Background noise in Echoview

This page outlines several methods with which you can address the issue of background noise in your acoustic data using Echoview.

Data thresholding
- Single beam data
- Multibeam data
Background noise subtraction
- On single beam echograms
- On multibeam echograms

Data thresholding

Where your targets have good signal to noise ratio (the background noise is low level when compared to the signal you are interested in) then it may be sufficient to ignore all data below a certain threshold.

Single beam data

Four thresholds can be applied to single beam Sv data in Echoview (handling is similar for other data types such as TS data):

on the Data page of the Variable Properties dialog box
- Minimum Sv threshold
- Maximum Sv threshold
- Time-varied threshold (TVT)
- Minimum TS threshold

Notes:

The time-varied threshold follows the time-varied gain curve and is defined by an Sv value at 1 m.
The time-varied threshold is applied to the data. There is no way to apply a time-varied threshold to the display without applying it to the data.
From Echoview 4.20 onwards, if a TVG range correction is applied on the Calibration 2 page (from Echoview 5.0 aka Calibration page) of the Variable Properties dialog box, then this correction is also used in the TVT calculations. Previously, Echoview did not account for TVG range correction in such calculations. As a consequence there may be small but significant errors in the TVT data. The errors are of the order of 0.5dB (at 2m) to 0.05 dB (at 17m).
Press CTRL+F11 or CTRL+F12 keys to change the Sv value at 1m to observe the effect on the data in the echogram window.
The Minimum TS threshold on the Data page of the Variable Properties dialog box of an Sv variable is a theshold which removes samples below a specified TS value. Refer to the Minimum TS threshold algorithm page for more information.

Multibeam data

Three thresholds can be applied to multibeam Sv data in Echoview:

on the Data page of the Variable Properties dialog box
- Minimum Sv threshold
- Maximum Sv threshold
using virtual variables
- a double threshold using the Double threshold operator

Background noise subtraction

Where there is a target with a low signal to noise ratio it may not be appropriate to correct for background noise by thresholding, an alternative (and physically correct approach) is to correct for noise by subtraction. The papers cited in the References page of this help file may provide useful reading on this topic.

Note: If correcting for noise by subtraction then NO thresholds should first be applied to the data.

On single beam echograms

Using analysis variables

In this case mean Sv and mean noise Sv (or equivalent area variables) are calculated over integration cells and the mean noise is subtracted from the total of noise plus target outside of Echoview as follows:

select the desired background noise analysis variables (on the Export page of the EV file properties dialog)
specify a noise Sv at 1 m (on the Analysis page of the Variable Properties dialog box)
specify the absorption coefficient (on the Calibration page of the Variable Properties dialog box)
specify the TVG range correction setting (on the Calibration page of the Variable Properties dialog box)
Export integration results (see Exporting integrations)
The noise analysis variables are exported along with integration results. You must perform the subtraction external of Echoview with whatever tool you are most comfortable with for downstream analysis.

Note: From Echoview 4.20 onwards, if a TVG range correction is applied on the Calibration 2 page (from Echoview 5.0 aka Calibration page) of the Variable Properties dialog box, then this correction is also used in the background noise analysis variable calculations. Previously, Echoview did not account for TVG range correction in such calculations. As a consequence there may be small but significant errors in the TVT data. The errors are of the order of 0.5dB (at 2m) to 0.05 dB (at 17m).

Using virtual variables

Virtual variables can also be used for background noise correction as follows:

create an new (noise) variable which estimates the background noise (use the data generator operator)
subtract this noise variable from the data you intend to analyze (use the linear minus operator)
continue your analysis as usual

For more information see About virtual variables, operators, or contact Echoview support.

Background noise removal operator

A multi-threaded operator that removes background noise for single beam Sv variables. The Background Noise Removal algorithm assumes that some portion of the sampled acoustic signal is dominated by background noise. It estimates the background noise value for each ping and subtracts it from the ping's samples. The algorithm analyzes the echogram by averaging the sample values withinn averaging cells around each ping. The averaging cells have a fixed horizontal extent in pings and a fixed vertical extent in samples. The noise estimate for a ping is the minimum of its cell averages.

You can specify six settings on the Background Noise removal page of the Variable Properties dialog box. Four settings specify the averaging cell used to determine the noise estimate. Two Threshold settings set the Maximum Noise and Minimum SNR.

The algorithm uses concepts discussed in "A post-processing technique to estimate the signal-to-noise ratio and remove echosounder background noise." by A. De Robertis and I. Higginbottom (2007). The algorithm page discusses background noise removal stategies that may be helpful with low SNR data.

On multibeam echograms

Two techniques using virtual variables are available for removing background noise and stationary objects in multibeam data.

To remove background noise or objects using a running mean estimate:

create an new (noise) variable which estimates the background noise (use the mean of n previous pings operator)
subtract this noise variable from the data you intend to analyze (use the minus operator)
continue your analysis as usual

To remove background noise or objects using a selective mean estimate:

create an new (background) variable which contains only pings which you have identified as containing only background noise or objects (use the ping subset operator)
subtract the average of all those pings from the data you intend to analyze (use the sample statistic subtract operator)
continue your analysis as usual