Hello all,

I followed the David's recipie. Unfortunately, it was not successful.

Scoords in onespec gives the following error:

on> scoords march042007  coords.dat  label=Wavelength  units=Angstroms
Warning: Wrong number of coordinate values in file

Timur
==

PS: I have checked "min_lenuserarea" and it is not due to this environment which defines the maximum size of the image header.

On Fri, Oct 17, 2008 at 3:08 PM, Malcolm Currie <[log in to unmask]> wrote:

ECHOMOP stores it calculations within an NDF extension.  Run

  % hdstrace <ndf>

where <ndf> is the name of your ECHOMOP output NDF to see what I mean.  This
structure permits reductions to be stopped and later resumed easily, and
allows a cloning system using such data as a template.

NDF2FITS does not know how it should interpret this specialised reduction
data structure.  NDF2FITS can retain all the ECHOMOP data in FITS form in a
binary table, but again some arbitrary FITS reader will not know how to
interpret the various columns.

For interchange with another package/system somewhere the semantics of the
structure must be known, and the relevant wavelength array identified, and
converted to a standard form.  As far as I can see the ECHOMOP reduction
extension is not properly documented.  This is naughty.  [ECHOMOP was
originally written by an astronomer outside of Starlink, and Starlink spent
considerable effort making it useable for regular users, not least through
its Cookbook series:
http://www.starlink.rl.ac.uk/star/docs/sc3.htx/sc3.html  and
http://www.starlink.rl.ac.uk/star/docs/sc7.htx/sc7.html .]  ECHOMOP should
have some export option, where the wavelength scale stored in component
.MORE.ECHELLE.ECH_2DWAVES (ref. ECHTRACT recipe in SC/3) is placed in a
standard NDF component, such as AXIS.  I would have thought ech_result to an
NDF would have done this.   David's suggestion offers another way to export
the wavelength information.  If this works out, please let us know.  We can
then include this interchange recipe in one of the cookbooks.

>googling on "onedspec splot" suggests you need the wavelength info in
>IRAF "multispec" format. Looking at the docs for ndf2iraf suggests it
>should be able to do this conversion.

The only time I've used splot was to test the NDF2IRAF conversions,
including multispec, many years ago.

David: interesting that multispec lives on despite the -TAB algorithm in
FITS, and the move from the Old IRAF Format to FITS.

Malcolm