead of 0;
2. same for the standard index
3. a standard index entry is 24 bytes long instead of 8, thus the
longs_pre_entry
value is 6 instead of 2; the semantic of the extra 4 longs is beyond my
comprehension, but at first sight they don't seem to be necessary.
I therefore came up with the attached patch. Perhaps I'm too optimistic,
but for now it seems to work. Use it at your own risk and post here your
findings.
Hopefully in the near future I'll start the massive conversion; I'll let
you know if the patch really works, so stay tuned...
Attachments:
avidec.patch 1.9 KB
ff...@googlecode.com
Comment #22 on issue 64 by fris...@yahoo.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
I forgot to mention that the above patch shall be applied to
libavformat/avidec.c of FFmbc-0.7-rc8
ff...@googlecode.com
Comment #23 on issue 64 by fris...@yahoo.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi all.
I work for a major swiss broadcasting company and we started a project to
convert a huge bunch of Matrox MMES AVI files into MXF, so this issue is
quite important for us.
The patch proposed by ilf (see comment #15), while setting the correct
number of frames in the index (and by consequence, the correct duration of
the video stream), seems not to solve the whole problem, as stated by cook
(see comment #19).
After some days of analysing different AVI files (Matrox and not, big and
small), studying the specs and trying, it seems to me that "big" MMES files
are different from "standard big" files in the following ways:
1. the super index "index subtype" (index_sub_type) is 16 instead of 0;
2. same for the standard index
3. a standard index entry is 24 bytes long instead of 8, thus the
longs_pre_entry value is 6 instead of 2;
(32bit);
the semantic of the additional field (or fields) is beyond my
comprehension, but at first sight it
doesn't seem to be necessary.
I therefore came up with the attached patch. Perhaps I'm too optimistic,
of FFmbc-0.7-rc8. I tested it against some files, ranging from a couple of
minutes (to assure that it's backward compatible) to almost 3 hours. Use it
Attachments:
avidec.patch 1.9 KB
ff...@googlecode.com
Comment #24 on issue 64 by ilf.mou...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi everyone...
I'm using ffmbc to generate a lot of transcoding for my work...
But I don't use directly Matrox avi file with ffmbc... I'm bypassing the
problem using LAVFilter (which I patched) to work with ffmbc using avisynth
script.. It's working but only in 32Bit...
I'm using it for 2 years now without any problems...
Hope this will help.
ff...@googlecode.com
Comment #25 on issue 64 by pa070pa...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
(Please see comment #23; I had to create a Google account since it's no
longer possible to use an external account)
Hello everybody.
As promised I'm back with some results
I wrote a bulk AVI to MXF conversion program; for each conversion, the
program tests the AVI file duration against the MXF one; so far, the
program has converted about 10'000 files, ranging from less than a minute
to a few hours.
The number of discrepancies found is... 0.
I didn't obviously looked at every file, just at some of them, and they
look OK.
I therefore think that the patch I submitted last November actually solves
the problem, but I would like to see these figures confirmed by other users.
@Ilf: I think this thread is to have FFmbc do the work; that said, is
always great to have alternatives, so thanks for your comment.
ff...@googlecode.com
Comment #26 on issue 64 by baptiste...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Nice, do you have other files I could validate the patch with ? Ill apply it
Thanks
ff...@googlecode.com
Comment #27 on issue 64 by pa070pa...@gmail.com: input Matrox avi problem
http://code.google.com/p/ffmbc/issues/detail?id=64
Hi Baptiste.
I sent you via Google Drive some links with the requested files.
Note that the AVI files are video-only; that's why the audio is in
separated files.
Files 399508* are a short clip, a little less than 2 minutes, used to test
the backward compatibility.
The other ones, 401605*, is a longer clip, more than 24 minutes, used to
test the patch: before applying it, ffprobe shows a duration of about 9:13,
while the patched version should show the correct one. ffmbc and other
programs should behave consequently.
Please keep these files for yourself, as the content might be copyrighted.
Please note that I submitted the patch in November, for the 0.7-rc8
version. I see that you released the 0.7 and the 0.7.1 versions; I've yet
to apply the patch to the newest release but I saw that it should be almost
the same as the November's one (just a matter of displacement in the source
file).
Keep up the good work ;-)