Splits BuiltCapture/sourceRateRefusalText into processor_snapshot.h; fixes stale lock-nesting notes; restamp now also publishes liveContours_ like reload; tightens a slack test bound and one drifted comment.
The splice crossfade now counts output frames and stops before the ring's far end (this also fixes a lap at matched rate). The search strides by the rate ratio. The editor takes its time base from its own decode. Same-rate output is pinned to hashes from before the change.
Editor close mid-drag now rolls back published contours like capture-loss does; bake detaches liveContours so a printed capture can't observe a live drag; a held note now receives a live publish that lands in the same block as a note-on; amp spline read clamped to [0,1].
The voice now converts the file's rate to the host's at note-on instead of assuming they match. Stage times resolve at the host rate, the filter solves at it, and the bake follows. Same-rate renders are unchanged bit for bit.
Contours ride their own seqlock block into a double-buffered engine copy, and the swap step is smoothed. Each drag move publishes contours without a model write; release still commits and reloads.
ReleaseProof{} and copy-reuse both compiled under this project's C++17;
user-provided ctor, deleted copy ctor and friend close them. The skip now
compares stored values, not norms, so it actually fires. Abort downgraded
to a debug assert.
Points toPlain's exponent arm at the undetented curve map so host reads
match the editor; adds a sizeof guard plus field-poison test for
LiveValues::operator==; skips the model write when an automation value
hasn't moved; corrects five stale doc citations.
The SDK's own single-component sample drains inputParameterChanges in
process() and implements setParamNormalized; automation was reading the
GUI channel alone. The audio thread now patches a block it solely owns.
30 Hz out-of-band energy 15.45% -> 0.00%; the 29 Hz rate-2.0 detune -133 -> +0 cents.
An unknown period keeps the fixed-window geometry bit for bit. The detector cannot
reach process(): sampler_core does not link it.
Retire the effectiveGain blend so the meter's minimum tracks smoothGain's own
reduction against real input, unscaled by the transition mute — a toggle over
quiet material now reads no reduction instead of pinning to 0.
Fold setLimiterEnabled's restart request into setInstrumentParams so every writer
keeps the host's latency report in sync. Pin the window-sizing identity, drop the
per-sample modulo, tighten the ceiling tolerance, publish the blended gain.
Zero-crossing counting was anti-correlated with the real defect (splice debris
fools it). Now asserts energy outside the fundamental, with an alignable control,
matching test_preserve_low_frequency.cpp's approach.
A splice can only relocate by [0.75, 1.25]*window, so periods with no multiple
in that interval never phase-align — at 50 ms, f < 16 Hz and 26.7-32 Hz.
Harness runs by hand; too slow to gate.
Re-derives the splice-cadence inequality and adds a corner probe that FAILS at
P=500 by design, pending a ruling. Names the baseline commit and harness edit,
fixes measurement methodology, corrects the Trigger-AHD/rate coupling framing.
Generalizes the correlation-aligned SOLA delay line so the feed and the shift are
independent rates over one ring. Unity is bit-identical to the shipped read, asserted
against a hash baseline captured pre-change.
Extracts resolveWaveformClaim (core/instrument/ui/spline_edit) so the shell's node/tab/marker click resolution is unit-tested directly, not just its input geometry; folds the staged-envelope node into it; fixes comment accuracy, a cost regression, and test fidelity issues.
Waveform overlay now resolves node/tab/marker click collisions by target area instead of check order; residue test now uses a distinguishing fixture; Gate-unavailable-while-drawn logic extracted to one pure helper shared by resolvePlay and applyControl.
Fixes the 2-point spline-EG early-free bug causing silent fade-ins, the pitch/filter enable-toggle Trigger-forcing hole, the contour-node/marker pixel shadow, missing deck-residue test coverage, and stale comments in knob_deck and spline_edit.
Live pitch depth, Gate/Spline enable-rule agreement, inert kTrigLength, NaN wire guards, hard-flag-tail corruption no longer wipes the record, RT/cold spline tie-break, retired alt-click, marker-shadow fix, plus new test coverage.
Cancels the curve-node drag whenever the popup closes so Esc mid-drag can't alias the amp curve; generalizes CurveTarget routing to one switch; fixes stale/overstated comments; clamps a pre-v12 depth fold; adds deck-inertness and filter-fold test coverage.
Payload v12 appends the new velocity->pitch curve and folds the retired filter velAmount into its now-bipolar curve, so pre-v12 projects reopen sounding identical. Preview button takes a drawn play triangle.
Normalizes crossfadeWeight over crossfade-1 so the last rendered frame lands at exactly the incoming tap instead of a residual step; corrects the CLAUDE.md invariant and seam test to match. Shares lerpSource/crossfadedSource/maxCrossfade, fixes stale docs/constants, and clears crossfade on the loop-OFF gesture.
Trigger's fade pair folds into the AHD (and goes live); the release anchors right;
Preserve rings its synthetic tail out instead of cutting it. Payload v10.
Cutoff-only re-solve (15.5 vs 56.9 ns/frame) makes the unquantized sweep
affordable, replacing the 2048-step mod quantizer. Live-compute parameters
remain blocked on a shell-architecture ruling.
Codec fallback for non-finite filter fields, a modAmount==0 early-out in tickFilterCutoff,
new stereo render tests for the dual-mono mirror, and comment/test accuracy fixes flagged
in review (stale deck-width claims, restated invariants, drifted CMake link comments).
daniel