Gaussian-windowed short-time Fourier transform (STFT) spectrogram. More...


Go to the source code of this file.
Classes | |
| struct | UTILSLIB::SpectogramInputData |
| Input parameters for short-time Fourier transform spectrogram computation. More... | |
| class | UTILSLIB::Spectrogram |
| Computes time-frequency spectrograms via short-time Fourier transform with configurable window and overlap. More... | |
Namespaces | |
| namespace | UTILSLIB |
| Shared utilities (I/O helpers, spectral analysis, layout management, warp algorithms). | |
Gaussian-windowed short-time Fourier transform (STFT) spectrogram.
SPDX-License-Identifier: BSD-3-Clause Copyright (c) 2026 MNE-CPP Authors
Spectrogram slides a Gaussian window of user-defined size along a 1-D input signal, applies an FFT to each windowed segment and stacks the resulting magnitude spectra column-wise into a time–frequency matrix. The window's standard deviation controls the classical time–frequency resolution trade-off (Heisenberg–Gabor uncertainty): wider windows give sharper frequency localisation at the expense of temporal blur, and narrower windows do the opposite.
The implementation parallelises across windowed segments via Qt Concurrent's map–reduce pattern; SpectogramInputData carries the per-window parameters and Spectrogram::reduce sums the partial spectrograms back into the final matrix. Use this when a single global STFT view is sufficient — for multitaper or wavelet representations see MultitaperTfr and MorletTfr.
Definition in file spectrogram.h.