Skip to content

Implementing video processing strategy with chunks#595

Closed
Copilot wants to merge 2 commits into
mainfrom
copilot/video-processing-chunks
Closed

Implementing video processing strategy with chunks#595
Copilot wants to merge 2 commits into
mainfrom
copilot/video-processing-chunks

Conversation

Copy link
Copy Markdown

Copilot AI commented May 26, 2026

Pull request created by AI Agent

Copilot AI added 2 commits May 26, 2026 22:16
… in VideoWriter

VideoNode._get_audio_chunk_for_frame now includes 'chunk_index' and
'step_duration' in the returned audio dict.  VideoWriterNode uses these
to deduplicate the sliding-window chunks: only the first step_duration
seconds of each new chunk_index is appended, reconstructing a
continuous, correctly-timed audio track (previously the same 5-second
chunk was appended once per frame, inflating audio by ~fps× the
chunk duration).

Also fixes the outdated assertion in test_video_audio_integration.py
(the return dict gained a 'timestamp' field in a prior commit) and adds
test_video_pipeline_audio_dedup.py covering the new deduplication logic.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants