Greetings to all,

Thank you Peter for your input. So, it should be 2.400.000 for the
maximum number of points in one trace.
How does this depend on the sample rate? I thought the space taken in
memory depended only on the number of points. Having a trace with 1000
samples at 50 sps takes as much space as one with 1000 samples at 25
sps, the difference should be that the first is half the time length
of the second, am I right?

By the way, I managed to cut the MiniSEED files with WAVETOOL using
the -cseed option. 12h files at 50sps got 2160000 samples which is
under the maxpoints limit. Thank you for the MSCUT idea but being
limited to 1h or 15min files might generate too many files.
Fortunately CORR can manage 12h files at that sps.

I have two further questions tough:

- What format should the event list ("SFILE INDEXFILE" parameter)
have? Should it look like the DIRF produced files? Can we address
other folders in this SFILE parameters? (I am asking this because
after much trials I can only input s-files which are on the same
folder from where I am running CORR, and can't use the s-files
lists...).

- It looks like for each phase in the S-File, and each event S-File,
CORR only picks the maximum, and not all the local maxima above the
minimum correlation threshold, am I right?
I am asking this because my aim is not to improve picks, nor do a
correlation matrix of events, but rather, to detect new events with
similar waveforms in a continuous waveform...

Have a great weekend,
Regards,

Emanuel Antunes

2015-03-20 7:10 GMT+00:00 Voss, Peter <[email protected]>:
> Hello Emanuel,
>
> Maximum samples: it depends on your sample rate, see limits in chapter 3.1.
> On unix you can use MSCUT to split your miniseed waveform files, see mscut -h
>
> Good luck,
> Peter
>
> -----Oprindelig meddelelse-----
> Fra: [email protected] [mailto:[email protected]] På vegne af 
> Emanuel Antunes
> Sendt: 19. marts 2015 13:46
> Til: [email protected]
> Emne: [seisan] CORR and Segmentation Faults
>
> Dear SEISAN community,
>
>
> I have been trying to run the CORR routine to identify unknown/uncatalogued 
> events with similar waveforms in a continuous database, where the master 
> event waveforms are already sliced, so only the scanned continuous waveforms 
> are whole.
> My problem, so far, is that I am constantly getting segmentation faults, even 
> when MULPLT is able to deal with the waveforms (I got segmentation faults 
> while using SEISAN 9.1 MULPLT on the same waveforms, by the way). Could this 
> be a limitation of the CORR routine? What can the maximum length (in samples) 
> be? Is the data format much relevant in this case?
> I must say I sliced one of the continuous waveforms in a 6 hours segment and 
> I was able to run the program. So I suppose I could do the same for all the 
> months of data I have, but before doing that, I would prefer to have some 
> expert opinion on what is the maximum length allowed. However, I couldn't 
> split the file using WAVETOOL (Segmentation Fault)  ou rdseed_many, nor 
> SEISEI, I had to resort to ObsPy to slice/split the waveform. Is there a way 
> to do it in SEISAN?
> There must be, if MULPLT can read the long SEED files, there must be a way 
> for other programs to read them...
>
> I am currently using SEISAN 10.3, and my waveforms are in miniseed.
> All advice is very welcome.
> Thank you all in advance,
>
> Best regards,
>
>
> PS : Dear developers, there is a typo in the version tag : " SEISAN
> 10.3 ,  Janauary 2015 "
>
> --
> Emanuel Antunes
>
> Seismology Laboratory
> Instituto Superior Técnico (IST)
> Universidade de Lisboa (UL)
>
> _______________________________________________
> seisan mailing list
> [email protected]
> http://mailman.uib.no/listinfo/seisan



-- 
Emanuel Antunes

MSc in Physics Engineering
Seismology Laboratory
Instituto Superior Técnico (IST)
Universidade de Lisboa (UL)

_______________________________________________
seisan mailing list
[email protected]
http://mailman.uib.no/listinfo/seisan

Reply via email to