From rm438 at cam.ac.uk Wed May 13 13:04:40 2009 From: rm438 at cam.ac.uk (Bob Myhill) Date: Wed, 13 May 2009 21:04:40 +0100 Subject: [SAC-HELP] BBFK alphanumeric output from SAC Message-ID: Dear all, I've been trying for some time to plot bbfk data retrieved from SAC (the 09/04/2008 version) in .a format, but do not understand the structure of the data output. The SAC manual informs me that the data is output in 'square mode', but I don't know how this relates to the azimuth and wavenumber, and how consecutive data points are chosen. I've tried both a simple azimuthal and wavenumber progression (rotating around the z-axis or radiating from it) but neither of these seem to recreate SACs own output. Could you give me some advice on how to read this data? Many thanks, Bob -------------- next part -------------- An HTML attachment was scrubbed... URL: From george at gly.bris.ac.uk Thu May 14 02:13:10 2009 From: george at gly.bris.ac.uk (George Helffrich) Date: Thu, 14 May 2009 10:13:10 +0100 Subject: [SAC-HELP] BBFK alphanumeric output from SAC In-Reply-To: References: Message-ID: <786773713a027ce5bf3f969d25c91c42@gly.bris.ac.uk> Dear All - This is one of the darker areas of SAC (in the sense of documentation). The reference cited in the code for the method is this: Nawab, SH, FU Dowla, and RT Lacoss, Direction determination of wideband signals, IEEE Trans. Acous. Speech Sig. Proc., 33: (5), 1114-1122, 1985. The WAVENUMBER parameter is a value in reciprocal kilometers, and is the maximum value that defines the extent of the X-Y grid of returned values. Assuming familiarity with reading SAC XYZ files, the value at every spatial point should be the F-K value, transformed by the method chosen: MLM, Music or PDS. MLM and Music will transform the F-K value to its reciprocal on output. PDS does no transformation. On 13 May 2009, at 21:04, Bob Myhill wrote: > Dear all, > > I've been trying for some time to plot bbfk data retrieved from SAC > (the 09/04/2008 version) in .a format, but do not understand the > structure of the data output. > > The SAC manual informs me that the data is output in 'square mode', > but I don't know how this relates to the azimuth and wavenumber, and > how consecutive data points are chosen. I've tried both a simple > azimuthal and wavenumber progression (rotating around the z-axis or > radiating from it) but neither of these seem to recreate SACs own > output. > > Could you give me some advice on how to read this data? > > Many thanks, > Bob > > _______________________________________________ > sac-help mailing list > sac-help at iris.washington.edu > http://www.iris.washington.edu/mailman/listinfo/sac-help > George Helffrich george at geology.bristol.ac.uk From re.rezaei at gmail.com Sun May 17 08:28:05 2009 From: re.rezaei at gmail.com (reza rezaei) Date: Sun, 17 May 2009 19:58:05 +0430 Subject: [SAC-HELP] sac scalling Message-ID: <5feb258a0905170828j69dce239i67ec19838015431@mail.gmail.com> Hi every one As you know sac changes its scale when we use fft command in it and we should correct it with div command. (Divide into dt) I want to use transfer to and transfer from. (Convolution and deconvolution) I want to know does sac change its scale or not here. Where changing scale is necessary and I have to pay attention it? I would be please if you help me. Thank you, Reza Iranian seismological center From carltape at fas.harvard.edu Sun May 17 13:35:47 2009 From: carltape at fas.harvard.edu (Carl Tape) Date: Sun, 17 May 2009 16:35:47 -0400 Subject: [SAC-HELP] SAC on 64-bit Linux Message-ID: <1242592547.4a107523bea9f@webmail.fas.harvard.edu> 17-May-2009 Dear SAC Help: I am having trouble using SAC libraries for a particular Fortran code. The problem may be related to 64-bit vs 32-bit processing. Here are the details of my situation. uname -a Linux denali #42-Ubuntu SMP Fri Apr 17 01:58:03 UTC 2009 x86_64 GNU/Linux Computer Dell Inspiron 530 Platform Ubuntu 9.0.4 sac SEISMIC ANALYSIS CODE [09/04/2008 (Version 101.2)] gfortran --version GNU Fortran (Ubuntu 4.3.3-5ubuntu4) 4.3.3 gcc --version gcc (Ubuntu 4.3.3-5ubuntu4) 4.3.3 I see from this file http://www.iris.edu/software/sac/SAC_History.txt that SAC should be possible with 64 bit Linux. And I can at least run sac and look at sac files, so something is working. The particular Fortran code is the FLEXWIN package from CIG (I am one of the main developers of this code): http://www.geodynamics.org/cig/software/packages/seismo/flexwin I used this code throughout my thesis at Caltech. Now, at Harvard on my new setup, when I compile the code, the pertinent error message is as follows: gfortran -O2 -Wall flexwin.f90 -o flexwin copy_float.o libmeasure2.a -Lttimes_mod -L/opt/sac/lib -lsacio -lsac -ltau -lm /usr/bin/ld: skipping incompatible /opt/sac/lib/libsacio.a when searching for -lsacio /usr/bin/ld: cannot find -lsacio collect2: ld returned 1 exit status make: *** [flexwin] Error 1 Qinya Liu (U. Toronto) is able to successfully compile the program with Ubuntu but using a slightly older (32 bit?) computer and a slightly older Ubuntu version: Linux river #1 SMP Wed Apr 1 21:47:28 UTC 2009 i686 GNU/Linux This led us to believe that the problem was 32-bit vs 64-bit. Is anyone aware of this problem? Do I need to try to compile SAC from the source code on my 64-bit computer? I would appreciate any help on this matter. Thanks, Carl Tape Harvard U. ----------- From chad at iris.washington.edu Sun May 17 14:02:22 2009 From: chad at iris.washington.edu (Chad Trabant) Date: Sun, 17 May 2009 14:02:22 -0700 Subject: [SAC-HELP] SAC on 64-bit Linux In-Reply-To: <1242592547.4a107523bea9f@webmail.fas.harvard.edu> References: <1242592547.4a107523bea9f@webmail.fas.harvard.edu> Message-ID: <36382461-909E-4A9C-A57B-6E8859AB9AE2@iris.washington.edu> Hi Carl, The gfortran compiler should be able to build both 32 and 64-bit binaries. I'm guessing your compiler is defaulting to 64-bit and that's why it's skipping the 32-bit libsacio.a. Try adding the '-m32' option to the compiler arguments to build a 32-bit binary. Unless you need the benefits of 64-bit (possible but unlikely), a 32-bit binary will probably work just fine. It is not generally possible to compile programs with a mix of 32 and 64-bit libraries; for a given application the constituent binary objects (libraries, .o files, etc.) must all be one or the other, otherwise the linker will complain as it did in your case. On most modern 64-bit operating systems their are parallel sets of system libraries, one 32-bit and one 64-bit. In some cases you need to install the 32-bit system libraries. regards, Chad On May 17, 2009, at 1:35 PM, Carl Tape wrote: > 17-May-2009 > > Dear SAC Help: > > I am having trouble using SAC libraries for a particular Fortran > code. The > problem may be related to 64-bit vs 32-bit processing. Here are the > details of > my situation. > > uname -a > Linux denali #42-Ubuntu SMP Fri Apr 17 01:58:03 UTC 2009 x86_64 GNU/ > Linux > > Computer Dell Inspiron 530 > Platform Ubuntu 9.0.4 > > sac > SEISMIC ANALYSIS CODE [09/04/2008 (Version 101.2)] > > gfortran --version > GNU Fortran (Ubuntu 4.3.3-5ubuntu4) 4.3.3 > > gcc --version > gcc (Ubuntu 4.3.3-5ubuntu4) 4.3.3 > > I see from this file > http://www.iris.edu/software/sac/SAC_History.txt > that SAC should be possible with 64 bit Linux. > > And I can at least run sac and look at sac files, so something is > working. > > The particular Fortran code is the FLEXWIN package from CIG (I am > one of the > main developers of this code): > http://www.geodynamics.org/cig/software/packages/seismo/flexwin > > I used this code throughout my thesis at Caltech. Now, at Harvard > on my new > setup, when I compile the code, the pertinent error message is as > follows: > > gfortran -O2 -Wall flexwin.f90 -o flexwin copy_float.o libmeasure2.a > -Lttimes_mod -L/opt/sac/lib -lsacio -lsac -ltau -lm > /usr/bin/ld: skipping incompatible /opt/sac/lib/libsacio.a when > searching for > -lsacio > /usr/bin/ld: cannot find -lsacio > collect2: ld returned 1 exit status > make: *** [flexwin] Error 1 > > Qinya Liu (U. Toronto) is able to successfully compile the program > with Ubuntu > but using a slightly older (32 bit?) computer and a slightly older > Ubuntu > version: > > Linux river #1 SMP Wed Apr 1 21:47:28 UTC 2009 i686 GNU/Linux > > This led us to believe that the problem was 32-bit vs 64-bit. Is > anyone aware > of this problem? Do I need to try to compile SAC from the source > code on my > 64-bit computer? > > I would appreciate any help on this matter. > > Thanks, > Carl Tape > Harvard U. > ----------- > > _______________________________________________ > sac-help mailing list > sac-help at iris.washington.edu > http://www.iris.washington.edu/mailman/listinfo/sac-help > From savage at uri.edu Sun May 17 14:51:12 2009 From: savage at uri.edu (Brian Savage) Date: Sun, 17 May 2009 17:51:12 -0400 Subject: [SAC-HELP] SAC on 64-bit Linux In-Reply-To: <1242592547.4a107523bea9f@webmail.fas.harvard.edu> References: <1242592547.4a107523bea9f@webmail.fas.harvard.edu> Message-ID: Carl, In the next couple of weeks, I hope, a new version of SAC (101.3) should be available which will be able to run natively on 64 bit machines. This includes a 64 bit version of the sacio library and the sac executable. Cheers Brian Savage On May 17, 2009, at 4:35 PM , Carl Tape wrote: > 17-May-2009 > > Dear SAC Help: > > I am having trouble using SAC libraries for a particular Fortran > code. The > problem may be related to 64-bit vs 32-bit processing. Here are > the details of > my situation. > > uname -a > Linux denali #42-Ubuntu SMP Fri Apr 17 01:58:03 UTC 2009 x86_64 GNU/ > Linux > > Computer Dell Inspiron 530 > Platform Ubuntu 9.0.4 > > sac > SEISMIC ANALYSIS CODE [09/04/2008 (Version 101.2)] > > gfortran --version > GNU Fortran (Ubuntu 4.3.3-5ubuntu4) 4.3.3 > > gcc --version > gcc (Ubuntu 4.3.3-5ubuntu4) 4.3.3 > > I see from this file > http://www.iris.edu/software/sac/SAC_History.txt > that SAC should be possible with 64 bit Linux. > > And I can at least run sac and look at sac files, so something is > working. > > The particular Fortran code is the FLEXWIN package from CIG (I am > one of the > main developers of this code): > http://www.geodynamics.org/cig/software/packages/seismo/flexwin > > I used this code throughout my thesis at Caltech. Now, at Harvard > on my new > setup, when I compile the code, the pertinent error message is as > follows: > > gfortran -O2 -Wall flexwin.f90 -o flexwin copy_float.o libmeasure2.a > -Lttimes_mod -L/opt/sac/lib -lsacio -lsac -ltau -lm > /usr/bin/ld: skipping incompatible /opt/sac/lib/libsacio.a when > searching for > -lsacio > /usr/bin/ld: cannot find -lsacio > collect2: ld returned 1 exit status > make: *** [flexwin] Error 1 > > Qinya Liu (U. Toronto) is able to successfully compile the program > with Ubuntu > but using a slightly older (32 bit?) computer and a slightly older > Ubuntu > version: > > Linux river #1 SMP Wed Apr 1 21:47:28 UTC 2009 i686 GNU/Linux > > This led us to believe that the problem was 32-bit vs 64-bit. Is > anyone aware > of this problem? Do I need to try to compile SAC from the source > code on my > 64-bit computer? > > I would appreciate any help on this matter. > > Thanks, > Carl Tape > Harvard U. > ----------- > > _______________________________________________ > sac-help mailing list > sac-help at iris.washington.edu > http://www.iris.washington.edu/mailman/listinfo/sac-help From snoke at vt.edu Mon May 25 19:32:07 2009 From: snoke at vt.edu (Arthur Snoke) Date: Mon, 25 May 2009 22:32:07 -0400 (EDT) Subject: [SAC-HELP] Does transfer FROM FAPFILE work? Message-ID: I prepared a FAPFILE in the format listed in the transfer help file, and it is as if the amplitude were 1 and phase 0 for all frequencies. Has anyone succeeded in getting it to work? What I was trying to do was running EVALRESP externally from SAC, which produces separate AMP. and PHASE. files for each channel. I wrote a program to convert them into a FAPFILE, but ... From peterg at llnl.gov Mon May 25 21:27:47 2009 From: peterg at llnl.gov (Peter Goldstein) Date: Mon, 25 May 2009 21:27:47 -0700 Subject: [SAC-HELP] Does transfer FROM FAPFILE work? In-Reply-To: References: Message-ID: <7.0.0.16.2.20090525212650.02849b40@llnl.gov> Art, I don't think fapfile will do anything. I'm pretty sure it got stripped out with the seismanger database functionality. Sorry if my previous message didn't make this clear. Peter At 07:32 PM 5/25/2009, Arthur Snoke wrote: >I prepared a FAPFILE in the format listed in the transfer help file, >and it is as if the amplitude were 1 and phase 0 for all frequencies. > >Has anyone succeeded in getting it to work? > >What I was trying to do was running EVALRESP externally from SAC, >which produces separate AMP. and PHASE. files for each channel. I >wrote a program to convert them into a FAPFILE, but ... >_______________________________________________ >sac-help mailing list >sac-help at iris.washington.edu >http:// www. iris.washington.edu/mailman/listinfo/sac-help From snokeja at vt.edu Tue May 26 03:42:03 2009 From: snokeja at vt.edu (Arthur Snoke) Date: Tue, 26 May 2009 06:42:03 -0400 Subject: [SAC-HELP] Does transfer FROM FAPFILE work? In-Reply-To: <7.0.0.16.2.20090525212650.02849b40@llnl.gov> References: <7.0.0.16.2.20090525212650.02849b40@llnl.gov> Message-ID: <4A1BC77B.5050905@vt.edu> Thanks Peter. I somehow missed that earlier message from you. (I have been having computer problems and have missed some messages.) It is deceptive as there is still code to check for the existence of a FAPFILE. I was thinking the coding was incomplete, not that code had been deleted. That is too bad, because I do not see another way output from external EVALRESP can be introduced to SAC. Happily, FROM EVALRESP works well. What is lost is some of the options within EVALRESP. Is TRANSFER FROM DBASE also a relic? Arthur Peter Goldstein wrote: > > Art, > > I don't think fapfile will do anything. I'm pretty sure it got stripped > out with the seismanger database functionality. Sorry if my previous > message didn't make this clear. > > Peter > > At 07:32 PM 5/25/2009, Arthur Snoke wrote: >> I prepared a FAPFILE in the format listed in the transfer help file, >> and it is as if the amplitude were 1 and phase 0 for all frequencies. >> >> Has anyone succeeded in getting it to work? >> >> What I was trying to do was running EVALRESP externally from SAC, >> which produces separate AMP. and PHASE. files for each channel. I >> wrote a program to convert them into a FAPFILE, but ... From peterg at llnl.gov Tue May 26 09:36:41 2009 From: peterg at llnl.gov (Peter Goldstein) Date: Tue, 26 May 2009 09:36:41 -0700 Subject: [SAC-HELP] Does transfer FROM FAPFILE work? In-Reply-To: <4A1BC77B.5050905@vt.edu> References: <7.0.0.16.2.20090525212650.02849b40@llnl.gov> <4A1BC77B.5050905@vt.edu> Message-ID: <7.0.0.16.2.20090526093449.028e9ce0@llnl.gov> Unfortunately "transfer from dbase" also would have gone away with the removal of the seismanager and the database connections". At 03:42 AM 5/26/2009, Arthur Snoke wrote: >Thanks Peter. I somehow missed that earlier message from you. (I >have been having computer problems and have missed some >messages.) It is deceptive as there is still code to check for the >existence of a FAPFILE. I was thinking the coding was incomplete, >not that code had been deleted. > >That is too bad, because I do not see another way output from >external EVALRESP can be introduced to SAC. Happily, FROM EVALRESP >works well. What is lost is some of the options within EVALRESP. > >Is TRANSFER FROM DBASE also a relic? > >Arthur > >Peter Goldstein wrote: >>Art, >>I don't think fapfile will do anything. I'm pretty sure it got >>stripped out with the seismanger database functionality. Sorry if >>my previous message didn't make this clear. >>Peter >>At 07:32 PM 5/25/2009, Arthur Snoke wrote: >>>I prepared a FAPFILE in the format listed in the transfer help >>>file, and it is as if the amplitude were 1 and phase 0 for all frequencies. >>> >>>Has anyone succeeded in getting it to work? >>> >>>What I was trying to do was running EVALRESP externally from SAC, >>>which produces separate AMP. and PHASE. files for each channel. I >>>wrote a program to convert them into a FAPFILE, but ...