refactor(audio): the review's cleanup list, and the design record behind the rack - #3176
Conversation
889add9 to
9fa270e
Compare
574d198 to
540056a
Compare
miga-heygen
left a comment
miga-heygen
left a comment
There was a problem hiding this comment.
Review: refactor(audio): the review's cleanup list, and the design record behind the rack — #3176
Verdict: LGTM
Four cleanups from the review's list plus the missing test that aligning the call sites exposed, plus a thorough design record.
enabledAudioFxNodes unification — three restatements of .filter((node) => node.enabled !== false) replaced with the canonical enabledAudioFxNodes(chain). This is the SSOT fix the skill flags: same predicate in two helpers, one owner. But lining up the call sites was not cosmetic — it exposed that buildFxChain's update filter had no test. The failure is silent and shape-preserving: a bypassed node shifts every later node's parameters into its neighbour, and the shape string is identical either way (bypassed nodes don't contribute), so nothing forces a rebuild.
The new test creates [highpass enabled, peaking disabled, lowpass enabled], updates the lowpass frequency, and verifies the highpass didn't move. Clean proof of the filter's correctness.
Dataset key derivation — HF_AUDIO_FX_DATA_KEY = HF_AUDIO_FX_ATTR.slice("data-".length) and HF_AUDIO_AUTOMATION_DATA_KEY likewise. The studio writes through the attribute and reads through the dataset key; the read side was carrying its own "fx-chain" literal and a rename would half-land. Both spellings now derive from one constant, with tests proving they match. Applied to audioFxSummary.ts, propertyPanelAudioFxGroup.tsx, useVolumeAutomation.ts, and propertyPanelAutomation.ts.
Shape reassignment removal — shape = shapeOf(next) was recomputed on every observer tick to write back the string that was already there (the early return proved it equal). Now const.
Narrowing vs assertion — resolveAutomationRange uses param directly after the param.kind !== "number" guard instead of casting to HfAudioFxNumberParam. The type system already narrowed it.
FxNodeRow keying — key={node.id ?? ${node.type}-${i}} instead of ${node.type}-${i}. Two effects of the same type kept their keys through a reorder, so React reused each row in place. Controls hold real state (a half-typed number, an in-flight drag), which stayed at the position and landed on whichever effect moved into that slot.
Cross-realm instanceof fix in analyse — el?.tagName !== "AUDIO" instead of instanceof HTMLAudioElement. The composition's elements belong to an iframe's document, so the constructor is from a different realm and instanceof is false for every one. Reading to plain values ({ src, start }) earlier eliminates the non-null assertions downstream.
Design record — plans/audio-fx-presets.md is a well-researched catalog proposal. The honesty about what the 15 effects can't do (no noise reduction, no pitch shift, no spectral repair) is the most valuable part — it prevents scope creep into Tier 1 territory that would need ML. The carve-as-precedent argument is strong: these scripts reuse analysis code that already ships, no new process, no new dependency, no new render-path failure mode.
No issues found.
Review by Miga
🤖 Generated with Claude Code
540056a to
382f657
Compare
9fa270e to
a2c3cb1
Compare
382f657 to
bd69e6f
Compare
a2c3cb1 to
e8d849a
Compare
bd69e6f to
d0770bf
Compare
e8d849a to
2181d38
Compare
d0770bf to
6d8c95f
Compare
2181d38 to
fea05e3
Compare
* feat(core): carve against every voice over a bed, always dynamically A bed usually runs under a whole sequence — a narrator, an interview answer, a second presenter — and carving against one of them left the others fighting it. `source` becomes `sources`, and `mixCarveSources` sums every voice onto the BED's clock before anything is measured. That is what keeps one analysis sufficient: the chain is fixed, so there is no per-voice filter to switch between, and bands drawn from all the speech there is with envelopes that rise wherever any of it happens answer the actual question — where and when is speech masking this bed. Summed rather than averaged: two people talking at once mask more than either alone. Audio before the bed starts is dropped rather than folded in at zero, since it plays over nothing and shifting it would put a cut where there is no voice. `dynamic` is gone. A fixed depth thins the bed through every pause, and once both have been heard there is no reason to want it, so every carve follows the speech. Two helpers the panel and the headless script now share instead of each carrying a copy — two definitions of "what does this name suggest" drift, and then the two disagree about which track is the voice: - `classifyAudioName` reads a track's kind from its id and filename together. `unknown` is deliberately common: treating an unrecognised name as "not a voice" would hide the one track somebody needs to pick. - `clipsOverlap` keeps out a voice that never plays while the bed does. An unwritten duration counts as unbounded, not zero — refusing a clip whose length the composition leaves to the media would drop the commonest case there is. Files written before this still load: a single `source` reads as a one-voice list, a stored `dynamic` is ignored, and an absent attribute means the defaults whole. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(core): stop \b from missing underscore-separated names, guard clipsOverlap's negative duration \b treats `_` as a word character, so \bbed\b never matched bed_01, music_bed_loop, or theme_song, and \bvo\b/\bvox\b/\btts\b had the same gap — an underscore-separated bed classified as "unknown" and could end up offered as its own carve source. Replaced the short hints with a boundary that actually excludes letters and digits on both sides. clipsOverlap computed end = start + duration without guarding sign, so a negative duration put end before start — an interval that does not describe anything, and one specific case showed it silently dropping a real overlap (a shorter, earlier broken end rejected a clip that genuinely contained the point). Duration clamps to zero instead: a clip cannot un-play time, and a zero-length clip at its start is the sane reading of "duration nobody wrote down as positive." Review by Miga (PR #3212). --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
#3212 (accidentally squash-merged into this branch instead of main) changed HfCarveSettings from a single `source` + `dynamic` toggle to a `sources` list with dynamic mode removed outright — the multi-voice UI consumer that goes with that shape lands in the very next PR, so this branch was left with a type that no longer matched its own code. Minimal port, not the multi-voice redesign that PR does properly: the "Listen to" picker and analyse() treat sources[0] as the one voice this UI still understands, and every dynamic-mode branch (the automated envelope lanes, the toggle, the checkbox) is gone along with the field — a carve is now always the static value the analysis computes, matching what the type change made permanent. Test suite trimmed the same way: the automation-lane and toggle tests covered behavior that no longer exists.
# Conflicts: # packages/studio/src/components/editor/propertyPanelAudioFxGroup.test.tsx # packages/studio/src/components/editor/propertyPanelAudioFxGroup.tsx # packages/studio/src/components/editor/propertyPanelFxSection.tsx
# Conflicts: # packages/studio/src/components/editor/propertyPanelAudioFxGroup.test.tsx # packages/studio/src/components/editor/propertyPanelAudioFxGroup.tsx # packages/studio/src/components/editor/propertyPanelFxSection.tsx # skills-manifest.json # skills/hyperframes-audio/SKILL.md # skills/hyperframes-audio/references/attributes.md # skills/hyperframes-audio/scripts/carve.mjs
Fallow audit reportFound 5 findings. Details
Generated by fallow. |
Blocks the regression workflow's required preflight gate. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2 participants
The last slice of the old #3156.
Four of the nine cleanups the review listed: the non-null assertion and two casts the guards already covered, one enabled-node helper instead of three restatements, and dataset keys derived from their attributes.
Lining up the three enabled-node call sites was not cosmetic — it exposed that
buildFxChain's filter had no test, and its failure is silent: a bypassed node shifts every later node's parameters into its neighbour with an identical shape either way. Now covered.Also carries the design record for the rack work that follows: the preset research and catalogue rationale, and the schematic direction.
🤖 Generated with Claude Code