[Seisunix] wrong to read segy file
IVAN DE LA CRUZ VARGAS CORDERO
segyread
tape=/data2/ALFILBLAU/RAW_DATA/Guatemala/ig2404/DBseis/ar06.0279.ig2404.gua-15.none.segy
|suximage
and indicate me:
1+0 records in
6+1 records out
segyread: discrepant tr.ns = 1434 with tape/user ns = 3000
... first noted on trace 401
segyread: discrepancy found in header and trace ns values
the value (3000) was used to extract traces
Could somebody help me to solve this problem, in fact if I try read
the first 401 traces is ok..After is not posible read them.
Regards Ivan Vargas Cordero
_______________________________________________
seisunix mailing list
seis...@mailman.mines.edu
https://mailman.mines.edu/mailman/listinfo/seisunix
Unsubscribe: seisunix-u...@mailman.mines.edu
Gery .
I've seen this problem before, however not sure what is the real problem. Anyway, I circumvet it with:
segyread tape=segy conv=0 | segyclean > su
Try conv=1 anyway.
Hope this helps,
> From: iva...@ub.edu
> To: seis...@mailman.mines.edu
> Subject: [Seisunix] wrong to read segy file
IVAN DE LA CRUZ VARGAS CORDERO
is the same..
"Gery ." <game...@hotmail.com> ha escrit:
Gery .
I forgot to mention it, but I got the same error and could read the file though. What is the error message you get after "suximage<su"?
> From: iva...@ub.edu
> To: game...@hotmail.com
> CC: seis...@mailman.mines.edu
> Subject: RE: [Seisunix] wrong to read segy file
John Stockwell
> Hello, I have a problem to read a segy file I 'm running the following script:
> segyread
> tape=/data2/ALFILBLAU/RAW_DATA/Guatemala/ig2404/DBseis/ar06.0279.ig2404.gua-15.none.segy
> |suximage
> and indicate me:
> 1+0 records in
> 6+1 records out
I would suggest a couple of changes. First
segyread
tape=/data2/ALFILBLAU/RAW_DATA/Guatemala/ig2404/DBseis/ar06.0279.ig2404.gua-15.none.segy verbose=1 | segyclean > outfile.su
The segyclean is needed to zero out the optional field. If there are values
in the optional trace header field, this will cause problems with
SU graphics programs.
Likely either the headers are wrong on the data. So I would try
segyread
tape=/data2/ALFILBLAU/RAW_DATA/Guatemala/ig2404/DBseis/ar06.0279.ig2404.gua-15.none.segy verbose=1 ns=3000 | segyclean > outfile.su
Then do
segyread < outfile.su
so you can see what the trace header values look like. If there are
actually a different number of samples per trace after trace 400, then
there will be weird large numbers in all of the header fields. If there
are only a select number of header fields set, then it is the ns value
in the headers that are wrong after trace 400.
>
> segyread: discrepant tr.ns = 1434 with tape/user ns = 3000
> ... first noted on trace 401
>
> segyread: discrepancy found in header and trace ns values
> the value (3000) was used to extract traces
> Could somebody help me to solve this problem, in fact if I try read the first
> 401 traces is ok..After is not posible read them.
>
> Regards Ivan Vargas Cordero
>
>
> _______________________________________________
> seisunix mailing list
> seis...@mailman.mines.edu
> https://mailman.mines.edu/mailman/listinfo/seisunix
> Unsubscribe: seisunix-u...@mailman.mines.edu
John Stockwell | jo...@dix.Mines.EDU
Center for Wave Phenomena (The Home of Seismic Un*x)
Colorado School of Mines
Golden, CO 80401 | http://www.cwp.mines.edu/cwpcodes
voice: (303) 273-3049
Our book:
Norman Bleistein, Jack K. Cohen, John W. Stockwell Jr., [2001],
Mathematics of multidimensional seismic imaging, migration, and inversion,
(Interdisciplinary Applied Mathematics, V. 13.), Springer-Verlag, New York.
John Stockwell
Gery .
Try sushift
From: iva...@ub.edu
To: jo...@dix.mines.edu
CC: seis...@mailman.mines.edu
Subject: Re: [Seisunix] wrong to read segy file
Ok I can read using trcwt=0 the result I enclose in this image..how
Could I to solve the variable lenght of traces?
John Stockwell <jo...@dix.mines.edu> ha escrit:
Fernando M. Roxo da Motta
<iva...@ub.edu> wrote:
Hi all,
> I have tried as suggested by Gery using conv=1 and conv=0, but error
> is the same..
>
Just some brainstorming...
Sometimes we should try to get another opinion in order to make
things a little bit clear.... Or not. :)
Let me suggest another approach, there is another interesting project
of seismic data processing :
SeaSeis - A simple seismic processing system
http://seaseis.com/
It is free software as well and is just starting. I am not
suggesting you to change packages. This project has a very
interesting tool that is the "seaview", a display program that reads
Seg-Y, Seg-D (some of it), SU, its internal format and some ASCII
format. What is interesting in this program is that it allow you
to investigate the values of all headers fields in a Seg-Y and even
trace values. I solved many problems using this tool.
My suggestion is to install the package, or at least the "seaview",
to investigate your file in order to get sure that it has different
trace length and not a difference in interpretation of header values
(conv=1 vs conv=0, endian=1 vs endian=0).
Another point is how this file was created ? Is it a concatenation
of many files ? If so, how was it done ? Just a 'cat' command ? If
people just used 'cat' to concatenate Seg-Y files there will be strange
consequences, as there will be a textual (3200 EBCDIC characters) and a
binary (400 bytes) header in the middle of the data. I don't know any
conversion program that can work this out. If (say) three files are
concatenated this way you will have three different blocks of data and
an strange behaviour at each included header (3600 bytes).
If the data really has a variable number of samples per trace, the
plot suggests that it happens in blocks, and that was the base for the
question about how the file was created. If it has variable trace
length and by blocks, you can ever use 'dd' command to get different
chunks of data with constant number of samples per trace. Those blocks
can be windowed (suwind) to an equal number of samples per trace and
then concatenated, as usual with SU data.
HTH
Roxo
--
---------------- Non luctari, ludare -------------------+ WYSIWYG
Editor ? Fernando M. Roxo da Motta <pe...@roxo.org> |
VI !! Except where explicitly stated I speak on my own behalf.| I see
text, ( Usuário Linux registrado #39505 ) | I get text !
Gery .
Thanks Fernando! this software looks very nice to me, and the introduction paragraph very honest and funny,
Cheers,
Gery
> From: pe...@roxo.org
> To: seis...@mailman.mines.edu
> Subject: Re: [Seisunix] wrong to read segy file
Fernando M. Roxo da Motta
<pe...@roxo.org> wrote:
Hi all,
Ivan sent me the data and I wrote a quick (sort of) Fortran 90
program to sweep the file. The program is attached in this message.
The program output was :
./a.out
DT= 4000
NS= 3000
FORMAT= 3
Byte per sample= 2
Number of samples changed from 3000 to 1434 at trace 401
Number of samples changed from 1434 to 3000 at trace 426
Number of samples changed from 3000 to 2331 at trace 2701
Number of samples changed from 2331 to 3000 at trace 2726
Number of samples changed from 3000 to 123 at trace 5426
EOF reading trace header at trace 5451
STOP 9999
So there are three sets with short traces :
from trace to trace Samples
401 425 1434
2701 2725 2331
5426 5426 123
So far it is unsupported by SU. (Just wondering, would it be too
hard to change ?....)
To make short a long history, I am sending attached a script that
divided the SegY file in three chunks with the same numbe of samples per
trace and concatenated them, producing an useful SU data.
$ surange < output.su
5375 traces:
tracl 1 5425 (1 - 5425)
tracr 1 5425 (1 - 5425)
fldr 3 222 (3 - 222)
tracf 1 25 (1 - 25)
cdp 3 222 (3 - 222)
trid 1
ns 3000
dt 4000
HTH
IVAN DE LA CRUZ VARGAS CORDERO
arief....@gmail.com
i also get some problem when loading segy data to seaseis,
error :
========================================
Run init phase...
Run INIT phase for module #1 INPUT_SEGY...
Input file # 0: Line_001.segy
Input file # 1: /home/---/su/openseaseis_v2.03/flows/Line_001.segy
Module #1 INPUT_SEGY:
FATAL ERROR occurred in INIT phase.
Could not open file '1'
========================================
Arief Kusman
Geophysicist