Skip to main content

RtNoiseWorker

Namespace: RTPROCESSINGLIB  ·  Library: DSP Library

#include <dsp/rt_noise.h>

class RTPROCESSINGLIB::RtNoiseWorker

Background worker that computes a noise power spectral density estimate from accumulated data blocks.

src/examples/ex_dsp_rt/main.cpp
RtNoiseWorker noiseWorker(64, info200, 10); // 64-point FFT over 10 accumulated blocks
MatrixXd spectrum;
QObject::connect(&noiseWorker, &RtNoiseWorker::resultReady, [&](const MatrixXd& psd) { spectrum = psd; });
for (int b = 0; b < 10; ++b) {
    noiseWorker.doWork(tones.middleCols(64 * b, 64)); // dB re unit^2/Hz, channels x (fft/2 + 1)
}

Inheritance​

Public Methods​

RtNoiseWorker(iFftLength, pFiffInfo, iDataLength)​

Creates the worker.

Parameters:

  • iFftLength : qint32 FFT length (number of samples per window).

  • pFiffInfo : FIFFLIB::FiffInfo::SPtr Associated Fiff Information.

  • iDataLength : qint32 Number of blocks to accumulate before computing the spectrum.

doWork(matData)​

Process one data block, accumulate, and emit the spectrum when enough blocks are collected.

Parameters:

  • matData : const Eigen::MatrixXd & The incoming data block (channels x samples).

Example​

Used in examples​

ex_dsp_rt​

DSP real-time workers: averaging, covariance, noise spectrum, connectivity, HPI fitting and inverse operators.

src/examples/ex_dsp_rt/main.cpp, line 217
RtNoiseWorker noiseWorker(64, info200, 10); // 64-point FFT over 10 accumulated blocks
MatrixXd spectrum;
QObject::connect(&noiseWorker, &RtNoiseWorker::resultReady, [&](const MatrixXd& psd) { spectrum = psd; });
for (int b = 0; b < 10; ++b) {
    noiseWorker.doWork(tones.middleCols(64 * b, 64)); // dB re unit^2/Hz, channels x (fft/2 + 1)
}
// scipy.signal.welch(x, 200, window=hann(66)[1:-1], nperseg=64, noverlap=0, detrend=False): 25 Hz -9.6521 dB, 50 Hz -29.6587 dB
ok &= expect(spectrum.cols() == 33 && std::fabs(spectrum(0, 8) + 9.652129129685868) < 1e-6 && std::fabs(spectrum(0, 16) + 29.658719433796943) < 1e-6,
             QString("RtNoiseWorker: 25 Hz at %1 dB, 50 Hz at %2 dB match scipy.signal.welch").arg(spectrum(0, 8), 0, 'f', 4).arg(spectrum(0, 16), 0, 'f', 4));

Authors of this file​