aboutsummaryrefslogtreecommitdiffstats
path: root/src/decoder (follow)
Commit message (Collapse)AuthorAgeFilesLines
* decoder/ffmpeg: detect and fix negative time stampsMax Kellermann2014-12-151-4/+7
| | | | | Works around assertion failure due to something that appears to be a (minor) FFmpeg bug.
* decoder/ffmpeg: support FFmpeg 2.5Max Kellermann2014-12-081-0/+5
| | | | | | Version 2.5 fixed an API oddity, however it broke API compatibility, at least with C++. Disable the workaround when a libavformat version is detected that is recent enough.
* decoder/ffmpeg: support opusMax Kellermann2014-11-101-1/+2
|
* decoder/faad: remove workaround for ancient libfaad2 ABI bugMax Kellermann2014-11-021-10/+2
| | | | | | | | | | | | | | | | | | | | | | | | Many years ago, FAAD had a serious ABI bug: the NeAACDecInit() prototype in its header declared the "samplerate" parameter to be "unsigned long *", but internally, the function assumed it was "uint32_t *" instead. On 32 bit machines, that was no difference, but on 64 bit, this left one portion of the return value uninitialized; and worse, on big-endian, the wrong word was filled. This bug had to be worked around in MPD (commit 9c4e97a6). A few months later, the bug was fixed in the FAAD CVS in commit 1.117 on file libfaad/decoder.c; the commit message was: "Use public headers internally to prevent duplicate declarations" The commit message was too brief at best; the problem was not duplicate declarations, but a prototype mismatch. No mention of the bug fix in the ChangeLog. The MPD project never learned about this bug fix, and so MPD would always pass a "uin32_t *" dressed up as a "unsigned long *". Nearly 6 years later, it's about time to fix this second ABI problem. Let's kill the workaround!
* decoder/ffmpeg: recognize MIME type audio/aacpSteven OBrien2014-11-021-0/+1
|
* decoder/ffmpeg: use memset() to initialize AVProbeDataMax Kellermann2014-09-091-0/+6
|
* decoder/ffmpeg: pass MIME type to ffmpeg/libav version 11Max Kellermann2014-09-071-0/+7
| | | | | | | That attribute was uninitialized before, which could crash libavformat. See Debian bug 760669
* decoder/audiofile: fix crash after seekingMax Kellermann2014-09-061-1/+1
| | | | | | Log call was added to the wrong branch. Fixes regression by commit ca1a1149
* decoder/gme: fix song durationMax Kellermann2014-08-291-1/+1
| | | | The unit of gme_info_t::length is milliseconds, not centiseconds.
* decoer/dsdiff: fix endless loop on malformed fileMax Kellermann2014-08-211-1/+1
| | | | Same bug as in the previous commit.
* decoer/dsf: fix endless loop on malformed fileMax Kellermann2014-08-211-1/+1
| | | | | | | When the data chunk size is not a multiple of the frame size, the last partial frame lead to an endless loop. We fix this by checking chunk_sze>=frame instead of chunk_sze>0. This way, the partial frame is simply skipped.
* decoder/ffmpeg: use avcodec_descriptor_get() to determine codec nameMax Kellermann2014-08-131-0/+9
| | | | | | In version 11, both ffmpeg and libav deprecate AVCodecContext::codec_name. The function avcodec_descriptor_get() has been introduced long ago.
* decoder/dsdiff: simplify dsdlib_skip() callMax Kellermann2014-07-121-4/+3
|
* decoder/dsdiff: simplify loop condition, merge branchesMax Kellermann2014-07-121-7/+2
|
* decoder/dsdiff: ignore garbage null byte at end of fileMax Kellermann2014-07-121-1/+1
| | | | | Failure to read another chunk header is not fatal. Continue to read metadata.
* decoder/dsdiff: fix metadata parser bug (uninitialized variables)Max Kellermann2014-07-121-1/+4
|
* decoder/faad: estimate song duration for remote filesMax Kellermann2014-07-121-0/+22
| | | | | | | | Previously, MPD tried to slurp the whole song file, count the number of frames and calculate the song duration from that. That however is extremely expensive for remote files, and will delay playback for a long time. Workaround: check only the first 128 frames and try to extrapolate from here. Fixes Mantis ticket 0004035.
* decoder/faad: bail out early if sample rate is invalidMax Kellermann2014-07-121-0/+2
|
* decoder/faad: use adts_check_frame() in faad_song_duration()Max Kellermann2014-07-121-2/+1
| | | | Eliminate more duplicate code.
* decoder/faad: test "seekable" after ADTS frame checkMax Kellermann2014-07-121-1/+5
| | | | | Don't bother to check for ADIF just because the stream is not seekable.
* decoder/faad: move code to faad_decoder_new()Max Kellermann2014-07-121-14/+17
| | | | Merge some duplicate code.
* decoder/faad: remove unnecessary readMax Kellermann2014-07-121-2/+0
| | | | Eliminate some overhead when the caller doesn't need the buffer.
* DecoderBuffer: add method _clear()Max Kellermann2014-07-121-8/+3
|
* decoder/faad: check sample_rate, not frames_per_secondMax Kellermann2014-07-111-2/+4
| | | | Checking the integer is faster, easier and more reliable.
* decoder/faad: make variables more localMax Kellermann2014-07-111-71/+49
|
* decoder/faad: use MAX_CHANNELSMax Kellermann2014-07-111-4/+2
| | | | .. instead of declaring a new constant.
* decoder/sndfile: use decoder_read_full()Max Kellermann2014-07-111-16/+6
| | | | | Replaces the loop in sndfile_vio_read(), eliminating duplicate and fragile code.
* decoder/sndfile: use decoder_read()Max Kellermann2014-07-111-16/+22
| | | | .. instead of InputStream::LockRead(). The former is cancellable.
* decoder/sndfile: log seek errorsMax Kellermann2014-07-111-1/+4
|
* decoder/audiofile: log seek errorsMax Kellermann2014-07-111-0/+1
|
* decoder/audiofile: fix typo in commentMax Kellermann2014-07-111-1/+1
|
* decoder/dsd: use decoder_read_full() where appropriateJoff2014-07-094-31/+19
| | | | | | | Addresses Mantis ticket 0004015. [mk: use decoder_read_full() only when needed, and a few formal changes]
* decoder/audiofile: use decoder_read_full()Max Kellermann2014-07-091-1/+6
| | | | | Works around WAV stream playback bug, because libaudiofile does not like partial reads (Mantis 0004028).
* DecoderAPI: add function decoder_read_full()Max Kellermann2014-07-091-12/+2
| | | | Move code from the "mad" plugin.
* DecoderAPI: add function decoder_skip()Max Kellermann2014-07-091-14/+1
| | | | Move code from the "mad" plugin.
* decoder/audiofile: use decoder_read()Max Kellermann2014-07-091-15/+24
| | | | | .. instead of InputStream::LockRead(). The former is cancellable.
* decoder/OpusDecoderPlugin: fix build failure due to missing stdio.h includeGustavo Zacarias2014-07-091-0/+1
| | | | Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
* decoder/AudiofileDecoderPlugin: fix build failure due to missing stdio.h includeGustavo Zacarias2014-07-091-0/+1
| | | | Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
* decoder/OggUtil: allow skipping up to 32 kB after seekMax Kellermann2014-04-291-1/+1
| | | | | | | Fixes missing song length on high-latency Opus files. According to tests with 320 kbit/s opus files with 60ms packets, we need to skip up to 29 kB.
* decoder/sndfile: work around libsndfile bug on partial readMarcello Desantis2014-04-091-6/+20
|
* decoder/ffmpeg: handle unknown stream start timeMax Kellermann2014-03-181-2/+25
|
* decoder/ffmpeg: pass AVSEEK_FLAG_ANY to av_seek_frame()Max Kellermann2014-03-181-1/+1
| | | | | | This corrects a major mistake from commit 724a59aa - there was one small thing that commit was supposed to do, and it failed. AV_TIME_BASE is not a seek flag.
* decoder/gme: fix memory leak in container_scan()Max Kellermann2014-03-061-0/+1
|
* decoder/ffmpeg: support libav v10_alpha1Max Kellermann2014-01-151-1/+11
|
* decoder/ffmpeg: include cleanupMax Kellermann2014-01-152-4/+1
|
* decoder/ffmpeg: check for av_samples_get_buffer_size() errorsMax Kellermann2014-01-151-0/+3
| | | | Fixes potential nullptr dereference.
* decoder/faad: fix memory leakMax Kellermann2014-01-081-0/+2
|
* fix FfmpegDecoderPlugin to use relative timestampsSteven O'Brien2013-12-201-4/+6
|
* decoder/ffmpeg: use IgnoreError instead of local Error instanceMax Kellermann2013-11-281-5/+2
|
* decoder/{dsf,dsdiff}: eliminate useless assignmentsMax Kellermann2013-11-232-5/+4
|