ISIS Tutorial - problem

[Andrew Smith]

Hi - I have been following the ISIS tutorial but when I compare my reduced Vega spectra with the reference spectrum they don't line up. The same happens if I use my own data. What am I doing wrong. I have followed the instructions line by line!

Thanks Andrew

[Andrew Smith]

I have done some more investigation. I am using v5.1.0 and find some differencies with the tutorial. Looking at the screen shot which compares the @@.fit with the responce curve looking at the smoothing factor 20 at wave length 4204 my intensity is 0.9 while the tutorial is 0.5316. Also when I display the flat_vega my "Domaine is 32767 and 0 rather than the tutorial 65535 and -512 (with the pointers at there extreem high & low values). I also get other small differences (3rd or 4th figure) e.g. in the calibration wavelengths. I have a Win7 PC with the latest updates. Regards Andrew

[Robin Leadbeater]

Interesting.

I worked through the ISIS tutorial for the Star Analyser the other day and had the same problem. The intensity of the response corrected spectrum did not match that of the libray spectrum used to calculate the instrument response.

Cheers
Robin

[Christian Buil]

Andrew,

Your problem is not a real problem and your processing is success full.

The difference between the two spectra is the relative scaling.
Remember, the vertical scale intensityvalues (Y-axis) are relative and arbitrary.
For compare intensities along the spectral axis it is necessary
to scale the compared spectra at the same value for a given wavelength. For example
we can decide to attribute the mean intensity 1.0 in the spectral
range 6650 A - 6750 A. All point of the spectrum are multiply by
a constant for respect these condition.

Look "Settings" tab and "Scaling profile spectral region" and
enter the values 6650 and 6750 for wavelength limits (see first
screen copy of LISA tutorial). (you can choose other values of courses,
a region free of major spectral lines is ideal, because the level
is representative of continuum).

You can also scale from the "Profile" tab by using "Normalize" command
(button "Normalize"). If necessary, for example, you can scale (normalize) to unity the reference
spectrum for the same spectral range and resave the scaled version before
comparison.

I hope to solve the difficulty.

Christian B

[Andrew Smith]

Thanks Christian - That explains it. Once I normalise them at the same point it works fine. Regards Andrew