Home

Welcome to the ffvvc wiki!

We are targeting to submit vvc native decoder to FFmpeg upstream.

How to contribute?

1. Check https://github.com/orgs/ffvvc/projects/1 Todo column to see anything you are interested.
2. If no, check https://github.com/ffvvc/FFmpeg for issues or fill some issue found by yourself.
3. make your changes
4. Make sure all conformances tests passed in the local
5. Create a new PR

How to read the code?

You can start with ff_vvc_task_run, see introductions.

Performance data

Command: perf.py -f ./ffmpeg performance/

d912ba17238eb7efb1892ad2b97e4443afe40cab (2023/12/17) on i7-12700K:

Clip	FPS (16 Threads)
BQTerrace_1920x1080_60_10_420_22_RA.vvc	141.3
Chimera_8bit_1080P_1000_frames.vvc	290.3
NovosobornayaSquare_1920x1080.bin	267.3
RitualDance_1920x1080_60_10_420_32_LD.266	268.3
RitualDance_1920x1080_60_10_420_37_RA.266	263.3
Tango2_3840x2160_60_10_420_27_LD.266	60.3

577acdea on M1 Mac Pro:

Clip	FPS
BQTerrace_1920x1080_60_10_420_22_RA.vvc	58.7
Chimera_8bit_1080P_1000_frames.vvc	153.3
NovosobornayaSquare_1920x1080.bin	150.3
RitualDance_1920x1080_60_10_420_32_LD.266	105.0
RitualDance_1920x1080_60_10_420_37_RA.266	133.0
Tango2_3840x2160_60_10_420_27_LD.266	21.7

580a67f on i7-11700k:

Clip	FPS
RitualDance_1920x1080_60_10_420_32_LD.266	138
RitualDance_1920x1080_60_10_420_37_RA.266	150
Tango2_3840x2160_60_10_420_27_LD.266	33

Contributors(based on patch merge time):

Nuo Mi <[email protected]>

Xu Mu <[email protected]>

frankplow <[email protected]>

Shaun Loo <[email protected]>