Star Analyser 100 efforts

[ejp1684]

Having played around with Star Analyser 100 and BASS Project software for a while, I decided it was time to do some serious work. I'd had trouble using a calibrated spectrum for other spectra, and I couldn't for the life of me get the instrument response to work. So a few nights ago I set everything up and fortunately the clouds were kind to me for a couple of hours. Once I got going I captured a dozen or so spectra in the space of one hour.

I then set to work calibrating using Vega. This is what I ended up with. And although I say it myself, I was quite pleased with it. But the next question was, could I use the calibration and instrument response curve on the other spectra with suitable results? For the calibration fit I used the Cubic option, it seemed to give the smallest errors.

 

Here's Alpheratz, and to my amazement using the calibration obtained from Vega and the image response curve seemed to work, with suitable alignment between peaks and troughs. I had to use some y-axis adjustment to get it to look right, and Lamda Offset as the spectra were in slightly different positions on the images.

 

I can't see all that much of the sky from my back garden, so I just took spectra of all the brighter stars I could see. Here are a few more examples.

 

Albireo 1 was exciting, the first time I'd done a K3 star.

 

Because I'd exposed for Albireo 1, Albireo 2 came out almost invisible, but I thought I'd have a crack at it anyway. It's a bit noisy, but the main H lines are clearly visible.

 

Breaking new ground with an F2 type.

 

This was the worst one, I just couldn't get it to align any better, but the peaks and troughs all line up - more or less.

 

Of course the forecast now looks dismal for most of the week, but hopefully I'll get a few more soon. Now that I've sussed out how to process the images, I'll have to read the BASS manual and my spectroscopy books and learn to interpret the information.

One thing about spectroscopy, processing the images gives you plenty to do if you can't get out under a clear night sky.

Eric.

[GuLinux]

Very nice job.. Did you use single images, or stacked?

[ejp1684]

Just single. I'll have a go at stacking them for less bright stars.

Eric.

[robin_astro]

Hi Eric,

That is an excellent start! you are already getting results than many Star Analyser users would be envious of.

 As you have found out, you can indeed use an instrument response, normally obtained using a hot A or B star, to correct any spectrum. If you want to get even closer though, there are a few extra factors you can take into consideration.

The instrument response includes the atmosphere as well as the instrumental effects.  The effect of  atmospheric extinction depends on wavelength (more absorption at the blue end of the spectrum) and varies with the altitude of the target (more air mass to travel through at lower altitude) and from day to day.  To reduce these atmospheric effects  it is best to chose a reference  star at a  similar altitude to your target and measure it on the same night.

All stars have subtly different spectra (including the shape of the continuum),  even if they designated as having the same classification so using a generic spectrum like  the Pickles spectra in BAAS, Vpec, RSpec etc as a comparison will only be approximate. (There are various reasons for the differences like interstellar extinction, the metalicity of the star, a variable star which changes its spectrum, or even an incorrectly catalogued spectral type)  To improve on this, you can use comparison stars which have actually measured reliable spectra. The MILES stars for example are often used by amateurs for low resolution spectra.  

There are also a couple of processing steps which you need to be careful of, particularly with  slitless spectrographs like the Star Analyser:-

Because the thickess of the spectrum varies along the spectrum due to focus , make sure you include all the rows with spectroscopic data otherwise you could lose some of the signal at places where the spectrum is wider which will distort the shape.

Take great care with background subtraction (not so much a problem with bright targets but can be problematic with faint targets) small errors in background subtraction can make a big error at the extreme red and blue ends where the spectrograph is less sensitive and so signal is low.  Keep the background zones close to the spectrum but make sure you have not got any spectrum spilling into the background zones and avoid as far as possible catching other spectra/star images in the background zones)

Cheers

Robin

[ejp1684]

Thanks Robin. Appreciate your advice and will make sure I keep to the points you made. Where do I get the MILES stars from, and can they be incorporated into BASS?

Eric.

[robin_astro]

Hi Eric,

I'm an ISIS user which has a subset of them integrated into its database but I'm not sure if they are integrated into BASS. You can get a nice spreadsheet though developed by Paolo Beradi and Marco Leonardi from the link given in this thread on the ARAS group

http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=941

You can use search criteria to find the most suitable one. There is not always one close to where you need it though so some planning is needed to catch one at the right elevation.

If you are interested in the history of amateurs starting using MILES stars you read about  it on  this thread

http://www.spectro-aras.com/forum/viewtopic.php?f=6&t=207

 

Alternatively, Francois Teyssier has generated a similar spreadsheet to select what should be typical stars from the large number of  stars with catalogued spectral types  to use with the Pickles standards which makes it easier to find one in the right part of the sky to match the target but there are no guarantees that the star spectrum is exactly the same as the generic one in the Pickles library.

http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=1227

Much of this is rather over the top though for the Star Analyser which is mainly about having fun and getting your feet wet in spectroscopy so dont worry about it too much.  What you are doing at the moment is certainly close enough to see the characteristics of the objects you are looking at. The rest is just finessing the results. Have fun with the Star Analyser and come back to this if you perhaps move on to more serious work with a slit spectrograph  

Cheers

Robin

[ejp1684]

Thanks again, Robin. Got a clearer picture now of how to approach it.

Eric.

[Merlin66]

The Miles catalogue should be available for download from the BASS site.

It works well with BASS.

 

Finding suitable Atype stars....

I found a neat way using CdC. By changing the default color palette for stars you can highlight A type stars - makes it easier to find them.

 

[ejp1684]

7 hours ago, Merlin66 said:

Finding suitable Atype stars....

I found a neat way using CdC. By changing the default color palette for stars you can highlight A type stars - makes it easier to find them.

Thanks for this, great idea!

Eric.

[robin_astro]

17 hours ago, Merlin66 said:

The Miles catalogue should be available for download from the BASS site.

It works well with BASS.

Hi Ken,

Can you please confirm/check that the non dereddened versions of the spectra are being used in BASS?  The prime use of the MILES database is to simulate the spectra of stellar populations (galaxies for example)  so for this the dereddened spectra are normally used (ie the spectra as they would appear without interstellar redenning). For our use of course we need the original spectra, including any redenning) . This caused problems initially when we were considering the use of the MILES database but is taken care of in the spreadsheet and ISIS

When choosing stars to use with generic databases like Pickles, take care to avoid stars with significant interstellar redenning (by checking B-V mag) as is done in Francois' spreadsheet

Cheers

Robin

[Merlin66]

Robin,

I'll check when I get home.

I'm pretty sure it includes the necessary reddening.

Ken

 

[Merlin66]

Robin, et al,

The MILES database I have for BASS Project shows both the original and de-reddening files.