Hello,
This adds the Opus post-filter. Opus deemphasis is not included as it (IMHO)
involves too many unknowns to write a good implementation without
benchmarking.
This post-filter implementation merely vectors the scalar product. It does
purposed not vector the memory accesses to avoid reconfiguring the vector unit
too often. It might be possible to work it out, but again, I cannot really
tell without real hardware
The following changes since commit b0c7352cd494c88d33f032be60b5e1c4e8b092a0:
avcodec/mjpegdec: Fix compilation (2022-09-30 19:30:29 +0200)
are available in the Git repository at:
git.remlab.net:git/ffmpeg.git rvv-vtype
for you to fetch changes up to 1aa3efa0563aaaed376a35c0e7c9fe53089c3a7e:
lavc/opusdsp: RISC-V V (256-bit vectors) postfilter (2022-10-01 15:24:42
+0300)
----------------------------------------------------------------
Rémi Denis-Courmont (4):
QEMU hack
lavc/opusdsp: RISC-V V postfilter
lavu/riscv: helper macro for VTYPE encoding
lavc/opusdsp: RISC-V V (256-bit vectors) postfilter
libavcodec/opusdsp.c | 2 ++
libavcodec/opusdsp.h | 1 +
libavcodec/riscv/Makefile | 2 ++
libavcodec/riscv/opusdsp_init.c | 49 +++++++++++++++++++++++++++
libavcodec/riscv/opusdsp_rvv.S | 59 ++++++++++++++++++++++++++++++++
libavutil/riscv/asm.S | 75 +++++++++++++++++++++++++++++++++++++++++
libavutil/riscv/cpu.c | 2 +-
7 files changed, 189 insertions(+), 1 deletion(-)
create mode 100644 libavcodec/riscv/opusdsp_init.c
create mode 100644 libavcodec/riscv/opusdsp_rvv.S
--
レミ・デニ-クールモン
http://www.remlab.net/
_______________________________________________
ffmpeg-devel mailing list
[email protected]
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
[email protected] with subject "unsubscribe".
