Wire SetComplexity to the spreading decision#150
Merged
Conversation
Encoder Quality ReportStatus: pass Tier 1 — SNR regression (96 kbps, pion encode → pion decode)Delta = baseline − current SNR; positive = regression. Fail threshold: 1.5 dB.
Tier 2 — opus_compare vs libopus (96 kbps CBR)Weighted error: lower is better. The gap reflects pion lacking constrained VBR; libopus ships with it enabled by default.
Run outputBaseline: |
RFC 6716 / 8251 conformationStatus: pass The action extracts the RFC 6716 reference implementation, applies the RFC 8251 decoder update patch, and then builds the patched reference tools. Legend: numeric cells are Inputs use the shared RFC 6716 / RFC 8251 bitstream corpus; accepted references follow RFC 8251 Section 11.
Run output |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #150 +/- ##
==========================================
+ Coverage 87.89% 87.96% +0.06%
==========================================
Files 32 32
Lines 7676 7692 +16
==========================================
+ Hits 6747 6766 +19
+ Misses 702 700 -2
+ Partials 227 226 -1
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
FrantaBOT
approved these changes
Jul 16, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
``SetComplexity
/WithComplexity` existed since #117 but were pure no-ops — the CELT encoder never looked at the value. This makes complexity actually do something, and bumps the default from 0 to 5 to match libopus's own default (`st->complexity = 5` in `celt_encoder.c`).First pass I gated the pitch pre-filter, dynalloc analysis, allocation trim, and dual stereo decision all behind
complexity > 0, figuring "low complexity = skip the expensive analysis". Ran the quality harness signals through it and complexity=0 came back better than complexity=5 by 5-10 dB SNR on most of them, which made no sense for a "cheaper/lower quality" setting. Isolated it by forcing each of the four back on one at a time — turns out it's entirely the pre-filter. libopus doesn't gate any of those four on complexity at all (they always run); the only thing complexity actually gates is the spreading decision, and even that isn't a simple on/off — below complexity 3 it falls back to a flatSPREAD_NORMALinstead of runningspreading_decision(), and only complexity exactly 0 getsSPREAD_NONE(celt_encoder.c~line 2317). Fixed it to match that.So today
complexityis fairly subtle: 0-2 skip the spreading estimator, 3-10 all behave identically (pion doesn't have the higher-complexity-only stuff libopus has — transient re-detection at ≥5, second MDCT pass at ≥8, etc. — those aren't ported). Still useful groundwork for wiring those in later without touching the public API again.Reference issue
Fixes #...