|
| | AccountedAllocator.h |
| | Internal allocator with optional per-owner payload accounting. Threading: the account belongs to one setup/worker operation. No global state.
|
| |
| | BoxAverage.h |
| | Internal moving-average arithmetic shared by live and offline gain computers. Threading: the caller owns the running sum and sample history.
|
| |
| | ClipperShape.h |
| | Shared stateless clipping curves, without gain, filtering or resampling.
|
| |
| | ContinuousClip.h |
| | Shared continuous reconstruction and integration of clipping curves.
|
| |
| | ContinuousStereoColor.h |
| | Source-rate rational color using continuous interval integration.
|
| |
| | ConvolutionRate.h |
| | Internal mono FIR rate conversion with an optional matched reference.
|
| |
| | DcBlock.h |
| | Shared one-pole DC-blocking state and difference-form recurrence.
|
| |
| | LogCosh.h |
| |
| | OfflineAttenuation.h |
| | Internal source-aligned minimum hold, release and four-box envelope. Threading: construction is worker-owned; completed curves are immutable.
|
| |
| | OfflineBoxSmoother.h |
| | Internal finite four-box kernel shared by source-aligned gain plans. Threading: worker-owned setup and cursors; a prepared kernel is immutable.
|
| |
| | OfflineClip.h |
| | Shared complete-source planning, calibration and transactional clipping. Threading: worker only; returned plans are immutable and independently reusable.
|
| |
| | OfflineClipProjection.h |
| | Bounded complete-source clipping with ideal sinc reconstruction. Internal offline worker engine; reuses Core curves, continuous integration, FIR compensation and finite-source sinc projection. No hidden gain or limiter. Threading: one worker owns each nonmovable instance.
|
| |
| | OfflineConvolutionWindow.h |
| | Exact three-leaf Toeplitz convolution on one requested output leaf. Uses Core FFTReal and complex multiplication; no stream state or PCM history. Threading: one offline worker owns each instance. Setup reserves all storage.
|
| |
| | OfflineGain.h |
| | Internal verified bandlimited gain rendering. Threading: synchronous worker-only operations; plans and sources stay immutable.
|
| |
| | OfflineGainSource.h |
| | Internal verified PCM/control cache and reusable offline product maps. Threading: worker-owned; the source and control plan remain immutable.
|
| |
| | OfflineProduct.h |
| | Internal finite-source bandlimited multiplication for offline workers.
|
| |
| | OfflinePunchEnvelope.h |
| | Internal sparse, source-aligned transient boost weights. Threading: worker-owned setup and cursors; completed curves are immutable.
|
| |
| | OfflineSinc.h |
| | Complete finite-source sinc phases and source-rate projection. Reuses Core convolution and Cauchy moments. Bounded PCM scratch; coarse maps grow with the source length. Readers supply immutable finite samples and may be rewound. Threading: one worker owns each nonmovable instance.
|
| |
| | OfflineStereo.h |
| | Shared complete-source delta transport for offline stereo generation. Threading: one worker owns each stream, source and provisional consumer.
|
| |
| | OfflineWorker.h |
| | Shared worker validation, exact exclusions and owning output transactions. Threading: synchronous worker operations; sources and plans remain immutable.
|
| |
| | RationalKnee.h |
| | Shared rational soft-knee curve and its antiderivative.
|
| |
| | RetimedDcBlock.h |
| | Internal source-clock DC transfer using shared Core recurrence and FIRs.
|
| |
| | ScaledSumSquares.h |
| | Shared overflow/underflow-safe finite-sample RMS accumulator. Threading: one stream owner. Fixed storage, no allocations. Dependencies: C++20 standard library only.
|
| |
| | StereoColor.h |
| | Internal rational color branch for stereo generation.
|
| |
| | StereoColorCoefficients.h |
| | Fixed source-rate continuous color integration filters. Generated by tools/generate_stereo_color.cpp; do not hand edit. Kaiser beta 18.5, cutoff 0.475 Fs, support +/-128 samples; degree-seven reconstruction and moment projection. Chebyshev fits on each unit interval. Projection includes the source-clock DC correction for S=1,2,4,8. Shared read-only payload: 82496 bytes, independent of instances.
|
| |
| | TptSvf.h |
| | Shared trapezoidal state-variable integrator, without parameter policy.
|
| |