Skip to content

Fix AudioProcessingTrack per-sample loop performance#633

Merged
BuffMcBigHuge merged 1 commit intomarco/feat/audio-sync-2from
ryan/fix/audio-track-numpy-perf
Mar 9, 2026
Merged

Fix AudioProcessingTrack per-sample loop performance#633
BuffMcBigHuge merged 1 commit intomarco/feat/audio-sync-2from
ryan/fix/audio-track-numpy-perf

Conversation

@ryanontheinside
Copy link
Copy Markdown
Collaborator

@ryanontheinside ryanontheinside commented Mar 9, 2026

Summary

  • Replace per-sample Python loops in AudioProcessingTrack with vectorized numpy operations, addressing review feedback
  • Add 42 tests covering interleaving, buffering, resampling, channel conversion, frame construction, and full recv() integration with adversarial inputs

Changes

src/scope/server/tracks.py

  • Audio buffer changed from deque[float] to np.ndarray (float32)
  • Channel interleaving now uses np.ravel(order="F") instead of nested for loops
  • Frame extraction uses numpy slicing instead of per-sample popleft()
  • _create_audio_frame accepts np.ndarray directly, skipping redundant conversion

tests/test_audio_processing_track.py (new)

  • Interleaving correctness (stereo ordering, mono passthrough, multi-channel)
  • Buffer accumulation (exact boundaries, undersized, multi-frame drain, small chunk accumulation)
  • Frame construction (clipping, silence, mono/stereo layout)
  • Adversarial inputs (zero-length, single sample, NaN, Inf, large chunks, float64, DC offset)
  • Resampling (passthrough, up/down, odd ratios, edge cases)
  • Full recv() integration (no audio, resampling, mono upmix, 1D tensor, pause, accumulation)

Test plan

  • All 42 new tests pass (uv run pytest tests/test_audio_processing_track.py)
  • Module imports cleanly
  • @BuffMcBigHuge Manual test with LTX-2 audio pipeline

@ryanontheinside
fix: replace per-sample Python loops with vectorized numpy ops in Aud…
38486db 
…ioProcessingTrack

Addresses review feedback on #534. The audio buffer interleaving and
frame extraction used O(n) Python loops over individual samples, which
is expensive for real-time audio. Now uses np.ravel(order="F") for
interleaving and numpy slicing for frame extraction.

Also adds 42 tests covering interleaving, buffering, resampling,
channel conversion, frame construction, and adversarial inputs.

Signed-off-by: RyanOnTheInside <[email protected]>
@coderabbitai
Copy link
Copy Markdown

coderabbitai bot commented Mar 9, 2026

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: e265dfbb-10bd-49c3-852a-5f17a4b1607e

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Post copyable unit tests in a comment
  • Commit unit tests in branch ryan/fix/audio-track-numpy-perf

Comment @coderabbitai help to get the list of available commands and usage tips.

@ryanontheinside ryanontheinside mentioned this pull request Mar 9, 2026
Merged
BuffMcBigHuge
BuffMcBigHuge approved these changes Mar 9, 2026
View reviewed changes
Copy link
Copy Markdown
Collaborator

@BuffMcBigHuge BuffMcBigHuge left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Mar 9, 2026

🚀 fal.ai Preview Deployment
App ID daydream/scope-pr-633
WebSocket wss://fal.run/daydream/scope-pr-633/ws
Commit 38486db

Testing

Connect to this preview deployment by setting the fal endpoint in your client:

FAL_WS_URL=wss://fal.run/daydream/scope-pr-633/ws

🧪 E2E tests will run automatically against this deployment.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Mar 9, 2026

✅ E2E Tests passed
Status passed
fal App daydream/scope-pr-633
Run View logs

Test Artifacts

Check the workflow run for screenshots.

@BuffMcBigHuge BuffMcBigHuge merged commit 2c445ba into marco/feat/audio-sync-2 Mar 9, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@BuffMcBigHuge BuffMcBigHuge BuffMcBigHuge approved these changes

@j0sh j0sh Awaiting requested review from j0sh

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@ryanontheinside @BuffMcBigHuge