Sensfunc: how is it defined?

Eduard Westra wrote on Mar 14, 2006

I have several standard stars taken on different nights, but all with the same instrument. First of all, when I go through the whole reduction, I get in the procedure

sensfunc

that the stars are not on the same magnitude.

This could potentially be explained by some clouds. Seeing is not a problem, since 5 arcsec slits were used, and some simple calculations show that even with 2.5 arcsec seeing there is only a 1% correction necessary. Anyway, this is of a side matter... What I now use is the

s

button to scale everything to the best throughput.

But what I really want to know is how to reproduce the image plot of sensfunc when using

:graph sri

(having the sensitivity function, residuals and image plotted). Or maybe as well: how do I reproduce the flux calibrated spectrum of the standard star from the sensitivity function, when I apply a sensitivity function to the standard star using

calibrate

. I tried something like: flux = 10^(-0.4*sensfie), but this didn't work...

Hope someone can help me out here. Thanks!

---

Francisco Valdes wrote on Mar 14, 2006

Hi,

Yes it is common to find that the standard stars don't overlay each other when you start sensfunc. As you said this is caused by non-photometric conditions. The usual procedure is then to do what you did with the 's' key.

For the main part of your question I am not sure what it is you are asking. Could you try to explain further what it is you want?

Yours,
Frank Valdes

---

Eduard Westra wrote on Mar 14, 2006

valdes

Yes it is common to find that the standard stars don't overlay each other when you start sensfunc. As you said this is caused by non-photometric conditions. The usual procedure is then to do what you did with the 's' key.

Yeah, I discussed this too and this seemed the most logical answer. Thanks for that..

valdes

For the main part of your question I am not sure what it is you are asking. Could you try to explain further what it is you want?

What I am after here is the sensfunc file... The contents of this file (when plotted with onedspec.splot) gives wavelength vs a magnitude. How do I convert this magnitude to a flux density and in what units is it? I tried before to guess that there is something of multiplying with a factor of -0.4 and raising this to the power of 10.

If I apply this to my extinction corrected standard star (instead of onedspec.calibrate) then I should get exactly the same as applying the sensitivity function with onedspec.calibrate.

---

Francisco Valdes wrote on Mar 14, 2006

What I am after here is the sensfunc file... The contents of this file (when plotted with onedspec.splot) gives wavelength vs a magnitude. How do I convert this magnitude to a flux density and in what units is it? I tried before to guess that there is something of multiplying with a factor of -0.4 and raising this to the power of 10.

If I apply this to my extinction corrected standard star (instead of onedspec.calibrate) then I should get exactly the same as applying the sensitivity function with onedspec.calibrate.

The spectrum is in magnitude units as described in the help for "calibrate:

When applying a flux calibration the spectra are divided by the
aperture sensitivity which is represented by a spectrum produced by
the task sensfunc. The sensitivity spectrum is in units of:

2.5 * Log10 [counts/sec/Ang / ergs/cm2/sec/Ang].

To apply to data that is in counts do

newval = oldval * 10^(-0.4*sensval)

where sensval is the value from the sensitivity spectrum. In calibrate there is a little more happening. The sensitivity spectrum is fit by interpolator because the spectrum being calibrated may not have the exact same wavelength sampling. Also the the calibration can be changed to Fnu and an extinction correction can be applied.

I hope this helps.

Frank

---

Eduard Westra wrote on Mar 14, 2006

valdes

[quote:4dba92e80a]
When applying a flux calibration the spectra are divided by the
aperture sensitivity which is represented by a spectrum produced by
the task sensfunc. The sensitivity spectrum is in units of:

2.5 * Log10 [counts/sec/Ang / ergs/cm2/sec/Ang].

To apply to data that is in counts do

newval = oldval * 10^(-0.4*sensval)

where sensval is the value from the sensitivity spectrum. In calibrate there is a little more happening. The sensitivity spectrum is fit by interpolator because the spectrum being calibrated may not have the exact same wavelength sampling. Also the the calibration can be changed to Fnu and an extinction correction can be applied.

I hope this helps.

It does! I think that I might have forgotten about the counts/sec/*ANG*... I thought it was counts/sec and that's it... I'll have a look at it!

Thanks very much!!

---

Last post on Mar 14, 2006