Changes to Simrad Ex60 sample range calculation in Echoview

Echoview 11 and later

The SimradEx60 TvgRangeCorrection equation is changed to better suit resampled Simrad Ex60 data.

Echoview 3.30 to Echoview 10

Ex60 sample data was fully supported first in Echoview 2.20. Versions of Echoview up to 3.25 interpreted sample range differently to later versions of Echoview. The first sample was considered to be one sample thickness distant from the transducer. Simrad informed us however that the first sample is at the transducer face, i.e. at a range of 0m and consequently one sample thickness closer to the transducer than was implemented in Echoview 3.25 and earlier.

As a consequence of this when working with Echoview 3.30 or later using any EV files created by 3.25 or earlier you must be careful to consider the impact of this change upon the data and your analyses. The following sections will help you identify if you are affected and what the consequences are:

• Does this affect my data?
• What are the consequences of this change?
• How far has my data moved (how big is one sample thickness)?
• How does this affect sample values?
• How does this affect my analyses?
• I'm using data compression in my .ek6 files, how does this affect me?

Does this affect my data?

This affects all variables derived from Ex60 sample data which were scrutinized with Echoview 3.25 or earlier and are now being scrutinized using Echoview 3.30 or later.

Specifically, that means the following raw variables and all virtual variables which derive from them:

Variables derived from Ex60 .raw files or Echolog 60 .ek6 files:

1. • Sv raw pings Tt

   • TS raw pings Tt

   • angular position raw pings Tt

Variables derived from Ex60 .DG files or Echolog 500 .ek5 files:

1. • angular position Bt telegrams Tt

   • Sv Wt telegrams (power) Tt

   • Sv Vt telegrams Tt

   • TS Wt telegrams (power) Tt

   • TS Pt telegrams Tt

No other data is affected.

What are the consequences of this change?

Put simply, every sample in each of the affected variables has moved closer to the transducer by one sample thickness and the very first sample (the sample closest to the transducer) is discarded.

The effects of this change can be summarized by:

• How far has my data moved (how big is one sample thickness)?
• How does this affect TVG range correction?
• How does this affect sample values?
• How does this affect my analyses?

How far has my data moved (how big is one sample thickness)?

Affected data is now closer to the transducer by one sample thickness. The sample thickness is typically in the order 0.1 to 10 centimeters but can be determined as follows.

If the data is from a .dg or .ek5 file this can be expressed as:

Sample thickness = Pulse duration × Sound speed / 8000

where pulse duration (ms) and sound speed (m/s) are taken from the Calibration page of the Variable Properties dialog box.

If the data is from a .raw, .ek60 or .ek6 file it can be expressed as:

Sample thickness = Sample interval × Sound speed/2

where Sample interval (s) is read from the data file (it can be seen on the Details dialog box) and sound speed (m/s) is taken from the Calibration page of the Variable Properties dialog box.

In either case, it can equally well be calculated from the start and stop range and sample count which are displayed on the Details dialog box as follows:

Sample thickness = (Stop range - Start range) / Number of data points

How does this affect TVG range correction?

Echoview 11 and later

SimradEx60 TvgRangeCorrection uses a Standard expression which is better suited to resampled Simrad Ex60 data than the 2 samples assumption used under the BySamples expression.

Echoview 3.30 to Echoview 10

SimradEx60 TvgRangeCorrection used a BySamples expression.

Application	Recommended value of TVG range correction
Sv data for echo integration	2
TS data for use with method 1 single target detection operators	2
TS data for use with method 2 single target detection operators	0

In Echoview 3.25 we recommend: three times the sample thickness for Sv data and one sample thickness for TS data to account for the error in Echoview's range calculation. If you applied those values you should now revise them to the current recommendations.

These recommendations reflect discussions between Echoview support and Simrad. Please contact Echoview support for further background on these recommendations if desired.

How does this affect sample values?

Data values for following variables will be unaffected:

• Sv Vt telegrams Tt

• TS Pt telegrams Tt

• angular position Bt telegrams Tt

• angular position raw pings Tt

Echoview 11 and later

SimradEx60 TvgRangeCorrection uses an equation which is better suited to resampled Simrad Ex60 data. The difference with respect to Echoview 10 for Sv data is of the order of 1E-7 dB. TS data is unchanged.

Echoview 3.30 to Echoview 10

Data values for following variables will be affected:

• Sv raw pings Tt

• TS raw pings Tt

• Sv Wt telegrams (power) Tt

• TS Wt telegrams (power) Tt

The magnitude of this change can be estimated as follows:

ΔSv = 20 log (R) + 2αR - 20 log(R+T) - 2α(R+T) = -(2αT + 20 log(1+T/R))
ΔTS = 40 log (R) + 2αR - 40 log(R+T) - 2α(R+T) = -(2αT + 40 log(1+T/R))

Where:

R = the range of the sample in Echoview 3.30 (m)
T = the sample thickness - can be calculated as described above (m)
α = the absorption coefficient (as entered on the Calibration page of the Variable Properties dialog box)

You can use these relations to estimate for your particular combination of pulse duration, absorption coefficient and operating range, the magnitude of the effect. For example: T=0.2m, α=0.01dB/m then ΔSv=-0.02dB and ΔTS=-0.04 at R=100m and ΔSv=-0.17dB and ΔTS= -0.34dB at R=10m.

How does this affect my analyses?

Echoview 11 and later

SimradEx60 TvgRangeCorrection uses an equation which is better suited to resampled Simrad Ex60 data. Refer to How does this affect my sample values?: Post Echoview 11

Echoview 3.30 to Echoview 10

All analyses may be affected.

If you have used lines or regions to define your analysis domain(s), then it is important to note that your data may have moved relative to these lines and regions. As a direct consequence of this, your analysis domains may now include some data which was intended for exclusion, or vice versa.

This may, or may not be of any consequence, depending entirely upon how and for what purpose you have defined your exclusion lines and analysis and bad data regions. The most significant effect for most users will be the possible inclusion of bottom echo trace in analyses.

You may be able to reposition all of your lines and regions as follows:

Repositioning regions: You can move the position of all regions by shifting your regions in the depth dimension by one sample thickness (as calculated above) using the Region Editor (note, if your transducer is not vertically oriented then you will have to convert one sample thickness in the range dimension to the equivalent in the depth dimension, see About depth, range and altitude).

Repositioning lines: You can move the position of each individual line by using the New line dialog box as described in Editing lines (you will need to move the lines closer to the transducer by one sample thickness in the range dimension).

It may not be necessary or possible to reposition your lines and regions however. Use the following table as a guide:

If your EV file uses:	Then:
a single pulse duration	All your data has moved by the same amount. You can reposition your lines and regions as needed and described above and continue working with this EV file. Alternately, you can define a transducer draft of one sample thickness in order to move the data relative to the lines and regions (essentially back to where it was before this change). The data values will still have changed (See How does this affect sample values) but they will have been repositioned in space to the same place relative to the lines and regions as they were prior to the change - thus ensuring that the inclusions and exclusions remain as they were.
multiple pulse durations	Each variable in Echoview has one pulse duration associated with it (on the Calibration page of the Variable Properties dialog box). This determines how far the data has moved towards the transducer (see How far has my data moved?). Lines and regions in Echoview are not specific to a particular variable and can be used on many variables. As a consequence, the data on variables of different pulse durations will have moved a different amount relative to each line and region. You will need to consider the impact of this upon your EV file. Take note that: The largest value of pulse duration will determine the largest move in data. So if you move your exclude-below line by this distance it will conservatively ensure that no samples on any variable have moved above it and are now included in your analysis. It is possible to move the data back to where it was prior to this change. To do this, define a transducer with a draft of one sample thickness for each unique sample thickness and associate the respective variables with the appropriate transducer. The data values will still have changed (See How does this affect sample values) but they will have been repositioned in space to the same place relative to the lines and regions as they were prior to the change - thus ensuring that the inclusions and exclusions remain as they were. Although the data in each variable will move by a different amount, this is in fact bringing the position of this data in line with Simrad's specification. This suggests that the bottom echo trace between the different echograms will be in better agreement now than they were.

Compression in .ek6 Files

The compression techniques used by Echolog 60 include the possible specification of:

• a start and stop range for data logging
  The logged samples will all have moved one sample thickness closer to the transducer. No other effect

• averaging of below threshold data
  Because all logged samples have moved one sample thickness closer to the transducer, they receive less TVG and their values have decreased. Consequently, samples that should not have been averaged were not, and some which should have been averaged were not. You have not lost any data in this process.

• averaging of below sounder-detected-bottom data
  Because the data samples have moved closer to the transducer by one sample thickness, some samples which were averaged because they were below the sounder-detected bottom line, should not have been averaged. This will affect at most one sample per ping, and it will now be above the sounder-detected bottom line but will reflect an averaged value.

Note: Echolog 60 now selects data correctly for compression in this manner. To help distinguish this fact it now logs to .ek60 files (not .ek6 files).

See also

Simrad Ex60 Sample Range Calculation
Simrad Ex500 Sample Range Calculation