The problem was that forcing CELT-mode for low bitrate CBR was done too late,
after the encoder had decided to use SILK. This was causing redundancy
to be allocated because the encoder didn't realize it was going to keep
using CELT.

Padding is now handled by the repacketizer.

Mark Harris authored 11 years ago and Jean-Marc Valin committed 11 years ago

Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>

http://lists.xiph.org/pipermail/opus/2013-November/002372.html

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.

denormalise_bands() can now produce signals close to the max MDCT amplitude.

Stereo mode, stereo width, min bandwidth, VBR damping

Code behaviour is unchanged

Defining FIXED_POINT and DISABLE_FLOAT_API now leaves no float instruction

Because it's indirectly used in the encoder (through the repackerizer).

Caps the savings at 1 bit per (coded) sample. This doesn't really increase
the bitrate and fixes some issues, e.g. with IgorC's "twilight" sample.

The library really depends on all the little fixed point math functions
 being inlined in order to get acceptable performance. It turns out that
 it's very easy for someone to compile with optimization disable when
 twiddling cflags or cooking up their own build system.

Newer versions of MSVC are unhappy with the strategy of the build
 environment redefining "inline" (even though they don't support the
 actual keyword). Instead we define OPUS_INLINE to the right thing
 in opus_defines.h.

This is the same approach we use for restrict.

Stefan Beller authored 11 years ago and Jean-Marc Valin committed 11 years ago

The compiler warned:
silk/float/pitch_analysis_core_FLP.c: In function 'silk_P_Ana_calc_corr_st3':
silk/float/pitch_analysis_core_FLP.c:499:36: warning: variable 'basis_ptr' set but not used [-Wunused-but-set-variable]
     const silk_float *target_ptr, *basis_ptr;

This means we can safely remove it.

Signed-off-by: Stefan Beller <stefanbeller@googlemail.com>
Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>

This fixes tansig_approx() to avoid crashing when the input is NaN.
The problem could only be triggered when calling the float API with
a float build at a complexity of 7 or more (i.e. analysis called).
Since the crash was due to an out-of-bound read (typically the index
is INT_MIN), it's unlikely to be exploitable in any other way than
causing a crash.

This define was replaces by OPUS_RESTRICT in opus_defines.h
in de0b5324 and subsequent commits. It is no longer necessary
to include it.

The LFE encoder would first decide to switch to SILK, and allocated
redundancy bytes. Then, it would force CELT because of LFE, but keep
the redundancy allocation, and end up with too few bytes, that would
end up being padded to the desired size for CBR.

Even when using SILK/hybrid.

This should fix decoding of padded multistream packets and (hopefully)
multistream fec.