- Only test keys when file is encrypted
- Add clHCA_TestBlock to test a frame with the current key
- Move some key detection code to hca_decoder.c and simplify hca.c
detection
- Decrease number of test frames due to increased accuracy of
clHCA_TestBlock
- rename clHCA_DecodeSamples16 to clHCA_ReadSamples16 since it isn't
really decoding
- divide clHCA_Decode into clHCA_DecodeHeader and clHCA_DecodeBlock to
clarify usage
- unify clHCA_IsOurFile0/1 since one would need to call
clHCA_DecodeHeader right after it anyway
- add clHCA_SetKey for easier key handling
- adapt vgmstream wrapper
Decoder has been modified a bit so it's easier to feed an arbitrary
number of discard samples (or at least, easier for me to understand),
since encoder delay can be >1024 (frame size) and would make some calcs
go all wobbly
Formats using custom layouts tend to be weird enough that it's a pain to
add support directly in the decoder.
Instead should use custom layouts and I/O streamfiles that, though a bit
wordy and unwieldy at the moment, are a lot more flexible.
It was decoding one 28-sample subframes and blocked layout handled
frames, doing some voodoo to skip CD sectors.
Now decodes one 28*8-sample frame and blocked layout handles CD sectors.
Decoding is byte-exact vs before.
- Code now more verbose, but easier to understand (by me, at least)
- Unify normal and badflags code
- Fix hist1 not being properly clamped, though not noticeable
- Fix inFamous (PS3) which seemingly wasn't using the extended table but
hitting some compiler weirdness? (failed in foobar and worked in test)
- Use int math for PSX-cfg, which should be minimally more accurate
On decoding start step_size was 0, but Yamaha ADPCM is supposed to start
at 127 so there was some deviation, usually minor as most files start in
silence.
Now writes header sample and skips last nibble, as seen in MS's XDK
code, and simplifies multichannel layouts and fixes some theoretical
configurations
Currently same as ms_ima_bytes_to_samples, but this will change; renamed
for consistency with all other IMA variations. Also clean a bit some
metas since I was testing anyway.
Switch Opus uses the same output format that opus_demo writes. Each frame has an 8-byte header.
The first 4 bytes contain the length of the frame. The second 4 bytes contain the final state of the range coder.
8299edfc0c/src/opus_demo.c (L658-L681)
Makefile.audacious.am:
* Add test/ subdirectory to the Autotools build
* Add version.h to EXTRA_DIST so that "make distcheck" passes
audacious/Makefile.audacious.am:
* Install Audacious plug-in into the appropriate directory
* No need to include CXXFLAGS in AM_CXXFLAGS, as the former is already
included in the compiler invocation
configure.ac:
* Need to link to libvorbis in order for static-library builds
(--disable-shared) to link correctly
* Only add GCC-specific compiler flags if GCC is detected
* Tightened up the GCC warning flags
* Don't set LIBS, so that different targets can link against different
libraries
* Generate makefile for test/ subdirectory
src/Makefile.audacious.am:
* Install libvgmstream into the library directory
* Removed GCC-specific flag from AM_CFLAGS
* No need to include CFLAGS in AM_CFLAGS, as the former is already
included in the compiler invocation
* Add libvgmstream dependencies via LIBADD instead of global LIBS var
test/Makefile.audacious.am:
* New makefile template for test/ subdirectory
unbootstrap:
* Also remove "compile" script from bootstrap
This is needed for blocked layout, as it can't do normal interleave.
Probably could be fixed in the future to remove several superfluous
_int/block decoders
Previously, the streams' offsets needed to be pre-adjusted with the
interleave; now assumes all streams start in the same offset (first
stream).
This simplifies short last interleaves (SCD/P3D) and makes
layout_mpeg_custom unnecessary (also allows theoretical variable-sized
interleaves).
