Re: [FFmpeg-devel] [PATCH] matroskadec.c: New patch for palettized QuickTime video (old one broken)
On 12/13/2015 09:02 PM, Mats Peterson wrote: On 12/13/2015 08:48 PM, Mats Peterson wrote: I just discovered that I've been incorrectly using the "i" loop variable in my patch, which is actually used by a track element loop at line 1718 in matroskadec.c. This will works as long as there is only video in the Matroska file, that's why I didn't discover it until I tried one with audio as well (all the samples so far are video only). Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel I forgot to restore the lower limit of A_QUICKTIME private data to 21 like Michael Niedermayer proposed. Sorry (it doesn't make much sense to me, but at least that broken file will pass.) Mats V_QUICKTIME (video) of course, not A_QUICKTIME. Mats -- Mats Peterson http://matsp888.no-ip.org/~mats/ ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] matroskadec.c: New patch for palettized QuickTime video (old one broken)
On 12/13/2015 08:48 PM, Mats Peterson wrote: I just discovered that I've been incorrectly using the "i" loop variable in my patch, which is actually used by a track element loop at line 1718 in matroskadec.c. This will works as long as there is only video in the Matroska file, that's why I didn't discover it until I tried one with audio as well (all the samples so far are video only). Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel I forgot to restore the lower limit of A_QUICKTIME private data to 21 like Michael Niedermayer proposed. Sorry (it doesn't make much sense to me, but at least that broken file will pass.) Mats -- Mats Peterson http://matsp888.no-ip.org/~mats/ ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] matroskadec.c: New patch for palettized QuickTime video (old one broken)
On 12/13/2015 09:16 PM, Mats Peterson wrote: On 12/13/2015 09:05 PM, Mats Peterson wrote: On 12/13/2015 09:02 PM, Mats Peterson wrote: On 12/13/2015 08:48 PM, Mats Peterson wrote: I just discovered that I've been incorrectly using the "i" loop variable in my patch, which is actually used by a track element loop at line 1718 in matroskadec.c. This will works as long as there is only video in the Matroska file, that's why I didn't discover it until I tried one with audio as well (all the samples so far are video only). Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel I forgot to restore the lower limit of A_QUICKTIME private data to 21 like Michael Niedermayer proposed. Sorry (it doesn't make much sense to me, but at least that broken file will pass.) Mats V_QUICKTIME (video) of course, not A_QUICKTIME. Mats Well, probably it will NOT pass, since I'm looking for the bit depth at offset 82 of the private data, which will of course be a totally bogus value with a private data size of 21. Oh well... Mats In any case, I've uploaded a sample file with both palettized video (Apple Graphics/SMC) and PCM audio called "montage.mkv" to both http://matsp888.no-ip.org/~mats/qtpalette-test and http://bit.ly/1I1Bw2I . Mats -- Mats Peterson http://matsp888.no-ip.org/~mats/ ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] matroskadec.c: New patch for palettized QuickTime video (old one broken)
On 12/13/2015 09:05 PM, Mats Peterson wrote: On 12/13/2015 09:02 PM, Mats Peterson wrote: On 12/13/2015 08:48 PM, Mats Peterson wrote: I just discovered that I've been incorrectly using the "i" loop variable in my patch, which is actually used by a track element loop at line 1718 in matroskadec.c. This will works as long as there is only video in the Matroska file, that's why I didn't discover it until I tried one with audio as well (all the samples so far are video only). Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel I forgot to restore the lower limit of A_QUICKTIME private data to 21 like Michael Niedermayer proposed. Sorry (it doesn't make much sense to me, but at least that broken file will pass.) Mats V_QUICKTIME (video) of course, not A_QUICKTIME. Mats Well, probably it will NOT pass, since I'm looking for the bit depth at offset 82 of the private data, which will of course be a totally bogus value with a private data size of 21. Oh well... Mats -- Mats Peterson http://matsp888.no-ip.org/~mats/ ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
