From: Rune Allnor on
On 19 Jun, 15:39, Walter Roberson <rober...(a)hushmail.com> wrote:
> Rune Allnor wrote:
> > Matlab? Forget it. Seismology is one of the really heavy
> > computational applications, where every drop of performance
> > matters. You can use matlab to play with the algorithms,
> > but once the going gets tough you need to throw it away.
>
> ... and code the algorithm in a language designed for numeric
> performance, such as Fortran.

Wrong. The main problems with seismic data sets is logistics.
You need to be able to access data freely and unconstrained,
which means that languages that give you fine control of
memory managment easily will beat systems that focus on mere
numerics.

But you are, unfortunately, right in that fortran is still
used in these systems. However, the pros had already switched
to C some 15 years ago, when I took my course on seismic
programming. Somebody at the system vendors told me they would
have preferred to go C++, but decided against it due to the
non-standard name mangling in C++ object code.

Rune
From: Steven Lord on

"Rune Allnor" <allnor(a)tele.ntnu.no> wrote in message
news:3d0ec7bb-251e-4e38-ad2c-3365ccb40ea2(a)w31g2000yqb.googlegroups.com...
> On 19 Jun, 11:11, "samee baig" <sameebai...(a)hotmail.com> wrote:
>> hello
>> I am a student of electronic engineering.I have been give project on the
>> relationship or application of digital signal processing and seismology
>> and how we can use matlab in this respect.please provide me with some
>> information on this project.
>
> Signal processing in seismology: There was a special issue of
> Proceedings of the IEEE some time in 1984 - I *think* it was
> september or october.
>
> Matlab? Forget it. Seismology is one of the really heavy
> computational applications, where every drop of performance
> matters. You can use matlab to play with the algorithms,
> but once the going gets tough you need to throw it away.

Well, admittedly only one of the user stories in the Earth and Ocean
Sciences category deals with earthquakes (more deal with "ocean sciences"
than "earth sciences") but you might find Geoscience Australia's experiences
an interesting read.

http://www.mathworks.com/products/matlab/userstories.html?code=earth&by=industry

--
Steve Lord
slord(a)mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com