geopsy.org

Marc,

I found a bug in H/V toolbox:
By default, the input parameters for H/V processing are set to the values used for the last H/V computation. When the option "At least" (for windowing) had been turn on , then when the H/V toolbox is launch, this option is still turn on, but only the minimum value of the time duration is displayed.
If I re-select the option "At least", then, this time, both minimum and maximum time durations are displayed.

This bug should not be too drastic to fix!

Sylvette

I wondered if this option was still necessary. The "At least" was first introduced by Gaillot and Bertrand in their original code in Fortran. At that time the fft computation was limited to power of 2 number of samples. Rather than systematically adding zeros at the end of a fixed length time window, the time windows themselves were enlarged to reach a power of 2 samples.

I think this option should be removed soon. What's your opinion?

This option is very useful in case of signals strongly dominated by transients. In such cases, the number of selected windows with sufficient time duration can be rather low. This option may thus enlarge the proportion of signal used for the processing. Of course, the best solution in such cases should be to perform new measurements in less noisy environment. It is however not always possible!

I agree with you, this option is not necessary in most cases. However I have some example which show that H/V results can be put in the garbage if I do not use this option ...

Ok, I'll fix the problem you reported in your first message and I'll keep this option available.