The encoder is fixed point, and uses an MDCT only for analysis. Due
to the slightly different rounding, the encoder makes a different
decision, so the tests have to be adjusted as well.
This patch replaces the transform used in AAC with lavu/tx and removes
the limitation on only being able to decode 960-sample files
with the float decoder.
This commit also removes a whole bunch of unnecessary and slow
lifting steps the decoder did to compensate for the poor accuracy
of the old integer transformation code.
Overall float decoder speedup on Zen 3 for 64kbps: 32%
The encoder seems to be trading blows with hevc_nvenc.
In terms of quality at low bitrate cbr settings, it seems to
outperform it even. It produces fewer artifacts and the ones it
does produce are less jarring to my perception.
At higher bitrates I had a hard time finding differences between
the two encoders in terms of subjective visual quality.
Using the 'slow' preset, av1_nvenc outperformed hevc_nvenc in terms
of encoding speed by 75% to 100% while performing above tests.
Needless to say, it always massively outperformed h264_nvenc in terms
of quality for a given bitrate, while also being slightly faster.
Support for building with older versions of MSVC (with the
c99wrap/c99conv frontend) was removed in
ce943dd6ac.
Signed-off-by: Martin Storsjö <martin@martin.st>
This effectively reverts 9b78abae19.
The underlying issue has been fixed properly in commit
cff480e49d.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The eatqi decoder uses a custom IDCT and actually does not
use the IDCTDSP API at all. Somehow it was nevertheless
used to simply apply the identity permutation on ff_zigzag_direct.
This commit stops doing so.
Reviewed-by: Peter Ross <pross@xvid.org>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The eatgq decoder uses a custom IDCT and actually does not
use the IDCTDSP API at all. Somehow it was nevertheless
used to simply apply the identity permutation on ff_zigzag_direct.
This commit stops doing so. It also renames perm to scantable,
because it is only the scantable as given by the spec without
any further permutation performed by us.
Reviewed-by: Peter Ross <pross@xvid.org>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The eamad decoder uses a custom IDCT and actually does not
use the IDCTDSP API at all. Somehow it was nevertheless
used to simply apply the identity permutation on ff_zigzag_direct.
This commit stops doing so.
Reviewed-by: Peter Ross <pross@xvid.org>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
librav1e provides a function to create extradata, so use it instead of
extracting the sequence header OBU from packets.
Signed-off-by: James Almer <jamrial@gmail.com>
The binkaudio decoders don't need mdct or sinewin at all;
and binkaudio_dct doesn't need rdft directly (but nevertheless
uses it indirectly via dct).
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
As long as ff_mpeg12_common_init() existed in mpeg12.c,
it added a dependency of mpeg12.o on mpegvideodata.o
(which provides ff_mpeg2_dc_scale_table, which is used
in ff_mpeg12_common_init()). mpegvideodata.o is normally
provided by the mpegvideo subsystem and therefore several
codecs and the MPEG-1/2 parser added a configure dependency
on said subsystem (additionally, the eatqi decoder just
added a Makefile dependency on mpegvideodata.o).
Given that ff_mpeg12_common_init() is no more, these dependencies
can be removed.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Starting with an h264 implementation. Can be extended to support other codecs.
A few caveats:
- OpenGOP streams are currently not supported. The firt packet must be an IDR
frame.
- In some streams, a few frames at the end may not get a reordered PTS when
they reference frames past EOS. The code added to derive timestamps from
previous frames needs to extended.
Addresses ticket #502.
Signed-off-by: James Almer <jamrial@gmail.com>
RISC-V defines the CLZ instruction as part of the ratified Zbb subset
of the (not yet ratified) bit mapulation extension (B). We can detect
it from the __riscv_zbb predefined constant. At least GCC 12 already
supports this correctly.
Note that the macro will be non-zero if supported, zero if enabled
in the compiler flags (e.g. -march=rv64gzbb) but not known to the
compiler, and undefined otherwise.
_Float16 support was available on arm/aarch64 for a while, and with gcc
12 was enabled on x86 as long as SSE2 is supported.
If the target arch supports f16c, gcc emits fairly efficient assembly,
taking advantage of it. This is the case on x86-64-v3 or higher.
Same goes on arm, which has native float16 support.
On x86, without f16c, it emulates it in software using sse2 instructions.
This has shown to perform rather poorly:
_Float16 full SSE2 emulation:
frame=50074 fps=848 q=-0.0 size=N/A time=00:33:22.96 bitrate=N/A speed=33.9x
_Float16 f16c accelerated (Zen2, --cpu=znver2):
frame=50636 fps=1965 q=-0.0 Lsize=N/A time=00:33:45.40 bitrate=N/A speed=78.6x
classic half2float full software implementation:
frame=49926 fps=1605 q=-0.0 Lsize=N/A time=00:33:17.00 bitrate=N/A speed=64.2x
Hence an additional check was introduced, that only enables use of
_Float16 on x86 if f16c is being utilized.
On aarch64, a similar uplift in performance is seen:
RPi4 half2float full software implementation:
frame= 6088 fps=126 q=-0.0 Lsize=N/A time=00:04:03.48 bitrate=N/A speed=5.06x
RPi4 _Float16:
frame= 6103 fps=158 q=-0.0 Lsize=N/A time=00:04:04.08 bitrate=N/A speed=6.32x
Since arm/aarch64 always natively support 16 bit floats, it can always
be considered fast there.
I'm not aware of any additional platforms that currently support
_Float16. And if there are, they should be considered non-fast until
proven fast.
Lynne
Timo Rothenpieler
Martin Storsjö
Andreas Rheinhardt
Wenbin Chen
Anton Khirnov
James Almer
Rémi Denis-Courmont
Roman Arzumanyan
Mohamed Khaled Mohamed
Hendrik Leppkes