Unable to specify ffmpeg hwaccel_args

[thinkloop]

When I set ffmpeg hardware acceleration like so:

ffmpeg:

  # hardware acceleration
  hwaccel_args: preset-vaapi

  # enable audio
  output_args:
    record: preset-record-generic-audio-aac

No image appears on the cam and I get the error:

Option hwaccel (use HW accelerated decoding) cannot be applied to output url vaapi -- you are trying to apply an input option to an output file or vice versa. Move this option before the file it belongs to.

I tried preset-intel-qsv-h264 as well as setting the environment variable in docker-compose LIBVA_DRIVER_NAME=i965, but got the same results.

I am running an Intel Core i5-12400 Desktop Processor and a Coral M.2 TPU, no GPU.

If I omit the hwaccel_args argument completely everything works fine.

Should I simply continue without it?

[blakeblackshear]

Can you post your entire config?

[thinkloop]

Hey,

Can you post your entire config?

ui:
  # Optional: Set the default live mode for cameras in the UI (default: shown below)
  live_mode: mse
  # Optional: Set a timezone to use in the UI (default: use browser local time)
  # timezone: None
  # Optional: Use an experimental recordings / camera view UI (default: shown below)
  #experimental_ui: False

database:
  path: /db/frigate.db
  
mqtt:
  host: 127.0.0.1
  user: uuu
  password: ppp
  port: 1883

detectors:
  coral1:
    type: edgetpu
    device: pci:0
    
ffmpeg:

  # hardware acceleration
  hwaccel_args: preset-vaapi

  # enable audio
  output_args: 
    record: preset-record-generic-audio-aac
    
detect:
  enabled: true
  # Optional: width of the frame for the input with the detect role (default: shown below)
  width: 720
  # Optional: height of the frame for the input with the detect role (default: shown below)
  height: 480
  # Optional: desired fps for your camera for the input with the detect role (default: shown below)
  # NOTE: Recommended value of 5. Ideally, try and reduce your FPS on the camera.
  fps: 5

motion:
  # Optional: The threshold passed to cv2.threshold to determine if a pixel is different enough to be counted as motion. (default: shown below)
  # Increasing this value will make motion detection less sensitive and decreasing it will make motion detection more sensitive.
  # The value should be between 1 and 255.
  threshold: 50
  
birdseye:
  # Optional: Enable birdseye view (default: shown below)
  enabled: True
  # Optional: Width of the output resolution (default: shown below)
  width: 1280
  # Optional: Height of the output resolution (default: shown below)
  height: 720
  # Optional: Encoding quality of the mpeg1 feed (default: shown below)
  # 1 is the highest quality, and 31 is the lowest. Lower quality feeds utilize less CPU resources.
  quality: 1
  # Optional: Mode of the view. Available options are: objects, motion, and continuous
  #   objects - cameras are included if they have had a tracked object within the last 30 seconds
  #   motion - cameras are included if motion was detected in the last 30 seconds
  #   continuous - all cameras are included always
  mode: objects
  
record:
  enabled: true
  retain:
    days: 30
    mode: motion  
  events:
    # Optional: Number of seconds before the event to include 
    pre_capture: 2
    # Optional: Number of seconds after the event to include 
    post_capture: 2
    # Optional: Objects to save recordings for. (default: all tracked objects)
    #objects:
    #  - person
    # Optional: Restrict recordings to objects that entered any of the listed zones (default: no required zones)
    required_zones: []
    # Optional: Retention settings for recordings of events
    retain:
      # Required: Default retention days (default: shown below)
      default: 365
      # Optional: Mode for retention. (default: shown below)
      #   all - save all recording segments for events regardless of activity
      #   motion - save all recordings segments for events with any detected motion
      #   active_objects - save all recording segments for event with active/moving objects
      #
      # NOTE: If the retain mode for the camera is more restrictive than the mode configured
      #       here, the segments will already be gone by the time this mode is applied.
      #       For example, if the camera retain mode is "motion", the segments without motion are
      #       never stored, so setting the mode to "all" here won't bring them back.
      mode: active_objects
      # Optional: Per object retention days
      objects:
        person: 730        
        
snapshots:
  enabled: true
  # Optional: print a timestamp on the snapshots (default: shown below)
  timestamp: true

rtmp:
  enabled: false

#restream:
  # Optional: Enable the restream (default: True)
#  enabled: true
  
timestamp_style:
  # Optional: Format specifier conform to the Python package "datetime" (default: shown below)
  #           Additional Examples:
  #             german: "%d.%m.%Y %H:%M:%S"
  format: "%Y/%m/%d %H:%M:%S"
       
cameras:
  lionel-driveway:
    ffmpeg:
      inputs:
        - path: rtsp:https://uuu:[email protected]:554/11
          roles:
            - record
        - path:  rtsp:https://uuu:[email protected]:554/12  
          roles:
            - detect
    detect:
      width: 640
      height: 352
      fps: 5
    ui:
      order: 1

This looks like #6008 (comment) which was a host driver issue

Now that you mention it I am seeing some threads that Debian might not have the latest drivers, not sure at all yet tho, will investigate further.

[NickM-27]

This looks like #6008 (comment) which was a host driver issue

[blakeblackshear]

So is this an issue when the automatic device detection doesn't work for the hwaccel args?

[NickM-27]

From what I can see on the issue the ffmpeg command looks like expected, not entirely sure

[thinkloop]

I tried installing various Intel drivers on the Debian instance, to no avail. I am not sure how to proceed. Is it an OS issue? A Frigate issue? Are there any commands I can try to see if everything is setup correctly? Would using Ubuntu solve the problem? It's my first time with Debian. Thanks for any advice.

[blakeblackshear]

I run all my stuff on Debian, so this should work fine. I don't know how you setup your server.

Are you running debian server? Did you install docker according to the official docker documentation for debian? Maybe try running docker compose with sudo?

[20BBrown14]

@thinkloop did you make any progres on this? I'm suddenly having the same issue with basically the same setup.

[thinkloop]

Sorry hadn't seen this until now. The issue for me was that stock stable Debian was too behind. Had to install one of the more advanced branches, Trixie I believe.

[jehy]

I had absolutely same issue on ubuntu 22 with docker, and adding privileged: true resolved it. I had CAP_PERFMON before, but looks like it is not enough.

[neoteknic]

privileged: true ? where exactly ? thank you

[iursevla]

@jehy yeah this solved my issue.

@neoteknic I assume you already fixed it but it should be in the docker compose

Documentation: https://deploy-preview-11419--frigate-docs.netlify.app/configuration/hardware_acceleration/#docker-compose---privileged

[Swizzler121]

Has anyone been able to resolve this? Just set up a frigate instance, the host container detects the intel stuff fine, but if I enable preset-vaapi all my video feeds die. I'm confused why this hasn't been escalated to a bug if there's no answer. I'm already running it as a privileged container so that's not the fix.

root@docker-frigate:/opt# ls -l /dev/dri/
total 0
drwxr-xr-x 2 root root        80 Mar 18 20:20 by-path
crw-rw---- 1 root video 226,   0 Feb 28 08:56 card0
crw-rw---- 1 root input 226, 128 Feb 28 08:56 renderD128

root@docker-frigate:/opt# vainfo
error: XDG_RUNTIME_DIR is invalid or not set in the environment.
error: can't connect to X server!
libva info: VA-API version 1.17.0
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so
libva info: Found init function __vaDriverInit_1_17
libva info: va_openDriver() returns 0
vainfo: VA-API version: 1.17 (libva 2.12.0)
vainfo: Driver version: Intel iHD driver for Intel(R) Gen Graphics - 23.1.1 ()
vainfo: Supported profile and entrypoints
      VAProfileMPEG2Simple            : VAEntrypointVLD
      VAProfileMPEG2Main              : VAEntrypointVLD
      VAProfileH264Main               : VAEntrypointVLD
      VAProfileH264Main               : VAEntrypointEncSliceLP
      VAProfileH264High               : VAEntrypointVLD
      VAProfileH264High               : VAEntrypointEncSliceLP
      VAProfileJPEGBaseline           : VAEntrypointVLD
      VAProfileJPEGBaseline           : VAEntrypointEncPicture
      VAProfileH264ConstrainedBaseline: VAEntrypointVLD
      VAProfileH264ConstrainedBaseline: VAEntrypointEncSliceLP
      VAProfileVP8Version0_3          : VAEntrypointVLD
      VAProfileHEVCMain               : VAEntrypointVLD

root@docker-frigate:/opt# lshw -c video
  *-display                 
       description: VGA compatible controller
       product: HD Graphics 530
       vendor: Intel Corporation
       physical id: 2
       bus info: pci@0000:00:02.0
       logical name: /dev/fb0
       version: 06
       width: 64 bits
       clock: 33MHz
       capabilities: pciexpress msi pm vga_controller bus_master cap_list rom fb
       configuration: depth=32 driver=i915 latency=0 resolution=1920,1080
       resources: irq:133 memory:f6000000-f6ffffff memory:e0000000-efffffff ioport:f000(size=64) memory:c0000-dffff

[peahonen]

I ran into the same error message. In my case it was sufficient to map the device into the container in docker-compose.yml:

services:
  frigate:
    devices:
      - /dev/dri/renderD128:/dev/dri/renderD128