Skip to Content

AV1 Ecosystem Update: December 2019

Table of Contents

dav1d 0.5.2 ‘Asiatic Cheetah’ released

dav1d ended the year with another performance improvement release. Dubbed the “Asiatic Cheetah” this release brings the following improvements:

  • ARM32 optimizations for loopfilter, ipred_dc|h|v
  • Add section-5 raw OBU demuxer
  • Improve the speed by reducing the L2 cache collisions
  • Fix minor issues, including compilation on some OSes

Release notes for dav1d 0.5.2 ‘Asiatic Cheetah’

rav1e v0.2.0: Winter Solstice released

Building on the official 0.1.0 release in November, rav1e had a busy December with 3 releases. Overall, rav1e got a lot faster which is only good news for AV1 encoding! Here are the improvements included in the 0.2.0 release.

  • Faster, overall 40%-70% faster than v0.1.0 depending on the encoding settings.
  • Optional serialization/deserialization of the encoding parameters through the feature serialize
    • Optional cli advanced commands to use it.
  • The builds are now using the dwarf debug format for the targets that support it, before it was a mixture of dwarf and stabs due to the nasm defaults.
  • Added a –benchmark hidden flag for the cli for MacOS and Linux.
  • documentation is now available on docs.rs.

Source: rav1e v0.2.0: Winter Solstice release notes

And, these are the improvements from the weekly pre-releases.

  • Faster, around 20% more compared to the last pre-release
    • More assembly optimizations from dav1d, mainly impacting AArch64
    • Forward transform SIMD, to be accounted for the largest part of this week speedup
    • Simplifications and refactoring in the Motion Estimation and Scene Change
  • Optional serialization/deserialization of the encoding parameters through the feature serialize
    • Optional cli advanced commands to use it.
  • The builds are now using the dwarf debug format for the targets that support it, before it was a mixture of dwarf and stabs due to the nasm defaults.
  • Slightly faster.
  • 60%-80% reduction in the resident set depending on the number of tiles in use. Now the resident set is nearly independent from the number of tiles in use.

Source: - 2019-12-15 Weekly Pre-release - 2019-12-01 Weekly Pre-release

Luca Barbato, one of the rav1e and dav1d contributors, recently presented detailed technical information about the improvements included in 0.2.0 release as well as the rav1e roadmap. It’s an interesting presentation with a lot more information on what features and the use cases the team is working on. You can view Luca’s presentation here https://people.xiph.org/~negge/LVS2019.pdf.

SVT-AV1 v0.8.0 released

Not to be left behind, SVT-AV1 project also ended the year on high with another official release. Here are the improvements included.

Encoder

  • Preset Optimizations
  • Single-core execution memory optimization [-lp 1 -lad 0]
  • Rate estimation update enhancements
  • On / off flags for feature run-time switching
  • Auto-max partitioning algorithm support
  • Multi-pass partitioning depth support
  • Remove deprecated RC mode 1 and shifter RC mode 2 and mode 3 to mode 1 and mode 2 respectively
  • Update cost calculation for CDEF Filtering
  • Intra-Inter compound for 10-bit
  • Eigth-pel optimization
  • AVX512 Optimizations
  • AVX2 Optimizations

Decoder

  • Initial multi-threading support
  • Decoder optimizations / cleanup

Source: SVT-AV1 codec 0.8.0 release notes

SDMC DV8919 Amlogic S905X4 Android TV 10 TV Box Supports AV1 Decoding

Over the past few months, we’ve seen reports of SOCs in development that support AV1 hardware decoding but no actual products. SDMC DV8919 appears to be the first Android TV box that leverages Amlogic’s new S905X4 chip which supports AV1 4K120p hardware decoding.

Source: https://www.cnx-software.com/2019/12/22/sdmc-dv8919-amlogic-s905x4-android-tv-10-tv-box-supports-av1-decoding/

Product Page for SDMC DV8919

Previous ecosystem updates

AV1 resources

Contact

Reach out if you have any questions! Feel free to follow me on

comments powered by Disqus