IRAF dealing with 2df data

Chen Jianjun wrote on Nov 22, 2000

Hi, all

I'm  a graduate student and now dealing with some sample data of 2df
with dofibers while I meet some difficulties.
It seems dofibers finds apertures along a line, and 2df along a column.
When I rotated the image by 90 degree with
the task rotate, the pipeline goes on and it does correctly find the
apertures. But it gives warning "the coordinate ignored, (rotated?),
pixel coordinate instead" while it goes to calibrate wavelength. So I
cannot get calibrated 1-D spectra.

Now I want to know is there any better way to rotate a 2-D image to
avoid this situation? Or, is there anybody who can give me some hints
about multifiber spectra reduction?

Thanks a lot

--
========================
Chen Jianjun
LAMOST, Beijing Astronomical Observatory
Beijing, 100012
Tel.:64877703 (O)

---

Frank Valdes wrote on Nov 22, 2000

Hello Jianjun,

The factor I think you are missing is that in the SPECRED package
(I assume this is the package you load to run DOFIBERS) you have to
do "epar specred" and set the parameter "dispaxis" to 1 if the spectra
are horizontal when displayed or 2 if the spectra are vertical.  The
way the software determines the dispersion direction is to first look
for a keyword DISPAXIS in the header (so you could use HEDIT to add it)
and if it is not there it uses the "dispaxis" package parameter.

Rotating or transposing the image can be made to work but there are
tricks that have to be done.  The simplest is to clear the coordinate
system after the rotate with "wcsreset <image> wcs=world".  I'm not
sure but you might also need "wcsreset <image> wcs=physical".

Let us know if you have further questions.  As you see your newgroup
message reaches us.  Alternatively you can write directly to us
at iraf@noao.edu.

Cheers,
Frank Valdes

---

Eon Chang Sung wrote on Nov 22, 2000

irafnet@noao.edu wrote:

> Hi, all
>
> I'm  a graduate student and now dealing with some sample data of 2df
> with dofibers while I meet some difficulties.
> It seems dofibers finds apertures along a line, and 2df along a column.
> When I rotated the image by 90 degree with
> the task rotate, the pipeline goes on and it does correctly find the
> apertures. But it gives warning "the coordinate ignored, (rotated?),
> pixel coordinate instead" while it goes to calibrate wavelength. So I
> cannot get calibrated 1-D spectra.
>
> Now I want to know is there any better way to rotate a 2-D image to
> avoid this situation? Or, is there anybody who can give me some hints
> about multifiber spectra reduction?
>
> Thanks a lot
>
> --
> ========================
> Chen Jianjun
> LAMOST, Beijing Astronomical Observatory
> Beijing, 100012
> Tel.:64877703 (O)

After rotating spectral images using by the task 'rotate', some header informations would be

produced and modified. Then you have to edit the image headers properly the header informations.

The newely formed header parameters by the task 'rotate' are 'ltv1, ltv2, ltm2_1, ltm1_2', and

 modified parameters 'ltm1_1 and ltm2_2'. Compare the header parameters before and after

 rotating. Set zero or none for the newely formed header parameters and same values of images before

rotating for the modified parameters.

--
==========================================================================
 Eon-Chang Sung, Ph.D.(Astronomer)    |     Tel.   :   82-42-865-3229
 Korea Astronomy Observatory          |     Fax.   :   82-42-861-5610
 61-1, Whaam-dong, Yusong-gu          |     email  :   ecsung@kao.re.kr
 Taejeon 305-348, R. of KOREA         |
==========================================================================

---

Last post on Nov 22, 2000