PSN-L Email List Message

Subject: R: PSN file format and long file names.
From: "Mauro Mariotti" mariottim@............
Date: Tue, 10 Jul 2001 13:49:05 +0200


In my experiments:

binary data can be compressed with a considerable amount of space.
ascii data can be compressed saving an enormous amount of space more than
70%
the result, due to the higher dimensions of ascii files for the same number
of samples,
are the same, in other words a compressed ascii file and a compressed binary
file
uses the same space.
This is obvious because the quantity of information is the same.
The interesting thing is that using compressed files AND long file names you
can archive
events that remains easily identifiable by the name itself, if need one you
can expand it
and view it with winquake or an other event viewer.

Mauro



----- Original Message -----
From: Bob Smith 
To: 
Sent: Tuesday, July 10, 2001 12:51 PM
Subject: Re: PSN file format and long file names.


> A brief comment on relative (binary vs. ASCII) file formats from a non
> seismologist, non-PSNer.
>
> I suspect that *once compressed* by a good file compression program such
> as PKZIP or WINZIP, the size difference will be much less that you might
> think at first.  Consider:
>
> 1.  Expanding a binary file to ASCII format does not, indeed *cannot*,
> add information to the file (unless, of course, the conversion program
> modifies the information content of the binary file in the process but,
> then you are not comparing apples to apples anymore.)
>
> 2.  The goal of a good *lossless* compression program program (PKZIP /
> WINZIP) is to eliminate redundancies and  minimize entropy.  The
> information content of the compressed file is identical to the original
> file (thus loss less).
>
> As a practical exercise, I suggest that someone, I don't have access to
> sample files, convert a few binary format files to ASCII, compress them
> both and compare both the pre compressed and post compressed files and
> their size ratios.
>
> Sincerely, Bob Smith
>
>
>
> Larry Cochrane wrote:
> >
> > Mauro,
> >
> > Not to beat a dead horse I will only make a few more comments about
ASCII
> > verses Binary. I do want to get some feedback from other members on long
> > file name standard for the PSN event files.
> >
> > Mauro Mariotti wrote:
> >
> > > On the expandibilty of PSN4 format (touchè) i'm not examined it well
> > > even if the expandibilty of a binary record is not too easy to manage
> > > especially
> > > for people that wants to stick one standard but is not so much expert
on
> > > programming tecniques.
> >
> > First, its not that much more difficult to deal with binary files....
Also,
> > I don't think a standard should be created or designed around the
ability of
> > someone's programming skills. This is not to say it should be so
> > complicated, like the SEED format, that even a skilled programmer has
> > problems extracting data from it.
> >
> > >
> > > Using archive compressor the 8 char filename length is a problem.
> > > Your winquake when access a database of events do a great work
analizying
> > > and reporting
> > > date time location and comments on the file you're opening.
> > > If the events file are compressed you cant'do that with winquake.
> > > Using long names you could create a database readable with Computer
> > Resource
> > > of windows
> > > creating filenames like:
> > >
> > > 20000701_145023_PG_E-W_KURILI-ISLANDS_M6.5.TXT
> > >
> > > I think you should modify SDR to use long names.
> > > If i'm not in wrong SDR is DOS based and this can't be done.
> > > Anyway this is my viewpoint:
> >
> > Since SDR is a DOS program it is limited to the 8.3 format. WinQuake and
my
> > archival system could save the files using a longer name. I would like
to
> > start a discussion on the new file name format. You can see Mauro
proposal
> > above. I propose a shorter format:
> >
> > YYYYMMDD.HHMMSS.COMPONENT_TYPE.SENSOR_ID.PSN
> >
> > Example: 20010705.234505.bhz.lc1.psn
> >
> > This is similar to how other seismic data archive systems on the net
store
> > event files.
> >
> > The first two set of numbers are the date and time followed by the
component
> > type like BHZ or LHN and then the sensor ID, this can be 2 to 6
characters
> > long. By ending the file in PSN, Windows can associate it as an event
file
> > and open it with a program like WinQuake if someone clicks on the file
name
> > using Explorer. Maybe just 2 characters for the year would be ok? This
will
> > become a problem in about 90 years! Anyone else want to suggest a
format?
> >
> > -Larry Cochrane
> > Redwood City, PSN
> >
> > __________________________________________________________
> >
> > Public Seismic Network Mailing List (PSN-L)
> >
> > To leave this list email PSN-L-REQUEST@.............. with
> > the body of the message (first line only): unsubscribe
> > See http://www.seismicnet.com/maillist.html for more information.
>
> --
>
> ---------  Avoid computer viruses  --  Practice safe hex  -------------
>  * * Specializing in small, cost effective embedded control systems * *
> Robert L. (Bob) Smith Smith Machine Works, Inc.
> internet   bobsmith5@........ 9900 Lumlay Road
> landline   804/745-1065                 Richmond, Virginia 23236+1004
> __________________________________________________________
>
> Public Seismic Network Mailing List (PSN-L)
>
> To leave this list email PSN-L-REQUEST@.............. with
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.
>

__________________________________________________________

Public Seismic Network Mailing List (PSN-L)


[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>