RESPITE: The CASA Toolkit Page: Documentation: Block Library Index:StationaryNoiseEstimation


This block calculates a stationary noise estimate.

The input data should be 1-D frames representing the estimate signal spectra. It is assumed that the first few frames of data contain only noise and these frames are used to form the noise estimate. The number of initial frames to employ can be set using the parameter NFRAMES.

The noise estimate is formed from the average of the initial frames and is sent to out1.

If IGNORE_LARGE is set then the values in the first NFRAMES are sorted and the larger 50% are discarded before calculating the average. (This feature may be useful for getting a crude estimate of the noise floor for signals that are not entirely stationary).

A clean signal spectrum estimate is calculated by subtracting the noise estimate from the noisy input signal. This is sent to out2. It should be noted that because the noise may be overestimated, the clean signal estimate may have negative values, even when working in the power spectral domain.

The input signal variance is estimated over the initial NFRAMES and is sent to out3.

Inputs Meaning Sample 1-D frame $\ge$2-D frame
in1 noisy spectral data Yes Yes Yes

Outputs Meaning
out1 stationary noise spectrum estimate
out2 signal spectrum estimate
out3 noise variance estimate

Parameters Type Default Meaning
NFRAMES Integer - Number of initial frames to use for forming noise estimate
IGNORE_LARGE Boolean False Ignore larger values when making initial estimate

Documentation for CTKv1.1.4 - Last modified: Thu Jun 28 15:43:35 BST 2001