Fortran - Read binary data generated by Scientific Atlanta 1590 Pattern Recorder

Author

Read binary data generated by Scientific Atlanta 1590 Pattern Recorder

Peter

2007-02-22, 10:06 pm

Greetings.

I would like to be able to read in Fortran or Matlab the contents of a
number of data files that I have been given. The files were generated
by the software that came with a Scientific Atlanta model 1590 pattern
recorder, a device used to record and plot antenna far-field
patterns. The software was written by MI Technologies and has a
version number of 1.2. When I edit one of these files with emacs, I
see quite a bit of recognizable character data, but also a large
amount of incomprehensible binary data. An example file is located
at
http://www.vcnet.com/~simonp/sa.dat
I think that the data is simply an x-y plot.

Thanks in advance for any help,

Peter

salgerman@gmail.com

2007-02-23, 4:15 am

It is not clear to me what you are asking.

I mean, sure you can read binary data with a computer program, but you
need to know the format of what you are reading.

If you don't know the format of such file, it could be because it is
proprietary. It is probably binary to save space, but it could
additionally be totally encrypted or not totally self contained and
you need their software to visualize it.

I had a similar task a long time ago with a file that came out of an
oscilloscope from Nikolai (?), and while the first recorded file had a
very specific format, they had a couple of utilities that converted
those files to plain ascii files that you could read or even simply
paste into a spreadsheet program.

So I would say: Look into their provided software/utilities and see if
they have a little program to translate the file into a more standard
format.

gsal

Peter

2007-02-23, 4:15 am

On Feb 22, 6:36 pm, Gary Scott <garylsc...@sbcglobal.net> wrote:
> Peter wrote:
>
>
> The best solution (using an extension) might be provided if you provide
> the compiler that your using.
>
>
>
>
>
> --
>
> Gary Scott
> mailto:garylscott@sbcglobal dot net
>
> Fortran Library: http://www.fortranlib.com
>
> Support the Original G95 Project: http://www.g95.org
> -OR-
> Support the GNU GFortran Project: http://gcc.gnu.org/fortran/index.html
>
> If you want to do the impossible, don't hire an expert because he knows
> it can't be done.
>
> -- Henry Ford

I'd actually prefer to do it in Matlab, but if necessary I could write
a short Fortran routine to convert the files to ascii first. I have
both the Intel v. 9.1 compiler and g95 available.

--Peter

Peter

2007-02-23, 4:15 am

On Feb 22, 8:18 pm, salger...@gmail.com wrote:
> It is not clear to me what you are asking.
>
> I mean, sure you can read binary data with a computer program, but you
> need to know the format of what you are reading.

Yes, that is the problem: I don't know the format of the data file.
I would like to read it in a meaningful way so that the data is
decoded from its unknown binary format.
>
> If you don't know the format of such file, it could be because it is
> proprietary. It is probably binary to save space, but it could
> additionally be totally encrypted or not totally self contained and
> you need their software to visualize it.

It is not encrypted because there are human-readable character strings
interspersed with the binary numerical data. Feel free to look at the
file I provided a link to and you will see this.

>
> I had a similar task a long time ago with a file that came out of an
> oscilloscope from Nikolai (?), and while the first recorded file had a
> very specific format, they had a couple of utilities that converted
> those files to plain ascii files that you could read or even simply
> paste into a spreadsheet program.
>
> So I would say: Look into their provided software/utilities and see if
> they have a little program to translate the file into a more standard
> format.
>
Thanks for the suggestion; it had already occurred to me to do this.
Unfortunately, I'm about 300 miles away from the facility that sent me
the data. I have asked the technician to send me any documentation on
the program and/or the pattern recorder and he hasn't been able to
locate any. The hardware dates from approximately 1994, I believe.
The technician is able to generate hard copies of the pattern cuts,
but I would like to read them into Matlab and compare with theoretical
predictions. The data format is that used by the recorder software to
store and retrieve files from the disk. In the worst case I could
digitize the hard copies but I would like to avoid doing so if
possible.

Thanks,
Peter

--Peter
> gsal

Herman D. Knoble

2007-02-23, 8:08 am

Peter, There is a free Hex Editor at: http://www.chmaas.handshake.de/
which shows that your data is a mixture of ASCII and binasry Integer data as
you suggested. After an ASCII flag: TRACE1 DATA
there is REAL (binary) data

By viewing this data (SA.DAT) you will then be able to write trial codes to
eventually handle this data.

Another suggestion would be to contact MI Technologies
http://www.mi-technologies.com/
and attempt by phone to contact a technical person who
might be able to give you a data descriptor for this data.

Skip Knoble

On 22 Feb 2007 18:20:46 -0800, "Peter" <petersamsimon2@hotmail.com> wrote:

-|Greetings.
-|
-|I would like to be able to read in Fortran or Matlab the contents of a
-|number of data files that I have been given. The files were generated
-|by the software that came with a Scientific Atlanta model 1590 pattern
-|recorder, a device used to record and plot antenna far-field
-|patterns. The software was written by MI Technologies and has a
-|version number of 1.2. When I edit one of these files with emacs, I
-|see quite a bit of recognizable character data, but also a large
-|amount of incomprehensible binary data. An example file is located
-|at
-|http://www.vcnet.com/~simonp/sa.dat
-|I think that the data is simply an x-y plot.
-|
-|Thanks in advance for any help,
-|
-|Peter

Steven Lord

2007-02-23, 7:10 pm

"Peter" <petersamsimon2@hotmail.com> wrote in message
news:1172205907.684533.291530@8g2000cwh.googlegroups.com...
> On Feb 22, 8:18 pm, salger...@gmail.com wrote:
>
> Yes, that is the problem: I don't know the format of the data file.
> I would like to read it in a meaningful way so that the data is
> decoded from its unknown binary format.

Since you said you've already asked the technician to look for documentation
on the format of that file, and he was not able to find any, I suggest you
check if Scientific Atlanta has any information they can give you:

http://www.sciatl.com/

They may have older manuals available for reference, and may be able to send
you a scan of the relevant section. It's worth a shot.

--
Steve Lord
slord@mathworks.com

Peter

2007-02-23, 7:10 pm

Peter

2007-02-23, 7:10 pm

On Feb 23, 5:46 am, Herman D. Knoble <SkipKnobleL...@SPAMpsu.DOT.edu>
wrote:
> Peter, There is a free Hex Editor at: http://www.chmaas.handshake.de/
> which shows that your data is a mixture of ASCII and binasry Integer data as
> you suggested. After an ASCII flag: TRACE1 DATA
> there is REAL (binary) data
>
> By viewing this data (SA.DAT) you will then be able to write trial codes to
> eventually handle this data.
>
> Another suggestion would be to contact MI Technologieshttp://www.mi-technologies.com/
> and attempt by phone to contact a technical person who
> might be able to give you a data descriptor for this data.
>
> Skip Knoble
>
> On 22 Feb 2007 18:20:46 -0800, "Peter" <petersamsim...@hotmail.com> wrote:
>
> -|Greetings.
> -|
> -|I would like to be able to read in Fortran or Matlab the contents of a
> -|number of data files that I have been given. The files were generated
> -|by the software that came with a Scientific Atlanta model 1590 pattern
> -|recorder, a device used to record and plot antenna far-field
> -|patterns. The software was written by MI Technologies and has a
> -|version number of 1.2. When I edit one of these files with emacs, I
> -|see quite a bit of recognizable character data, but also a large
> -|amount of incomprehensible binary data. An example file is located
> -|at
> -|http://www.vcnet.com/~simonp/sa.dat
> -|I think that the data is simply an x-y plot.
> -|
> -|Thanks in advance for any help,
> -|
> -|Peter

Thanks Steven and Skip for your useful suggestions. I tried both SI
and MI (apparently, MI was spun off of SI about 10 years ago, so MI is
the right contact at this point). The guy who might know the format
is out today, but the person I spoke to sent me a copy of the software
documentation. It turns out that the software is capable of exporting
CSV files, which I am able to read. I will have the technician do
this. In the meantime, for future reference, could you (Skip) please
explain how you were able to distinguish the binary integer data from
the binary floating point data? How many bytes for each? Were you
able to discern the particular floating point representation?

Thanks,
Peter

Terence

2007-02-23, 7:10 pm

I like puzzles. I had a look at the data and it has a structure.
Starting with a binary/ascii dump.

Text is followed by a binary zero, or a #0A for further lines.
A variable name is followed by a binary zero, then none or four binary
characters, then always a 24-byte set byes which are interger
variables which seem to be

increase increase slowly Measure? slow increase Measure?
< rapidly ><---------><---><-----><----------><----><----><--------->
63 02 00 00 00 00 00 00 0B 00 0B 00 53 02 00 00 05 00 05 00 5E 02 00
00

Later sample
A3 03 00 00 00 00 00 00 07 00 07 00 98 03 00 00 04 00 04 00 9F 03 00
00

Later the format switches to un-named varaiables of 16-bytes groups
which seem to an id and a measure of individual plots.

Typical texts and variable names are:-
SA MODEL 1590
BEAMWIDTH
XAXIS TYPE
XAXIS ANG RNG..
XAXIS SLOTWIDTH
YAXIS AMP MODE.
YAXIS PHS RNG..
LBL TITLE.CMB STAR SLOT ARRAY
LBL TEXT1.Engr:Dan Arroyo
LBL TEXT2.Freq:2.646 GHz
LBL TEXT3.CMBS028, AUT 0 D TX 0
LBL TEXT4.Remarks: CMB Star Slot A
LBL PREFIX.CMBS
LBL DATE.01/23/07
LBL SEQNUMBER.025
LBL COMMENTS.
These comments are retained in.sa
They may be viewed using the.
Remarks Dialog..
The comments won't appear on the.
TGIF BLUE SKIES
OFF X1.
OFF Y1.
OFF X2.
OFF Y2.
OFF X3.
OFF Y3.
OFF X4.
OFF Y4.
OFF X5.
OFF Y5.
OFF X6.
OFF Y6.
OFF X7.
OFF Y7.
OFF X8.
OFF Y8.
OFF X9.
OFF Y9.
OFF X10
OFF Y10
GAIN CORRECTION
GAIN COMPUTED
SGH FILENAME d:\sgh\SA12_09 sgh
SGH FREQUENCY
SGH GAIN
SGH TRACE
AUT POLZN
AUT TRACES1
AUT TRACES2
GRID TYPE
GRID RESOLN
GRID AXES
GRID REVSCALE
REF CHART
REF AMPTD
REF PHASE
CHART SPAN
CHART CENTER
AMPTD SPAN
AMPTD MAX
PHASE SPAN
PHASE CENTER
AXISLBL CHART Angle (Deg)
AXISLBL AMPTD Relative Power (dB)

Sponsored Links