GPR data in SEGY format: Z factor adjustment

324 views
Skip to first unread message

NBK

unread,
Aug 23, 2016, 4:25:00 AM8/23/16
to OpendTect Users
Hi all,
I am working with a GPR dataset in SEGY format and trying to import it to OpendTect. When I import my data, I cannot set the appropriate z-range in the survey setup, because I am working on very high frequency radar, i.e., I only have some 10s of nanoseconds of data. This means in the visualisation my data are not displayed correctly (no depth).
From what I understood in the manual, it is possible to use a Z factor adjustment when working with GPR data, so that the data are scaled. However, it seems to be possible only for data in .DZT format. Unfortunately, I do not have the option to convert my data to .DZT format (if I do, I lose the geographical information), so my question is: Is there a way to use this Z factor adjustment when importing SEGY data? I do not see it anywhere, but I might have missed it somewhere.
Best,
N

Maurizio Ercoli

unread,
Aug 23, 2016, 5:34:53 AM8/23/16
to us...@opendtect.org
Hello Nabjka,

it doesn' t matter which time window you are using.
You can simply use the Z-scale tool (in menu "View") to adjust and stretch the Z, also after importing a (GPR) SEG-Y (you don't need a conversion in .DTZ format).

Immagine incorporata 1

Best r.

maurizio.



--
You received this message because you are subscribed to the Google Groups "OpendTect Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to users+unsubscribe@opendtect.org.
To post to this group, send email to us...@opendtect.org.
Visit this group at https://groups.google.com/a/opendtect.org/group/users/.
To view this discussion on the web visit https://groups.google.com/a/opendtect.org/d/msgid/users/8f308f7e-1e29-4966-a8c1-8de5fd4a754c%40opendtect.org.
For more options, visit https://groups.google.com/a/opendtect.org/d/optout.



--

Maurizio Ercoli, Ph. D
DIPARTIMENTO DI FISICA E GEOLOGIA
Università degli Studi di Perugia
06123 Perugia PG, Italia

Julien Moreau

unread,
Aug 23, 2016, 5:34:53 AM8/23/16
to us...@opendtect.org
Good morning,

When importing your data, overide the sampling rate. A tip is to use a conversion factor to transform directly into depth so that the offsets and the depth are in the same unit assuming a constant velocity

Cheers,

Julien

--

Bert Bril

unread,
Aug 23, 2016, 5:34:53 AM8/23/16
to us...@opendtect.org
Hi N,


> Is there a way to use this Z factor adjustment when importing SEGY
> data? I do not see it anywhere, but I might have missed it somewhere.

According to the hacker's rules, I should now answer 'Yes'.

Still. In the wizard, there is a column where you can override the
values from the SEG-Y file. It has a Z start/step, so you can basically
give your data any range you want. I'd adjust the Z scale to a multiple
of 10 of the original, so you can recognize the scale. The best thing to
do is to make stuff comparable to seismic data set ranges (i.e. in the
order or 1000s of milliseconds).


Cheers, Bert


--
OpendTect: selected recipient of SEG's
<http://dgbes.com/index.php/key-words/news/winner-of-seg-s-distinguished-achievement-award>
Distinguished Achievement Award 2016
<http://dgbes.com/index.php/key-words/news/winner-of-seg-s-distinguished-achievement-award>

NBK

unread,
Aug 23, 2016, 9:31:40 AM8/23/16
to OpendTect Users
Hi all,
Thanks for your replies. Ideally, I was looking for a solution where the "wizard" automatically finds the correct survey settings etc., because I want to use the software for teaching purposes. I am aware of the z-scale in the View menu, but if the data is very shallow (10s of ns) then a simple stretching of the z-scale makes it very coarse. Ideally I want my students to use the 2D viewer to check the depth of some horizons, but that is not possible if the data appear pixelated, so to speak. "Pretending" that the data is in depth (or using a conversion factor) also means that the students have to do a conversion back again... I am not sure I trust them to be capable of that, but that is of course an option.
For those with similar problems: In the end, I found that the easiest (if somewhat convoluted) solution was to change my SegY files so that the z-range now (for the purposes of OpendTect) appears to be in ms rather than ns.
Thanks for your help everyone. It is nice with a supportive user community.
Best
N
Reply all
Reply to author
Forward
0 new messages