fix(audio): code-review fixes — clip-before-duck, phaser lanes, rate reschedule - #3156

Closed

vanceingalls wants to merge 28 commits into

wa-18-lane-stretchfrom

wa-19-review-fixes

Commits on Aug 10, 2026

fix(engine): hand an empty audio track back instead of failing the render
vanceingalls
committed
726fd13
fix(engine): take the browser lease inside the try that releases it
vanceingalls
committed
e5fbac4
fix(core): aim the phaser's trim lanes at the trims, not at the pinned wet/dry
vanceingalls
committed
213c5f6
fix(core): move a node's id with its slot when the chain updates in place
vanceingalls
committed
f83b20e
fix(core): sample the volume lane at clip-local time, the way the render bakes it
vanceingalls
committed
2fb80b1
fix(studio): let an FX parameter be typed, and stop a bare blur re-persisting it
vanceingalls
committed
480c43e
fix(lint): stop the double-automation rule blaming the wrong element
vanceingalls
committed
c9c2d3f
build(core): regenerate the audio-fx runtime before the tests run
vanceingalls
committed
1b57a95
fix(core): dispose a clip's FX graph when it ends, not only when it is stopped
vanceingalls
committed
40679e9
fix(studio): lock the FX rack while the carve is measuring
vanceingalls
committed
4872956
fix(engine): give the ffmpeg children the signal that actually aborts them
vanceingalls
committed
7a1adbc
fix(engine): duck the FX output before quantising it, not after

vanceingalls
and
claude
committed
32f9f01
fix(engine): move a track's PCM across the wire in bounded chunks

vanceingalls
and
claude
committed
d9229b5
fix(core): reschedule FX automation when the transport changes rate

vanceingalls
and
claude
committed
b1041a0
perf(core): reuse the FFT scratch arrays across carve windows

vanceingalls
and
claude
committed
8b68ce8
fix(core): let a failed worklet registration be retried

vanceingalls
and
claude
committed
674927a
fix(core): retire a worklet processor on dispose instead of only disconnecting it

vanceingalls
and
claude
committed
644633c
fix(core): stop reading a blank FX parameter as zero

vanceingalls
and
claude
committed
267161c
fix(producer): classify the audio failure that takes the whole mix down

vanceingalls
and
claude
committed
222390d
chore: drop a dead audio-FX ignoreExports entry

vanceingalls
and
claude
committed
889add9
fix(studio): key FX rows by node id so a mid-edit stays with its effect

vanceingalls
and
claude
committed
cfc993c
refactor: drop the non-null assertion and two casts the guards already cover

vanceingalls
and
claude
committed
0da747e
refactor(core): route the graph's enabled-node filter through one helper

vanceingalls
and
claude
committed
702dfa5
refactor: derive the dataset keys for the FX and automation attributes

vanceingalls
and
claude
committed
1de1afb
docs(plans): research audio treatment presets over the shipped effect set

vanceingalls
and
claude
committed
cbdc7c6
docs(plans): add rack module identity and the preset menu to the presets doc

vanceingalls
and
claude
committed
ddf2dff
docs(design): pick the schematic direction for the FX rack and record why

vanceingalls
and
claude
committed
509d789
docs(design): automated parameters read live instead of drawing their lane

vanceingalls
and
claude
committed
574d198

Navigation Menu

Uh oh!

fix(audio): code-review fixes — clip-before-duck, phaser lanes, rate reschedule - #3156

fix(audio): code-review fixes — clip-before-duck, phaser lanes, rate reschedule#3156
vanceingalls wants to merge 28 commits into
wa-18-lane-stretchfrom
wa-19-review-fixes

Commits on Aug 10, 2026

fix(engine): hand an empty audio track back instead of failing the render

fix(engine): take the browser lease inside the try that releases it

fix(core): aim the phaser's trim lanes at the trims, not at the pinned wet/dry

fix(core): move a node's id with its slot when the chain updates in place

fix(core): sample the volume lane at clip-local time, the way the render bakes it

fix(studio): let an FX parameter be typed, and stop a bare blur re-persisting it

fix(lint): stop the double-automation rule blaming the wrong element

build(core): regenerate the audio-fx runtime before the tests run

fix(core): dispose a clip's FX graph when it ends, not only when it is stopped

fix(studio): lock the FX rack while the carve is measuring

fix(engine): give the ffmpeg children the signal that actually aborts them

fix(engine): duck the FX output before quantising it, not after

fix(engine): move a track's PCM across the wire in bounded chunks

fix(core): reschedule FX automation when the transport changes rate

perf(core): reuse the FFT scratch arrays across carve windows

fix(core): let a failed worklet registration be retried

fix(core): retire a worklet processor on dispose instead of only disconnecting it

fix(core): stop reading a blank FX parameter as zero

fix(producer): classify the audio failure that takes the whole mix down

chore: drop a dead audio-FX ignoreExports entry

fix(studio): key FX rows by node id so a mid-edit stays with its effect

refactor: drop the non-null assertion and two casts the guards already cover

refactor(core): route the graph's enabled-node filter through one helper

refactor: derive the dataset keys for the FX and automation attributes

docs(plans): research audio treatment presets over the shipped effect set

docs(plans): add rack module identity and the preset menu to the presets doc

docs(design): pick the schematic direction for the FX rack and record why

docs(design): automated parameters read live instead of drawing their lane