summaryrefslogtreecommitdiffstats
path: root/libavcodec
Commit message (Collapse)AuthorAgeFilesLines
* libopenjpeg: introduce lowres and lowqual private optionsLuca Barbato2012-07-201-0/+28
| | | | | OpenJPEG can decode in lower resolution or decode only a number of enhancement layers.
* alac: fix channel pointer assignment for 24 and 32-bitJustin Ruggles2012-07-191-4/+4
| | | | Needs to be done separately for each element.
* alac: cosmetics: general pretty-printing and comment clean upJustin Ruggles2012-07-191-31/+18
|
* alac: calculate buffer size outside the loop in allocate_buffers()Justin Ruggles2012-07-191-2/+2
|
* alac: change some data types to plain intJustin Ruggles2012-07-191-5/+5
|
* alac: cosmetics: rename some variables and function namesJustin Ruggles2012-07-191-93/+77
|
* alac: multi-channel decoding supportAndrew D'Addesio2012-07-191-18/+43
| | | | Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
* alac: split element parsing into a separate functionJustin Ruggles2012-07-191-40/+81
| | | | | This will make multi-channel implementation simpler. Based partially on a patch by Andrew D'Addesio <modchipv12@gmail.com>.
* alac: support a read sample size of up to 32Justin Ruggles2012-07-191-10/+11
| | | | | Use get_bits_long() in decode_scalar(). Use unsigned int for decoded value.
* alac: output in planar sample formatJustin Ruggles2012-07-191-27/+19
| | | | | Avoids unneeded interleaving and allows for reusing the AVFrame output buffer as the internal buffer for 24-bit and 32-bit sample size.
* alac: add 32-bit decoding supportJustin Ruggles2012-07-191-1/+14
|
* alac: simplify channel interleavingAndrew D'Addesio2012-07-191-43/+14
| | | | Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
* alac: use AVPacket fields directly in alac_decode_frame()Justin Ruggles2012-07-191-6/+5
|
* alac: fix check for valid max_samples_per_frameJustin Ruggles2012-07-191-3/+3
|
* alac: use get_sbits() to read LPC coefficients instead of castingJustin Ruggles2012-07-191-1/+1
|
* alac: move the current samples per frame to the ALACContextJustin Ruggles2012-07-191-23/+22
| | | | This will simplify the multi-channel implementation.
* alac: avoid using a double-negative when checking if the frame is compressedJustin Ruggles2012-07-191-4/+3
|
* alac: factor out output_size check in predictor_decompress_fir_adapt()Justin Ruggles2012-07-191-5/+3
|
* alac: factor out loading of next decoded sample in LPC predictionJustin Ruggles2012-07-191-3/+4
|
* alac: use index into buffer_out instead of incrementing the pointerJustin Ruggles2012-07-191-8/+6
|
* alac: simplify lpc coefficient adaptationJustin Ruggles2012-07-191-31/+10
|
* alac: reduce the number of local variables needed in lpc predictionJustin Ruggles2012-07-191-12/+12
|
* alac: simplify 1st order prediction and reading of warm-up samplesJustin Ruggles2012-07-191-13/+5
|
* alac: cosmetics: reindent after last commitJustin Ruggles2012-07-191-45/+45
|
* alac: remove unneeded conditionals in predictor_decompress_fir_adapt()Justin Ruggles2012-07-191-3/+0
|
* alac: use sizeof() instead of hardcoded data sizesJustin Ruggles2012-07-191-2/+4
|
* alac: make block_size signedJustin Ruggles2012-07-191-1/+1
| | | | It does not need to be unsigned.
* alac: remove a duplicate local variableJustin Ruggles2012-07-191-1/+0
|
* alac: conditionally set sign_modifier to 1Justin Ruggles2012-07-191-4/+2
| | | | | It is already unconditionally set to 0 prior to this, so we can modify it only when needed.
* alac: eliminate 2 unneeded local variables in bastardized_rice_decompress()Justin Ruggles2012-07-191-16/+7
| | | | | x_modified is just unnecessary, and final_val can be removed by simplifying the unsigned-to-signed conversion.
* alac: adjust conditions for updating entropy decoder historyJustin Ruggles2012-07-191-3/+3
| | | | avoids some unnecessary arithmetic in certain situations
* alac: cosmetics: reindent after last commitJustin Ruggles2012-07-191-8/+8
|
* alac: limit the rice param before passing to decode_scalar()Justin Ruggles2012-07-191-9/+7
| | | | | reduces the number of parameters to decode_scalar() and slightly simplifies the code
* alac: reduce the number of parameters to bastardized_rice_decompress()Justin Ruggles2012-07-191-11/+5
| | | | Use the ALACContext fields directly instead.
* alac: cosmetics: rename some ALACContext parametersJustin Ruggles2012-07-191-64/+63
|
* alac: clean up and update comments leftover from reverse-engineeringJustin Ruggles2012-07-191-31/+18
|
* flacdec: reverse lpc coeff order, simplify filterMans Rullgard2012-07-192-15/+14
| | | | | | | Reversing the lpc coefficient order simplifies indexing in the filter. Signed-off-by: Mans Rullgard <mans@mansr.com>
* x86: dsputil: drop some unused CPU flag debug codeDiego Biurrun2012-07-191-15/+0
|
* ppc: fix build with altivec disabledMans Rullgard2012-07-181-1/+3
| | | | Signed-off-by: Mans Rullgard <mans@mansr.com>
* vp3: move idct and loop filter pointers to new vp3dsp contextMans Rullgard2012-07-1819-169/+220
| | | | | | | | This moves all VP3-specific function pointers from dsputil to a new vp3dsp context. There is no reason to ever use the VP3 IDCT where an MPEG2 IDCT is expected or vice versa. Signed-off-by: Mans Rullgard <mans@mansr.com>
* build: add CONFIG_VP3DSP, reduce repetition in OBJS listsMans Rullgard2012-07-183-10/+7
| | | | Signed-off-by: Mans Rullgard <mans@mansr.com>
* tscc2: do not add/subtract 128 bias during DCTKostya Shishkov2012-07-181-2/+2
| | | | | | | It turns out that the reference decoder subtracts 128 from DC during block decode but adds it back during reordering block with zigzag pattern. Transforming block with incorrect DC caused heavy visual artifacts for many quantisers.
* tscc2: fix typo in DCTKostya Shishkov2012-07-181-1/+1
|
* libopenjpeg: introduce encoding supportMichael Bradshaw2012-07-174-2/+440
| | | | | | | Based on FFmpeg version from commit 713a7854e06964abc8f7d015b94acbed27769d24 Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
* libopenjpeg: rename decoder source file.Michael Bradshaw2012-07-172-1/+1
| | | | | Signed-off-by: Michael Niedermayer <michaelni@gmx.at> Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
* tscc2: DCT output should not be clippedKostya Shishkov2012-07-171-1/+1
| | | | This fixes decoding some TSCC2 files with large quantisers.
* libfdk-aac: Check if cutoff value is validMohammad Alsaleh2012-07-151-0/+5
| | | | | | | | Passing a cutoff value < sample_rate/256 will cause a crash. Also, values >20000 will have no effect and 20000 will be used anyway. Signed-off-by: Mohammad Alsaleh <msal@tormail.org> Signed-off-by: Martin Storsjö <martin@martin.st>
* vc1dec: Do not use random pred_flag if motion vector data is skippedMashiat Sarker Shakkhar2012-07-151-1/+1
| | | | | | | This fixes SA10143.vc1 from test-suite. Also partially fixes MC-VC1.ts from videolan streams archive. Signed-off-by: Martin Storsjö <martin@martin.st>
* vp8: Enclose pthread function calls in ifdefsMartin Storsjö2012-07-152-0/+9
| | | | | | This fixes building with threads disabled. Signed-off-by: Martin Storsjö <martin@martin.st>
* snow: refactor code to work around a compiler bug in MSVC.Ronald S. Bultje2012-07-141-1/+9
| | | | | This fixes the compiler error "cannot convert from 'BlockNode' to 'int16_t'".
OpenPOWER on IntegriCloud