Short time Fourier transform

Canonical STFT — batch / stream / analyse with normalized Hann OLA; stream ≡ batch under any chunking

Spectral editing — gain on time×frequency regions via COLA-normalized STFT (Audacity/afftfilt class)

Sinusoidal analysis — STFT peaks (parabolic) + nearest-frequency partial tracking (McAulay-Quatieri 1986)

A Node-RED node to decode wav file and perform short-term Fourier transform (STFT)

Lightweight monophonic pitch (f0) detection for Node and the browser. 106k-parameter ONNX model with a dual-resolution STFT + instantaneous-frequency frontend.

Shared DSP utilities for the buffered-audio-nodes ecosystem.

Phase-vocoder engine — peak-gated and peak-locked spectral bin scatter (Bernsee, Laroche-Dolson), phase locking, peak finding; powers @audio/shift-pvoc* and @audio/stretch-pvoc*

Audio-facing window kit — typed fills (periodic/symmetric), apply, COLA check over window-function

Noise PSD estimation — minimum statistics (Martin 2001), IMCRA (Cohen 2003), and quiet-segment profiling for spectral denoisers

Short time Fourier transform

Spectral flux onset detection

Voice activity detection — frame-level speech/non-speech decision (energy + spectral flatness) plus per-bin speech-presence probability and decision-directed a-priori SNR

Onset detection — spectral-flux and energy-flux onset detection functions (ODFs) + adaptive peak picking

Full beat detection pipeline: onset detection → tempo estimation → beat grid

Real-time AI noise suppression (DeepFilterNet3) for the browser/Electron: observability, AutoGain, offline cache, licensing

Shared DSP utilities for the buffered-audio-nodes ecosystem.

Tempogram — local tempo over time via onset-envelope autocorrelation

This module for Node-RED contains a set of nodes which offer audio feature extraction functionalities. Such nodes have a python core that runs on Librosa library.

denoise — content-aware auto-selector that classifies the dominant noise type