Noise zones are defined from the analysis of selected regions in the signal. A model containing the spectral description of the noise spectra is then used to substract noisy components in the signal.
A model of the noise is defined with noise fingerprints. These noise fingerprints then allow to identify and remove undesired components in a sound. Ideally, a noise zone is defined in the part of the signal that has no sinusoidal content, so that the noise model is relevant. We suppose that the noise will remain constant in the whole processed part of the signal.
When processing the signal, only the components whose amplitude is superior to the threshold defined by the fingerprints, relatively to the signal's amplitude, will be preserved. This threshold can be modified before processing.
The analysis parameters are defined in the Real Time Processing Settings dialogue window.
Choose Processing/Realtime Processing Settings
.
The FFT settings determine the results of the noise zone analysis, just as in the noise removal dialogue window when processing the noise zones. The other parameters have no influence upon the quality of the result.
Select a zone proceeding as usual : click and drag the pointer in the waveform zone.
Select Analysis/Create Noise Zone from Selection
. A noise zone will determine a noise fingeprint, which is located in the center of the zone, as shown by the red dotted line.
Once a noise zone has been defined, it is represented in the sonogram window, and in the processing sequencer, as a block, just like a treatment.
Although the noise zone is represented in the processing sequencer, its size or location cannot be modified. This restriction is designed to limitate possible confusions between the position of the noise zone at the analysis on one hand, and at the noise removal, on the other hand. The noise zone wil be applied to the whole file or selected part or the file when processing the noise removal.
As shown in the inspector, the beginning, ending or duration values of the noise zone can only be displayed, but they cannot be edited.
A noise zone can be deleted, copied or cut and pasted from a location to another in the same sound file, or from a sound file to another.
Use the usual menus and shortcuts, and click at a desired location in the waveform window to paste a noise zone
Note that several noise zones can be generated or pasted at the same place, which might lead to a crash.
Noise zones can be exported as -noisezone.sdif files one by one to any location of the computer.
Select the noise zone you wish to export
Choose Analysis/Export Selected Noise Zone
Select a name and location for the file.
To import a noise zone into a file, no noise zone must have been created before. This means that only one noise zone can be imported to a file.
Select the region where you want to have the noise zone applied with a click and drag in the waveform area.
Choose Analysis/Import Noise Zone
.
Select the file you want to import.
This feature is designed to interpolate only one noise fingerprint with the other fingerprints of the file.
Noise zones can be saved as treatments. This can be an alternative to the fact the noise zones importations are limited.
Select File/Save Treatment as...
By default, the noise zone will be saved as a -trt.sdif file in the Treatments
directory of the current User Home folder.
To open the noise zone again in the same file, or in another file
select a region to apply the noise zone – if not, it will be applied to the whole file
choose File/Open Treatment
.
Note that the location of the noise zone will be similar to that of the saved treatment, and cannot be modified.
The interpolation is possible only if the fingerprints have been calculated with the same FFT settings. The fingerprints generated in the file are updated automatcially at the nolise removal, and can thus necessarilly be interpolated.
As the parameters of the imported fingerprints cannot be modified, it may happen that some imported items cannot be interpolated. This is why the number of possible importations is restricted.