Feature request: Add hardware acceleration for FFmpeg #190
Labels
enhancement
New feature or request
Comments
|
What is missing to make it possible? |
|
Isn't passing the |
|
I am not an expert but I assume some hardware acceleration related arguments need to be added to the ffmpeg command: https://trac.ffmpeg.org/wiki/Hardware/QuickSync (In addition to passing the device to the container) |
C4illin
added a commit
that referenced
this issue
Dec 11, 2024
|
I added FFMPEG_ARGS as en env variable, set it to "-hwaccel qsv" and pass the gpu to the container. Hopefully it works, I dont have an intel cpu but I tested with vulkan. |
When trying to convert something I'm getting this error: Here is my docker compose: convertx:
user: "701:701"
image: "ghcr.io/c4illin/convertx:main"
container_name: "convertx"
restart: "unless-stopped"
devices:
- "/dev/dri:/dev/dri"
environment:
- "FFMPEG_ARGS=-hwaccel qsv"
- "TZ=*****"
- "JWT_SECRET=*******"
volumes:
- "./convertx:/app/data"
networks:
- externalOutput of crw-rw---- 1 root 701 226, 1 dec 9 11:40 card1
crw-rw-rw- 1 root 701 226, 128 dec 9 11:40 renderD128And inside the container crw-rw---- 1 root 701 226, 1 Dec 9 11:40 card1
crw-rw-rw- 1 root 701 226, 128 Dec 9 11:40 renderD128Hardware acceleration is working fine in other containers with the same setup. |
Does ffmpeg work with hardware acceleration outside the container? Maybe some more args is needed |
It should be enough I think. This works outside the container |
|
Ok, if the command works by itself but not in the container with the same input there is only two possibilities: you added the wrong device or the alpine version of ffmpeg doesn't inclue qsv |
|
Maybe try again and add |
|
You could also test |
|
Hello Everyone. To make it work on my synology with intel celeron J4125 that uses i965 intel video driver (vaapi): I also downloaded the i965 driver file : i965_drv_video.so and mapped it inside the container. Compose file: This worked with my synology and converts with hardware accelerated ffmpeg just fine. In any case i would like to say thank you to the developer since its a great and handy tool! |
Thanks for taking a look at this. I added it! :) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I have just started using this software and it seems great. However, using FFmpeg for conversion of large video files can take a lot of time if using CPU only for transcoding. Using e.g. Intel Quick Sync for hardware acceleration would probably speed things up a lot.
The text was updated successfully, but these errors were encountered: