Hi Marc,
in the version of Geopsy 3.4 there is a new plug in: Phase lt, but no information on its meaning and use.
Would it be possible to have a brief explanation of the new tool?
Thank you
Luigi
trying this version i verified that spac2disp doesn't work.
a new plug in in the 3.4 version
Re: a new plug in in the 3.4 version
Hi Luigi,
PhaseIt is a tool for comparing signals of Huddle Tests (all sensor at the same place, max 1-2 m), below some frequency the response of all sensors should be the same. This tools computes the coherence, the phase shifts and the spectral ratios versus time and frequency. I hope to submit a short paper on it soon.
Can you detail what's wrong with spac2disp? I checked quickly with vertical data and I noticed no problem. It works also with 3C data.
Marc
PhaseIt is a tool for comparing signals of Huddle Tests (all sensor at the same place, max 1-2 m), below some frequency the response of all sensors should be the same. This tools computes the coherence, the phase shifts and the spectral ratios versus time and frequency. I hope to submit a short paper on it soon.
Can you detail what's wrong with spac2disp? I checked quickly with vertical data and I noticed no problem. It works also with 3C data.
Marc
Re: a new plug in in the 3.4 version
Hi Marc,
it just doesn't start. When in geopsyland I type spac2disp nothing happens and the same happens if I click directly from outside of geopsyland on spac2disp.
I also tried the new Frequency Rejection utility. I also read the paper and I noticed that the parameter n (or v) is crucial to significantly modify the morphology of the HV curve. I am attaching three pics where the raw HV curve (geometric mean) is represented and the result of the Frequency Rejection with SD 2 and 1.5 where the peak at 0.6Hz gains in amplitude exceeding the value 2 and satisfy the SESAME criterions for natural HV peak.
Which criterion should be used for this parameter? Can the result I obtained with SD1.5 be considered correct?
And finally, when is it preferable to adopt the geometric average rather than the squared average?
Thank you
Luigi
it just doesn't start. When in geopsyland I type spac2disp nothing happens and the same happens if I click directly from outside of geopsyland on spac2disp.
I also tried the new Frequency Rejection utility. I also read the paper and I noticed that the parameter n (or v) is crucial to significantly modify the morphology of the HV curve. I am attaching three pics where the raw HV curve (geometric mean) is represented and the result of the Frequency Rejection with SD 2 and 1.5 where the peak at 0.6Hz gains in amplitude exceeding the value 2 and satisfy the SESAME criterions for natural HV peak.
Which criterion should be used for this parameter? Can the result I obtained with SD1.5 be considered correct?
And finally, when is it preferable to adopt the geometric average rather than the squared average?
Thank you
Luigi
- Attachments
-
- spac2disp.jpg
- (101.58 KiB) Not downloaded yet
-
- SD_1.5.jpg
- (69.65 KiB) Not downloaded yet
-
- SD_2.jpg
- (80.22 KiB) Not downloaded yet
-
- geom_mean.jpg
- (82.19 KiB) Not downloaded yet
Last edited by luigiV on Thu May 06, 2021 10:03 am, edited 3 times in total.
Re: a new plug in in the 3.4 version
spac2disp is now fixed in 3.4.2-preview (git repo).
The geometric average suggests that the E and N are two quantities that are better represented on a log scale. However, I see E and N as two components of a vectorial field, hence the combination through "Total horizontal energy" sounds better. The "Square average" is the same with a factor sqrt(2). I would say that for historical reasons the last one is mainly used.
I also noticed that a factor 2 is not removing many time windows. 1.5 is more effective. Is it better? The uncertainty is reduced but is it more accurate? I have no answer.
The geometric average suggests that the E and N are two quantities that are better represented on a log scale. However, I see E and N as two components of a vectorial field, hence the combination through "Total horizontal energy" sounds better. The "Square average" is the same with a factor sqrt(2). I would say that for historical reasons the last one is mainly used.
I also noticed that a factor 2 is not removing many time windows. 1.5 is more effective. Is it better? The uncertainty is reduced but is it more accurate? I have no answer.
Re: a new plug in in the 3.4 version
Thanks,
so this approach to window selection is interesting, but not conclusive.
Finally I found that gphistogram opens the max files of HVTFA, but it represents them in an incomprehensible way. I am attaching a picture.
Luigi
so this approach to window selection is interesting, but not conclusive.
Finally I found that gphistogram opens the max files of HVTFA, but it represents them in an incomprehensible way. I am attaching a picture.
Luigi
- Attachments
-
- max_ell_gphistogram.jpg
- (56.73 KiB) Not downloaded yet
Re: a new plug in in the 3.4 version
Yes, there is still some work to really read those .max files into gphistogram.