From timmyd at netcom.com Tue May 7 10:20:41 1996
From: timmyd at netcom.com (Tim DeBenedictis)
Date: 7 May 1996 14:20:41 GMT
Subject: Std. date/time/RA/Dec FITS keywords?
Message-ID: <4mnm7n$pf@pecos.msfc.nasa.gov>
Greetings all-
I am writing some astrometry software which needs to run with a variety
of different FITS-format images. It would be nice to use the R.A., Dec.,
date, and time information stored in the images' FITS headers, but
unfortunately there does not seem to be any standard FITS keywords which
provide this information.
I read the defining FITS paper, which describes the DATE-OBS keyword;
however, there is no corresponding TIME-OBS keyword, and one-day time
resolution is not sufficient.
I also read about the CRVAL, CRPIX, CRDEL, and CROT keywords, which sem
to specify the coordinates, rotation, and image scale. However, as far
as I can tell, no one actually seems to use these in practice! I've
worked with a variety of images from institutions like Kitt Peak, Lowell
Observatory, as well as amateur images from SBIG cameras. The only thing
I've seen is keywords like RA and DEC which then list an R.A. and Dec. in
hours/minutes/seconds as an ASCII string.
Is there any standard way of handling these things? What is the
"correct" way to do it? Please send e-mail, as I am not able to check
responses to all the groups I've posted this to.
-Tim DeBenedictis
timmyd at netcom.com
From psb at ast.cam.ac.uk Thu May 9 12:20:44 1996
From: psb at ast.cam.ac.uk (Peter Bunclark)
Date: 9 May 1996 16:20:44 GMT
Subject: Std. date/time/RA/Dec FITS keywords?
Message-ID: <4mt60p$1pn@pecos.msfc.nasa.gov>
Tim DeBenedictis wrote:
>
> Greetings all-
>
> I am writing some astrometry software which needs to run with a variety
> of different FITS-format images. It would be nice to use the R.A., Dec.,
> date, and time information stored in the images' FITS headers, but
> unfortunately there does not seem to be any standard FITS keywords which
> provide this information...[deletia - mod.]
Read :
FITS
Documents
[moral: check the 'net before pressing "post" - mod.]
From sla at umbra.ucolick.org Mon May 13 10:36:56 1996
From: sla at umbra.ucolick.org (Steve Allen)
Date: 13 May 1996 14:36:56 GMT
Subject: Std. date/time/RA/Dec FITS keywords?
Message-ID: <4n7he5$3m8@pecos.msfc.nasa.gov>
In article <4mt60p$1pn at pecos.msfc.nasa.gov>,
Peter Bunclark wrote:
>Tim DeBenedictis wrote:
>> of different FITS-format images. It would be nice to use the R.A., Dec.,
>> date, and time information stored in the images' FITS headers, but
>> unfortunately there does not seem to be any standard FITS keywords which
>> provide this information...[deletia - mod.]
>Read :
>FITS
>Documents
This reply may give the impression that there is a definitive answer
to the question. There is not.
There are no adopted standards specifying the keyword name or format
of RA and Dec in a FITS header. There exists decimal degrees, decimal
hours, and hexadecimal strings in hours and degrees. The given RA and
Dec may be expressed in a catalog reference frame or as notably
inaccurate telescope encoder readings.
Similarly there is no adopted standard specifying the keyword name or
format of observation time in a FITS header. Again, existing practice
varies rather widely as do the accuracies of the clocks used to obtain
the time.
There is a draft document at
http://fits.cv.nrao.edu/documents/wcs/wcs.html
whose adoption should solve this problem, but its recommendations
are not generally found in any existing FITS headers.
--
Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064
sla at ucolick.org Voice: +1 408 459 3046 FAX: +1 408 454 9863
WWW: http://www.ucolick.org/~sla PGP public keys: see WWW
From psb at ast.cam.ac.uk Tue May 14 03:46:58 1996
From: psb at ast.cam.ac.uk (Peter Bunclark)
Date: Tue, 14 May 1996 08:46:58 +0100
Subject: Std. date/time/RA/Dec FITS keywords?
References: <4n7he5$3m8@pecos.msfc.nasa.gov>
Message-ID: <31983A72.2310@ast.cam.ac.uk>
Steve Allen wrote:
>
> In article <4mt60p$1pn at pecos.msfc.nasa.gov>,
> Peter Bunclark wrote:
> >Tim DeBenedictis wrote:
> >> of different FITS-format images. It would be nice to use the R.A., Dec.,
> >> date, and time information stored in the images' FITS headers, but
> >> unfortunately there does not seem to be any standard FITS keywords which
> >> provide this information...[deletia - mod.]
>
> >Read :
> >FITS
> >Documents
>
> This reply may give the impression that there is a definitive answer
> to the question. There is not.
>
> There are no adopted standards specifying the keyword name or format
> of RA and Dec in a FITS header. There exists decimal degrees, decimal
> hours, and hexadecimal strings in hours and degrees. The given RA and
> Dec may be expressed in a catalog reference frame or as notably
> inaccurate telescope encoder readings.
>
> Similarly there is no adopted standard specifying the keyword name or
> format of observation time in a FITS header. Again, existing practice
> varies rather widely as do the accuracies of the clocks used to obtain
> the time.
>
> There is a draft document at
> http://fits.cv.nrao.edu/documents/wcs/wcs.html
> whose adoption should solve this problem, but its recommendations
> are not generally found in any existing FITS headers.
>
> --
> Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064
> sla at ucolick.org Voice: +1 408 459 3046 FAX: +1 408 454 9863
> WWW: http://www.ucolick.org/~sla PGP public keys: see WWW
I agree that there hasn't been much standardisation up to present; but
anyone writing new software should study the available documents on
wcs, including the section in NASA's ``A Users's Guide for the Flexible
Image Transport System (FITS)'', and apply the recomendations to his
instrument/system.
The disagreements seem to be in how exactly to represent the
image skew, rotation and scale. The actual convention for representing
the coordinates of the reference pixel are as laid down in the
mystical PAPER I , ie CRVALn in degrees for ra and dec. Meanwhile,
most observatories have actually used RA and DEC keywords in more or
less human-readable format, but these shouldn't in fact be part of
a wcs.
Since widely-distributed software (eg SAOTNG) is becoming available
to exploit images with wcs, there is much more motivation than
previously to exploit the potential.
Peter.
From rlm at astro.physics.uiowa.edu Thu May 16 20:02:48 1996
From: rlm at astro.physics.uiowa.edu (Robert Mutel)
Date: Fri, 17 May 1996 00:02:48 GMT
Subject: Std. date/time/RA/Dec FITS keywords?
References: <4n7he5$3m8@pecos.msfc.nasa.gov>
Message-ID: <319BC228.1E5A@astro.physics.uiowa.edu>
Readers of sci.astro.research may be interested in a collection of programs for FITS images called ATFtools.
It includes a tool for inserting world coordinates into a FITS image header by pattern matching to the
Guide Star Catalog. The program (wcs) uses the nominal ra, dec, keywords (along with keyword crpix for
plate scale) to roughly locate the image center and either an online or CD-ROM GSC database to do the pattern
matching. The programs have been ported to several UNIX systems (Linux, Solaris, Unixware) and Windows-95.
There are also tools for batch photometry, FITS header editing, and multiple image alignment. and cropping.
The Windows-95 port is available at