Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Unable to create task with image quality 100 #4461

Open
2 tasks done
Marishka17 opened this issue Mar 15, 2022 · 6 comments
Open
2 tasks done

Unable to create task with image quality 100 #4461

Marishka17 opened this issue Mar 15, 2022 · 6 comments
Assignees
Labels
bug Something isn't working good first issue Good for newcomers

Comments

@Marishka17
Copy link
Contributor

My actions before raising this issue

Expected Behaviour

Task should be created

Current Behaviour

[Errno 22] Invalid argument; last error log: [libx264] Error setting profile baseline.

Possible Solution

Remove 'profile': 'baseline'
Note that usage of -profile:v is incompatible with lossless encoding

Steps to Reproduce (for bugs)

  1. select video file
  2. use_zip_chunks=False
  3. set image quality to 100
  4. try to create task

Context

Your Environment

  • Git hash commit (git log -1): 2b815922
  • Docker version docker version (e.g. Docker 17.0.05):
  • Are you using Docker Swarm or Kubernetes?
  • Operating System and version (e.g. Linux, Windows, MacOS):
  • Code example or link to GitHub repo or gist to reproduce problem:
  • Other diagnostic information / logs:
    Logs from `cvat` container

Next steps

You may join our Gitter channel for community support.

@Marishka17 Marishka17 added the bug Something isn't working label Mar 15, 2022
@bsekachev bsekachev added the good first issue Good for newcomers label Mar 11, 2024
@EBayego
Copy link

EBayego commented Mar 28, 2024

Hi! I'll take this issue to resolve it. @nmanovic

@EBayego
Copy link

EBayego commented Mar 30, 2024

Hi there,
After several attempts, I haven't been able to reproduce the error. It works for me, and I can create the task following the steps you mentioned, using different videos too. Although these are basic questions, are you using the latest version? Is it possible that the video is damaged and cannot be encoded because of this? Have you tried with multiple videos to check this last point?

@nmanovic
Copy link
Contributor

nmanovic commented Apr 1, 2024

@Marishka17 , could you please provide a way to reproduce the issue?

@nmanovic
Copy link
Contributor

nmanovic commented Apr 1, 2024

@EBayego , please select another issue to contribute. Thank you.

@EBayego
Copy link

EBayego commented Apr 1, 2024

@EBayego , please select another issue to contribute. Thank you.

Hi Nikita, I commented in this other issue #4747

@Sanjay6388
Copy link

Solutions:

  1. Remove or Adjust the profile Parameter
    Edit the CVAT configuration to remove or adjust the profile parameter in the video encoding settings.

Locate the CVAT configuration file or scripts used for task creation (e.g., cvat/settings/base.py or any custom configuration).
Search for the ffmpeg options or settings related to profile.
Remove or adjust the 'profile': 'baseline' setting to use a more compatible profile, like main or high.

  1. Use use_zip_chunks=True or Lower image_quality
    If image_quality=100 is not a strict requirement:

Set image_quality to a lower value (e.g., 90 or 95) to avoid triggering lossless encoding.
Enable use_zip_chunks=True to bypass video re-encoding entirely. This approach uses pre-extracted image frames instead of processing the video.

  1. Update CVAT and FFmpeg
    Ensure your CVAT installation and FFmpeg version are up-to-date:

Update CVAT to the latest version if possible. This ensures any bugs or misconfigurations in the encoding pipeline are resolved.
Check the version of FFmpeg installed within the CVAT Docker container:
bash

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working good first issue Good for newcomers
Projects
None yet
Development

No branches or pull requests

5 participants