[Ffmpeg-devel-irc] ffmpeg-devel.log.20191201

Mon Dec 2 03:05:03 EET 2019

[00:21:18 CET] <thardin> 420 is the weed number
[12:15:29 CET] <ubitux> sorry i haven't been here for a while, is "for (int i ..." allowed now?
[12:15:41 CET] <ubitux> is it limited to the for loop or can we do modern c declaration too?
[12:16:25 CET] <durandal_1707> it is allowed, please follow our discussion and developer manual
[12:16:52 CET] <ubitux> yeah sorry the first one was a rethorical question, i'm seeing the thousands of matches, that's nice
[12:17:14 CET] <durandal_1707> second one, is not because it is ugly
[12:17:31 CET] <thardin> considering c99-to-c89 exists it shouldn't be a problem
[12:17:54 CET] <ubitux> durandal_1707: opinion, but ok :)
[12:27:09 CET] <durandal_1707> ubitux: use { }
[14:01:23 CET] <durandal_1707> ubitux: i'm not goint to send new nlmeans iteration, very bad experience
[14:30:26 CET] <ubitux> durandal_1707: what do you mean?
[14:33:01 CET] <durandal_1707> lost interest, i'm not going to wait days for negative review
[16:42:59 CET] <kierank> durandal_1707: could be bug
[16:54:20 CET] <durandal_1707> kierank: what?
[16:55:08 CET] <kierank> I don't see how they make braw 420
[16:56:03 CET] <durandal_1707> 64 + 64 + 32 + 32: coded coefficients in this order
[16:56:58 CET] <durandal_1707> perhaps last is just split...
[16:57:22 CET] <durandal_1707> so its 64+64+(32+32)
[16:57:29 CET] <durandal_1707> but that is very weird
[16:57:44 CET] <kierank> yes odd
[17:10:49 CET] <cone-519> ffmpeg 03Steven Liu 07master:b26225a3c710: avformat/dashenc: remove unused check of avformat_free_context
[17:10:50 CET] <cone-519> ffmpeg 03Steven Liu 07master:e880f4fb387b: avformat/hdsenc: removed unused check of avformat_free_context
[17:10:51 CET] <cone-519> ffmpeg 03Steven Liu 07master:0f79a71353fa: avformat/rtpenc_mpegts: removed unused check of avformat_free_context
[17:10:52 CET] <cone-519> ffmpeg 03Steven Liu 07master:9cc88ed4b71c: avformat/smoothstreamingenc: removed unused check of avformat_free_context
[17:13:02 CET] <Boist> 174
[17:20:17 CET] <cone-519> ffmpeg 03Andreas Rheinhardt 07master:35005a4af1eb: avformat/flac_picture: Simplify checks
[17:20:18 CET] <cone-519> ffmpeg 03Andreas Rheinhardt 07master:84a4261cd842: avformat/flac_picture: Switch to bytestream2 API
[17:20:19 CET] <cone-519> ffmpeg 03Andreas Rheinhardt 07master:5946243fa893: avformat/flac_picture: Return directly if nothing has been allocated
[17:20:20 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:13816a1d085f: avformat/mxfdec: Clear metadata_sets_count in mxf_read_close()
[17:20:21 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:47d963335eb2: avcodec/vmdaudio: Check chunk counts to avoid integer overflow
[17:20:22 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:0e010e489b70: avcodec/vc1_block: Fix integer overflow in AC rescaling in vc1_decode_i_block_adv()
[17:20:23 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:589cb44498b5: avcodec/wmaprodec: Fix buflen computation in save_bits()
[17:20:24 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:7686ba1f149a: avcodec/alac: Fix integer overflow in lpc_prediction() with sign
[17:20:25 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:d468da8d7960: avcodec/g729dec: Check for KELVIN && 6k4
[17:20:26 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:f64be9da4c8b: avcodec/g729dec: require buf_size to be non 0
[17:20:27 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:576746b4e300: avcodec/g729dec: Factor block_size out
[17:20:28 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:336f9461df7d: avcodec/g729dec: Avoid using buf_size
[17:20:29 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:0ddef0045737: avcodec/g729dec: Avoid one multiply by using init_get_bits8()
[17:20:30 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:fd3c34ff304a: avcodec/alsdec: Avoid 1 layer of pointer dereferences in INTERLEAVE_OUTPUT()
[17:20:31 CET] <cone-519> ffmpeg 03Michael Niedermayer 07master:a11aa5f3ed7e: avcodec/alsdec: Discard frames for which no channel could be decoded
[18:57:18 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:e2def5f382b0: avcodec/cngdec: Remove AV_CODEC_CAP_DELAY
[18:57:19 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:650ce5047cf3: avutil/lfg: Correct index increment type to avoid undefined behavior
[18:57:20 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:f3e553e36a98: avcodec/rawdec: Check bits_per_coded_sample more pedantically for 16bit cases
[18:57:21 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:66e701af697d: avcodec/wmavoice: Fix integer overflow in synth_frame()
[18:57:22 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:9162df98e1cf: avcodec/nuv: Move comptype check up
[18:57:23 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:3f186986520f: avcodec/mxpegdec: Check for multiple SOF
[18:57:24 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:a67d997ad735: avcodec/g729dec: Use 64bit and clip in scalar product
[18:57:25 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:6ce537910543: avcodec/ralf: Fix integer overflows with the filter coefficient in decode_channel()
[18:57:26 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:ed4f48535335: avcodec/ffwavesynth: Fix integer overflow with pink_ts_cur/next
[18:57:27 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:db659cdf4789: avcodec/nuv: Use ff_set_dimensions()
[18:57:28 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:0eff731d8d35: avformat/mxfdec: Clear metadata_sets_count in mxf_read_close()
[18:57:29 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:2f942ddf7415: avcodec/vmdaudio: Check chunk counts to avoid integer overflow
[18:57:30 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:5b43aa7a80f5: avcodec/vc1_block: Fix integer overflow in AC rescaling in vc1_decode_i_block_adv()
[18:57:31 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:f4daf42c1a5c: avcodec/wmaprodec: Fix buflen computation in save_bits()
[18:57:32 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:4d932eb66b03: avcodec/alac: Fix integer overflow in lpc_prediction() with sign
[18:57:33 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:0b49c74fe177: avcodec/g729dec: require buf_size to be non 0
[18:57:34 CET] <cone-519> ffmpeg 03Michael Niedermayer 07release/3.4:289a79d545e8: Changelog: update
[19:47:27 CET] <mateo`> <èè
[19:47:31 CET] <mateo`> oups
[20:59:59 CET] <cone-519> ffmpeg 03James Almer 07master:eced91afa5fc: avcodec/cbs_av1: implement missing set_frame_refs() function
[21:00:00 CET] <cone-519> ffmpeg 03James Almer 07master:553c1431ac7d: Revert "avcodec/cbs_av1_syntax_template: Check ref_frame_idx before use"
[21:00:01 CET] <cone-519> ffmpeg 03James Almer 07master:af7ab32b897b: fate/cbs: add a switch frame AV1 test
[21:00:02 CET] <cone-519> ffmpeg 03James Almer 07master:637742b45dc8: fate/cbs: add a decode model AV1 test
[23:10:19 CET] <mkver> Anybody with knowledge about tta here?
[23:24:18 CET] <durandal_1707> mkver: whats wrong?
[23:25:44 CET] <mkver> The Matroska demuxer allocates 30 bytes of extradata for it, but uses only 18 of them. The tta demuxer uses 22 bytes for the extradata (the difference is in a CRC-32, but even then eight bytes are unaccounted for). This is weird
[23:27:08 CET] <durandal_1707> mkver: note that out tta demuxer can't handle correctly decoding of last frame after seeking
[23:27:13 CET] <durandal_1707> *out
[23:27:14 CET] <durandal_1707> *our
[23:27:51 CET] <mkver> And you think that this has something to do with this?
[23:28:14 CET] <durandal_1707> that is only major issue with it i think, and same applies to matroska
[23:28:33 CET] <durandal_1707> or only to tta in matroska, long time ago it was
[23:28:44 CET] <jamrial> tta in mkv is a mess because the spec strictly requires the header to be dumped, which forces us to somehow calculate the last frame size
[23:28:48 CET] <jamrial> but yeah, that 30 is probably wrong
[23:29:17 CET] <durandal_1707> last frame size calculation is not working, last time i tested it
[23:29:38 CET] <jamrial> yup
[23:31:21 CET] <jamrial> the decoder says "30bytes includes TTA1 header"
[23:31:40 CET] <jamrial> and rejects any extradata smaller than 22 bytes
[23:32:21 CET] <mkver> And this here: http://tausoft.org/wiki/True_Audio_Codec_Format has only 22 bytes.
[23:35:23 CET] <mkver> The 30 bytes is with a seektable with one frame; see c4e0e314248865830ec073e5a3ef08e0a40aabf2
[23:44:27 CET] <jamrial> mkver: if you want, change the 30 in the mkv demuxer to 22. the tta demuxer doesn't store the seek table in extradata either
[23:44:52 CET] <mkver> Already did so.
[00:00:00 CET] --- Mon Dec  2 2019