Sheratan (beta ARI)

[James Foster]

To: Robin,

RE:"Find the star you want eg Castor from http://www.eso.org/sci/observing/tools/ ... tonow.html
and click on "plot"

Thanks! I ended-up buying the LISA visual unit so this will be my training tool for Isis spectra work.
However, I "upgraded" my L-200 spectroscope with an Alpy calibration module, so I'm still using this for
my Solar Analog work (CaK & Hy regions).

I was able to open up the ESO link for Castor in the spectral region I work at in BASS so I'll see if I can get
more "accurate" instrument response curves.

I'll probably cross-over to Isis when I get a LHRESIII with a 1800 ln/mm grating next year or so or sooner once
I go through the tutorials Christian puts out for the LISA spectroscope. Thanks again for the ESO links!

[Ken Harrison]

Olivier,
On what basis do you say ""For good accuracy and good process, ISIS is better than BASS ""???
I agree for batch processing ISIS is very useful, but I've yet to see any data which shows that BASS Project is any less accurate than any of the other software mentioned.

[Robin Leadbeater]

To get you started here is the UVES Castor spectrum (from 3500-8000A) in 1D fits format. Note that this is a "real" spectrum and still includes telluric lines so if you are working in a region with telluric lines it is a good idea to clean them out of the UVES spectrum first. VSpec and ISIS have a tool for this, not sure about BASS. (Note that the tellurics have been removed in UVES H alpha spectra in the ISIS database.)

Cheers
Robin

EDIT : Hmm... the file seems to have undergone a wavelength shift in the conversion to fits so here is the dat version instead
EDIT: Still a problem so I have removed the attached file while i sort out what has gone wrong

Just for completeness, the reason for the problem I had here was that with the UVES spectra the wavelength calibration is non linear, with unequal bin widths so are not compatible with spectral potting programs which expect equal bin widths. (I recommend checking how BASS deals with this ISIS plots correctly but cannot save with unequal bin widths. It does have a tool to convert unequal to equal bin widths though.)
http://www.spectro-aras.com/forum/viewt ... f=6&t=1648
(This is one of the many reasons why I now use ISIS for all my processing. Christian Buil has an excellent understanding of the subtleties of spectroscopy and data reduction which is the main reason I trust my data to ISIS.)

Cheers
Robin

[John Paraskeva]

recommend checking how BASS deals with this ISIS plots correctly but cannot save with unequal bin widths. It does have a tool to convert unequal to equal bin widths though

This is a potentially complex area. I have carried out some initial analysis using BASS that may be relevant to the above.

The UVES text format can be modified to a DAT format quite easily using the Excel text wizard and some copy/paste with Notepad. I quickly ended up with a 3 column tab delimited test file without column headers and renamed it with a DAT extension.

BASS reads only the first 2 columns (1=0 or lamda, 2=flux), of a DAT profile, the 3rd column is ignored. The 3 column DAT file therefore loaded straight in (ignoring 0.1 values in last column).

Also worth clarifying that BASS does not assume points are equally spaced (unless its a FIT with calibration keywords populated). The X location is determined by the first column if a DAT file and/or by calibration calculations.

I used VSpec to export the same UVES data to DAT for comparison purposes. (I could not create a VSpec FIT due to a run time error).

Both DAT profiles look identical when plotted in BASS even when zoomed in for scrutiny and without any calibration offset.

The identical results imply that the 3rd (sigma) column in UVES file does not appear to be used in VSpec, suggesting the first two columns are just x,y pairs and not Gaussian curves wit sigma=0.1? If each row was a Gaussian point than we may expect a Gaussian to be plotted centred on the Y value. Each point would have overlapping points summed, say between -3*sigma and +3*sigma? Perhaps sigma is used in a different way?

Depending on clarification of the above, it would be very simple for BASS to load UVES format files. I think keeping the existing .uvs file extension as standard would be good.

Saving files
I also tested using BASS to save the DAT file imported from VSpec. On examining the output, the same wavelengths and flux were shown. The non-linearity is preserved for DAT files having lamda as the first column. (Note: it took a long time to export 409 thousand rows, but the fix to speed this up is ready for the next release).
Saving the DAT file as a FIT showed a offset issue (similar to ISIS).

Investigation to the FITS offset issue to follow....

regards
John

[Olivier GARDE]

Olivier,
On what basis do you say ""For good accuracy and good process, ISIS is better than BASS ""???
I agree for batch processing ISIS is very useful, but I've yet to see any data which shows that BASS Project is any less accurate than any of the other software mentioned.

Hi Ken

I don't use bass but I have read the manual and have not seen anything about wavelength calibration with a standard lamp (Th/Ar or Neon lamp), how to selected the best emissions lines with this software ?
Bass use element present in the target or reference star, (for example Hydrogen HAlpha Hbeta and Hgamma lines) it works in low resolution but not in high resolution because lines is shifted with heliocentric velocity and also proper motion of the target. This is not an absolute wavelength calibration.Of course I think you can do the job manually, selecting the lines in Neon lamp, but it's long to do...

With ISIS, the software use Neon or Ne/ar or Th/ar lines of the calibration lamp, you can choose the lines from the calibration lines and remove for example blended lines. During process wavelength calibration, you only tell to the software the position of a typical line of Neon lamp.

If a SA, LISA, Alpy, LHires or eShel spectrograph is used, ISIS has functions and treatments dedicated to each spectro and the process depend of witch spectrograph you use.

For example, Alpy need to correct smile and tilt and LISA need to correct slant and tilt.

ISIS is also able to process echelle spectrum.

From my point of view, ISIS is very good in spectrum processing for the repeatability of measurements from one spectrum to another. (using exactly the same parameters and the same process).

[Andrew Smith]

I have used BASS with Th Ar lamp to calibrate my echelle spectra. ISIS does not like my grating which is not the same as used in the eshel.

It seems accurate based on the position of Th Ar not used in the calibration.

Regards Andrew

[Ken Harrison]

Robin, et al,
BASS Project is very good for calibration and for slit spectrographs can make use of any of the standard calibration lamps (I use the FILLY, RELCO and Neon).
The usual tilt, slant, smile corrections can be corrected as well.
The visible lamp lines are not "pre-selected" but easily recognised and calibrated from the drop down list of standard lines. Telluric and others are also included.
Heliocentric corrections can be also be applied.
John has just released a new update V1.9.6

I accept that Christian has done a great job with the ISIS package and that it is used by many amateurs, but that doesn't mean that other packages will not be developed and used moving forward. We already have Tom's basic RSpec for novices and Benji's SPCAudAce as well as Valerie's VSpec - all these have been developed over a period of time with great effort to support the amateur astronomical community.
We should help and actively support these efforts.

[Francois Teyssier]

Dear Ken,

The aim of ARAS project is to provide and share high level spectra of various classes of stars.
We all know that the learning curve is often long and difficult. Basic softwares can represent an alternative for first steps, why not.
One of the goal of ARAS forum is helping the arrivals in spectroscopic world.
But, we consider that ISIS should be the reference software, for its high quality treatment.
This is the axis developed by Francois Cochard in his book (soon translated in English)
Beyond the quality of Christian's software, the use of the same software has a major interest : reducing the dispersion of the final results
This is the raison why I strongly recommend ISIS as the reference software for the treatment of spectra accpeted in ARAS data base.
Of course, this not mandatory.

Best regards,

François Teyssier

[Andrew Smith]

Unfortunately for me ISIS will not process my spectra. It is very focused on shelyak products and if you deviate from them at all your success with it may well be limited.

While not as slick as ISIS in some ways BASS if very good and not focused on a particular brand.

I you want ARAS to be a Shelyak only site then please say so and I will withdraw.

I had hoped it was more than that.

If you want the best maybe we should use IRAF.
.
Regards Andrew

[Joan Guarro Flo]

Hi Andrew:

At this time, I have six operating spectrographs, all I have built I included my the Echelle Mussol, (most of these spectrographs have nothing to do with designs Shelyak), and I have no problem with ISIS. On the contrary, in my personal opinion, ISIS is the best software for data reduction.

Possibly this Forum, have Shelyak members, but all of them have always had an exquisite behavior, and they have not done more of proselytism of their mark.

If you want to reduce your spectra with ISIS, and you have a problem, you can submit and someone can help.

Best regards, Joan.