Detecting and measuring faint point sources with a CCD

[Macavity]

In part answer to a recent question of mine (PDF Paper, 6 pages):

Detecting and measuring faint point sources with a CCD

by Herbert Raab, Astronomical Society of Linz.

Now, if I can find something similar for extended sources...

[FraserClarke]

If you wanna go the pro-way, try sextractor (which stands for 'S'ource 'Extractor' -- not some dodgy west country fetish... );

Terapix - SExtractor

'professional' front end ; i.e. not intuitive...

[Merlin66]

Craig Stark and others have also presented similar information.

The ability to accurately determine and apply all the variables to the SNR calculations, to me, is the problem. Adding multiple exposures, bias, darks and flats all add to the confusion.

A good analogy is mixing blue and yellow paint...what type of paint, how to mix it, temperatures etc etc - the end result will be green, but which green?? Light green, medium green, dark green. You really need a reliable end method of measurement. Same for SNR.

I've been investigating SNR in spectra, which is very similar to the problems of SNR in extended objects; we have the light from a star spread out along a band a few microns/pixels high and hundreds of microns/ pixels long. In that faint band there are bright lines, dark lines and a "background" continuum. We would like to accurately (?) determine the SNR for the spectrum. Taking into account all the various factors we may "estimate" a SNR say of 40; but is this correct???

The professionals have the same problem...

There's a paper on the DER_SNR method which claims to be a "standard" method for SNR measurement in spectra.

We have done some measurements of various spectra using different SNR methods and find if nothing else the DER_SNR is consistant and COULD be used as a reference point for comparison of different spectra. If you are interested in trying DER-SNR I can send you details of the required Python/ Numpy programs to run it under WinXp.

StandardSNR calcder_snr.pdf

[dph1nm]

If you wanna go the pro-way, try sextractor

If you wanted to do this I would recommend the GAIA image display tool, which is part of the Starlink package and has sextractor built in.

NigelM

[Macavity]

Belated thanks for the responses. In truth, I am probably unlikely to get too deeply into real measurements. But I did like the "paper" format... All that curve fitting and SNR reminded me of "bump" (mass resonance) hunting in particle physics. A bit of Nostalgia for me anyway?

I sense though, it is a good idea to postpone the effects of darks, flats etc. Otherwise these thing do indeed become rather convoluted?

A few ideas emerged strongly. Interestingly (obviously, in retrospect?) the width of faint sources is the same as that of bright sources. I think I now better understand the dependences of image quality? DSO images might need both a fast scope AND maximum aperture? I may not yet buy that small fast APO! I'm afraid, I might still persist with my Watec camera though - Even if it's not the best tool.

But, with bright images, and since you mention(ed) it (Merlin!), I may not able to resist having a go with a "Star Analyser 100" diffraction grating? Looks like there might be other and cool "measurement" things, as an amateur, I (one) can do? Might even have time to help beta-test software, if it's ever of help.

[Merlin66]

Spectroscopy is well overdue for some attention by amateurs.

Sure, we have a few dedicated souls doinf some excellent work... but we need an influx of new observers to build up the global numbers and get into the mainstream. The AAVSO is considering expanding its variable star observation from photometry to spectroscopy.

A Rainbow Optics/ Star Analyser grating is a great starting point and allows you to understand and better appreciate the basics. There are many freeware programs available to assist in the analysis of spectra - you don't need PhD in astro-physics!! - all you need is a desire to better understand this universe around us.

More than 85% of ALL knowledge we have on the stars, galaxies and universe was obtained via the spectroscope.....