Viswanath Puttagunta authored 9 years ago and Jean-Marc Valin committed 9 years ago

Optimize opus encode (float only) usecase using ARM NE10
library. Mainly effects opus_fft and ctl_mdct_forward
and related functions.

This optimization can be used for ARM CPUs that have NEON
VFP unit. This patch only enables optimizations for ARMv7.

Official ARM NE10 library page available at
http://projectne10.github.io/Ne10/



To enable this optimization, use
--enable-intrinsics --with-NE10=<install_prefix>
or
--enable-intrinsics --with-NE10-libraries=<NE10_lib_dir> --with-NE10-includes=<NE10_includes_dir>

Compile time checks made during configure process to make sure
optimization option available only when compiler supports NEON
instrinsics.

Runtime checks made to make sure optimized functions only called
on appropriate hardware.

Signed-off-by: Timothy B. Terriberry <tterribe@xiph.org>

This fixes an actual error in the downmix (using the float version even
for the int API), as well as a bunch of conversion warnings.

Should fix delayed decision on surround, but not well tested.

Code is still float, but at least tonality esitmation seems to work.
Speech/music analysis is still disabled.

I checked that it's not *completely* busted, but it probably needs more testing.

This should prevent errors in the bandwidth detection from affecting quality
when we have enough bits to be close to transparent.

This is done using an adaptive beta and an estimate of the speech
and music detection confidence

Variable duration option renamed to OPUS_SET_EXPERT_FRAME_DURATION,
with new API.
Also moves up the analysis to avoid having to do int->float conversion
on large buffers.

Can now work up to full-band and uses lsb_depth to fix the noise issue.

Also, reducing the VBR rate on panned mono

This reverts commit 9f407afa.

Also added 3rd clause to "master" COPYING file

In f9265ddf opus_get_version_string() was moved
to libcelt, but the fallback definition for OPUS_VERSION
stayed in src/opus.c. This is helpful for those using
and external build system, which may not duplicate
all the defines from the included build files.

Thanks to Tim Terriberry for pointing out the issue.

This better distinguishes the reference implementation,
which is what this string is meant to do. Thanks to
Greg Maxwell for the suggestion.

This lets users print it out without context, which is a little
easier for clients.

The new opus_get_version_string() call just returns a static string
defined by OPUS_VERSION, which is passed in from the build system
through config.h (or a custom compile line). Provide a fallback to
"unknown" if the build system fails to actually provide that definition.

This restores compilation with Makefile.draft.

In general, this means there will be builds out there with non-specific
version strings, since ports won't be forced to update the string. While
that's unfortunate, I think it's more valuable that the library be simple
to build.

Timothy B. Terriberry authored 14 years ago and Jean-Marc Valin committed 14 years ago

These were used because the entropy coder originally came from
 outside libcelt, and thus did not have a common type system.
It's now undergone enough modification that it's not ever likely to
 be used as-is in another codec without some porting effort, so
 there's no real reason to maintain the typedefs separately.
Hopefully we'll replace these all again somedate with a common set
 of Opus typedefs, but for now this will do.

This fixes an issue caused by commit 6c8acbf1, which moved the
 ec_ilog() prototype from entcode.h to ecintrin.h, where the
 ec_uint32 typedef was not yet available.
Thanks to John Ridges for the report.

Got authorization from all copyright holders

by POSIX. The other _t types that are not part of the API are still there
for now. Also, got rid of all that was left of the 64-bit types.

Timothy B. Terriberry authored 15 years ago and Jean-Marc Valin committed 15 years ago

Adds specialized O(N*log(K)) versions of cwrsi() and O(N) versions of icwrs()
 for N={3,4,5}, which allows them to operate all the way up to the theoretical
 pulse limit without serious performance degredation.
Also substantially reduces the computation time and stack usage of
 get_required_bits().
On x86-64, this gives a 2% speed-up for 256 sample frames, and almost a 16%
 speed-up for 64 sample frames.

fine energy bits. Also, doing better rounding of the bits.

Timothy B. Terriberry authored 16 years ago and Jean-Marc Valin committed 16 years ago

It is no slower with a large number of pulses, and as much as 30% faster with
 a large number of dimensions.

Timothy B. Terriberry authored 16 years ago and Jean-Marc Valin committed 16 years ago

This lets us encode and decode directly from the pulse vector without an
 intermediate transformation.
This makes old streams undecodable.
Additionally, ncwrs_u32() has been sped up for large N by using the sliding
 recurrence from Mohorko et al.
ncwrs_u64 could be sped up in a similar manner, but would require a larger
 table of multiplicative inverses (or several 32x32->64 bit multiplies).
Note that U(N,M) is now everywhere 1/2 the value it used to be.