Staranalyzer et polynôme...

[Robin Leadbeater]

The common reference is, in fact, the dispersion, which doesn’t change. I begin by building a polynomial using the Balmer lines from an A-type star.

The dispersion alone NOT sufficient to calibrate the spectrum of a slitless system. You are guessing at the wavelength of one of the features in the spectrum ie you need to know in advance something about the object you are measuring.

Robin

[Robin Leadbeater]

Jim,

This is just an education exercise that Etienne is running, not advanced use of the Star Analyser where assuming a feature in a known star type might be acceptable. Suggesting students have to guess a feature in advance in order to use the Star Analyser is a bad idea. I recommend you look at the work here Etienne has been doing with the ALPY on very faint extra galactic objects. Some of these are accessible with the Star Analyser. Assuming a wavelength there would be a complete nonsense.

Cheers
Robin

[James Ley]

The common reference is, in fact, the dispersion, which doesn’t change. I begin by building a polynomial using the Balmer lines from an A-type star.

The dispersion alone NOT sufficient to calibrate the spectrum of a slitless system. You are guessing at the wavelength of one of the features in the spectrum ie you need to know in advance something about the object you are measuring.

Robin

In my very limited experience with bright stars and planetary nebulae, I have found that advance knowledge is not necessary either to make a guess or to confirm it.

Very roughly, if I guess that the large absorption slightly blueward of the middle of the spectrum is H-beta and, unknown to me, if the star is O-type through F-type, then I am rewarded with all the Balmer and Ca II K and H lines of a reference spectrum aligning with my spectrum, confirming my guess. The Ca II H and K lines also make an easy guess if the pair is visible. Finally, the Na I D line, is often a large absorption in the middle of the spectrum.

I am not suggesting that such a method is appropriate for all users under all circumstances.

The educational aspect is also interesting. In half an hour I have taught many students who had never even seen a telescope before to calibrate a star analyzer spectrum using the zero order. It was easy. They didn't have to think about the absorption features or the spectral type of the star. They just had to click on the dot. Certainly there are lots of occasions where this would be desired. But perhaps, with appropriate "unknown" targets, students could learn a bit more with the above method where they had to look for what features they can find in the raw spectrum, see the consequences of guessing correctly and incorrectly and compare how different spectral types have characteristic absorptions.

[Robin Leadbeater]

The educational aspect is also interesting. In half an hour I have taught many students who had never even seen a telescope before to calibrate a star analyzer spectrum using the zero order. It was easy. They didn't have to think about the absorption features or the spectral type of the star. They just had to click on the dot. Certainly there are lots of occasions where this would be desired. But perhaps, with appropriate "unknown" targets, students could learn a bit more with the above method where they had to look for what features they can find in the raw spectrum, see the consequences of guessing correctly and incorrectly and compare how different spectral types have characteristic absorptions.

I think students actually learn more about spectroscopy and astrophysics by following the simple linear calibration procedure using the zero order. Before you go to the telescope, You start by explaining that a diffraction grating is a simple device that shifts the light along the spectrum by an amount proportional to its wavelength. (You can explain with a diagram the constructive and destructive interference of the waves depending on the angle if you want to, it is pretty simple ). You then give them that dispersion figure (or get them to calculate it (A/pixel) calculated from the grating l/mm and spacing) You then go to the telescope and get them to record a spectrum of say Vega. You get them to measure the distance of the absorption lines from the zero order (no fancy software needed) and calculate the wavelengths of the absorption lines. You then show them a calibrated spectrum of hydrogen measured in the lab and ask them what they have just found out for themselves about Vega without any prior knowledge. Real science !

Cheers
Robin