What would you like to see in Demetra ?

Re: What would you like to see in Demetra ?

Postby Robin Leadbeater » Wed Apr 29, 2020 11:50 am

I do not use Demetra yet because I need a fully flexible wavelength calibration like the file mode in ISIS (for the LHIRES and modified ALPY 200) and I do not like to use different programs but I am watching closely :-)

ISIS has an optimised binning feature which is useful for faint objects. Does Demetra have this ?

One feature which is missing in all amateur software is a calculation of signal/noise ratio. (Except for Spc Audace I think which has a calculation based on variability between adjacent bins) This is very important in professional work (many fits files of professional spectra have both the spectrum and the SNR profile by bin) This could be calculated for example from the sigma variability in each bin between individual profiles in the sub exposures for example (similar to the method used in PlotSpec)

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
 
Posts: 1481
Joined: Mon Sep 26, 2011 4:41 pm

Re: What would you like to see in Demetra ?

Postby Francois Cochard » Mon May 11, 2020 8:24 am

Hi Robin,

Several good points...

- Regarding the flexible calibration, we have it in Demetra! You can make your own "reference lines list" - but the advantage is that the lines are detected automatically. Well... this is probably something we could show you in live. It could be a good opportunity to have an "open workshop" with Zoom - would you be interested?

- Yes, the binning algorithm of Demetra is the same as ISIS (thank you, Christian ! ;>)

- The question of SNR is tricky. So far, I've not seen a universal method for calculating the SNR - I mean working for any kind of spectrum. A Ha high resolution emission profile (Be star) cannot be measured the same way as a low resolution VIS spectrum. If you have the magic reciepe, please tell me ! Anyway, we'll add very probably something, because I agree this is an important information. The "something" will be probably different calculations, like in ISIS. The method you mention (PlotSpec) could be one of these method.

Friendly,

François
Francois Cochard
 
Posts: 35
Joined: Fri Sep 30, 2011 2:30 pm

Re: What would you like to see in Demetra ?

Postby Robin Leadbeater » Fri May 15, 2020 12:56 pm

Hello Francois,

Francois Cochard wrote:
- The question of SNR is tricky. So far, I've not seen a universal method for calculating the SNR - I mean working for any kind of spectrum. A Ha high resolution emission profile (Be star) cannot be measured the same way as a low resolution VIS spectrum. If you have the magic reciepe, please tell me !


Yes all methods which try to measure the uncertainty based on the final profile have problems but we can do much better if we use the data available from the individual sub exposures during the reduction. (I used similar techniques for analysing variability on paper machines which have cross machine profiles which vary with time). The difference between the individual sub exposures has everything we need to estimate the uncertainty in any spectrum and analyse the different origins of the variability (both stochastic and systematic). This information is normally ignored and thrown away after the final spectrum has been produced but can be found in the individual profiles produced by ISIS for example. The method used in Plotspec is a simple example of this but it is possible to refine this to reveal much more information about the variability.

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
 
Posts: 1481
Joined: Mon Sep 26, 2011 4:41 pm

Previous

Return to Software Demetra

Who is online

Users browsing this forum: No registered users and 1 guest

x