On 9/17/16, Michael Niedermayer <mich...@niedermayer.cc> wrote: > On Sat, Sep 17, 2016 at 03:06:57PM +0200, Carl Eugen Hoyos wrote: >> 2016-09-15 20:48 GMT+02:00 Michael Niedermayer <mich...@niedermayer.cc>: >> > On Thu, Sep 15, 2016 at 03:47:49PM +0200, Carl Eugen Hoyos wrote: >> >> 2016-09-15 15:42 GMT+02:00 Michael Niedermayer >> >> <mich...@niedermayer.cc>: >> >> >> - { AV_CODEC_ID_ADPCM_SWF, ('S' << 8) + 'F' }, >> >> >> /* HACK/FIXME: Does Vorbis in WAV/AVI have an (in)official ID? >> >> >> */ >> >> >> { AV_CODEC_ID_VORBIS, ('V' << 8) + 'o' }, >> >> >> { AV_CODEC_ID_NONE, 0 }, >> >> > >> >> > does this affect adpcm_swf in nut ? >> >> >> >> Yes, indeed. >> >> >> >> Is it possible to fix adpcm_swf in wav? >> > >> > maybe if block_align is set (and is constant) >> >> Attached fixes encoding, decoding still has the issue >> that the decoder doesn't like more than one packet. >> (Iiuc, it works if I limit the packet size in the wav demuxer.) >> >> Carl Eugen > >> adpcmenc.c | 1 + >> 1 file changed, 1 insertion(+) >> dccfa2d54e899c17dcb57b8e37658955dfa6ea9c >> 0001-lavc-adpcmenc-Set-block_align-for-adpcm_swf.patch >> From 168bc2f828ffd7f85c0db736e6097e4198854c5d Mon Sep 17 00:00:00 2001 >> From: Carl Eugen Hoyos <ceho...@ag.or.at> >> Date: Sat, 17 Sep 2016 15:01:19 +0200 >> Subject: [PATCH] lavc/adpcmenc: Set block_align for adpcm_swf. >> >> --- >> libavcodec/adpcmenc.c | 1 + >> 1 file changed, 1 insertion(+) > > LGTM > > can you add a fate test?
This patch is invalid. Block align should be set to what ADPCM_SWF uses in adpcm_encode_frame as pkt_size. And ADPCM_SWF should not be special, it should use block_align as pkt_size. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel