© ElektroPhysik
QuintSonic 7
84 von 184
Unknown sound velocities represent another issue. Though the measuring results will provide the
correct number of layers (on condition any other problems have not been encountered), you
cannot be sure about the correctness of layer thickness unless you know the sound velocities in
the materials concerned.
For that reason, it is always recommended to make a material cross-section. Based on the results
of such cross section, you can interpret readings correctly and set parameters accordingly. This is
the only way to obtain reliable and sound readings.
9.9 Interaction of preprocessor, optimisation and
postprocessor parameters
This section is dedicated to experienced users wishing to obtain maximum system performance by
using optimal parameter sets. A good understanding of the system functions is assumed.
The processing sequence
– preprocessor, optimisation, postprocessor is of great significance for
the determination of the optimal parameter set. The input data for the optimisation stage represent
the output data of the preprocessor stage. In turn, the output data of the optimisation stage
represent the input data for the postprocessor stage.
Using input data such as position and amplitudes of the extracted echoes, the optimisation
process creates a synthetic receive signal. After this, the process tries in small steps to change the
positions and amplitudes of the extracted echoes in a way that the respective, newly calculated
synthetic receive signal deviate as little as possible from the true receive signal. This process is
verified by using the least squares method.
The optimisation will succeed more effectively to the extent that the input signal of the optimisation
stage is complete. If bigger portions of the input signal are missing (e. g. through a preceding
clipping with amplitudes limits adjusted very high), the optimisation processor might try to close this
“gap” by displacing and enlarging extracted echoes located in the proximity. This will unnecessarily
deteriorate the accuracy of the neighbouring echoes. If the optimisation function is enabled, it is
recommended not to suppress strong interference echoes by means of the preprocessor. They
should rather be left in the signal for further use of such signal portions through the optimisation
function. The interference echo portions can be suppressed later on by the postprocessor.
The postprocessor, with its various stages, represents the last processing level before calculation
of layer thickness. The output signal of the postprocessor has no effect on the previous stage
(preprocesor and optimisation). The processing steps of the postprocessor will only have an effect
on the final result. For that reason, it is of no significance to the optimisation, how the
postprocessor has been set.
On the other hand, the sequence of postprocessing stages is important. Please see section 8.5. An
echo can be marked as being invalid only once. Once an echo has been identified as being invalid
in a certain stage, it will be marked by an icon of this stage accordingly. As a consequence, it will
not be considered any more in any further subsequent processing stage. If you find an echo
marked other than expected, it is recommended to check at which stage it has actually been