commit 5b64cf7c4d67af9050bc959996ef976ce6326888
Author: Jean-Marc Valin <jean-marc.valin@octasic.com>
Date:   Thu May 19 16:40:35 2011 -0400

    Fixes a few bugs introduced in the latest update

commit b91eb86025d2de3c4cf04f4fb1aa48a28ad676a3
Author: Timothy B. Terriberry <tterribe@xiph.org>
Date:   Thu May 19 16:12:02 2011 -0400

    build_draft.sh fixes

commit 0b10c30db7ad60ff7d4bfc1139bdb86e0543c035
Author: Jean-Marc Valin <jean-marc.valin@octasic.com>
Date:   Thu May 19 15:30:21 2011 -0400

    Fixes the draft build

commit 164424cded5978e657105104e171ac202dde5e5b
Author: Jean-Marc Valin <jean-marc.valin@octasic.com>
Date:   Thu May 19 15:00:23 2011 -0400

    Updated build system

commit 188d9fcba8c3a1b1ad67c67c430e1b37e3658a40
Author: Koen Vos <koen.vos@skype.net>
Date:   Thu May 19 17:14:57 2011 -0400

    SILK/Opus update

Includes better error handling in fft/mdct init

defining HALF16() too

Pitch search can now go up to 720 samples

The encoder would crash in the PVQ search if fed NaNs via the float interface. This patch protects against it in two sufficient ways: Making the PVQ search robust against NaNs and by squashing NaNs to zero on input.

Thanks to David Richards for reporting this failure mode.

Timothy B. Terriberry authored 13 years ago and Jean-Marc Valin committed 13 years ago

This ensures we don't have to rely on build-system support for
 disabling these warnings in libcelt.

commit dfe4d46f9abf6b0e96e66370d428da4e283204ce
Author: Jean-Marc Valin <jean-marc.valin@octasic.com>
Date:   Mon May 2 09:59:45 2011 -0400

    SILK file list update

commit 5faf541d86445eb8ce1775ea206e88afe83c7107
Author: Jean-Marc Valin <jean-marc.valin@octasic.com>
Date:   Mon May 2 09:51:29 2011 -0400

    Makefiles update

commit f5d70b3ccbe209cc07a1a604fffca93ea34bc64b
Author: Koen Vos <koen.vos@skype.net>
Date:   Mon May 2 09:50:33 2011 -0400

    SILK stereo update and MSVC build update

Gregory Maxwell authored 13 years ago and Jean-Marc Valin committed 13 years ago

Stop adjusting the VBR drift during silent frames, this prevents overshoot after silent spans. This also results in undershoot on files with lots of silence, but this is desirable in terms of increasing quality consistency. It will be even better if we later include some open-loop correction of the targets.

We don't allow *encoding* a stereo stream with a mono encoder
because that's just silly.

Adds a CELT_GET_LOOKAHEAD() ctl() call instead. Other uses of
celt_mode_info() should not be needed anymore.

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.

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

We were trying to normalize bands that didn't actually exist (e.g.,
 the last band with 320-sample frames at 32kHz).
Thanks to John Ridges for the report.

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

Mark _BitScanReverse as an intrinsic, as this is not done by default
 in _DEBUG mode (thanks to Bjoern D. Rasmussen for the suggestion).
Move the declaration of ec_ilog() to where the macro using it is
 defined, to ensure it is available when the macro is used.
Thanks to John Ridges for the report.

The return code has been eliminated in favor of using the existing
 error flag in the ec_enc struct.
This also adds the declaration to the header, which JM's importation
 of Koen's adaptation of my original code in b3c05b6b omitted.

Correct a number of uninitialized value problems when end!=nbEBands. Now 8khz/20ms audio can be decoded at 48kHz and vice versa (with the PF turned off).