First light with a Star Analyzer 200 - Lots of filters

[Ags]

This is a quick test run of my star analyzer 200 with a bunch of filters - the scope is an FMA 135 and the camera is an ASI120MM. The Analyzer is about 17 mm from the chip and I just screwed various filters on the front of the FMA 135 and took 10 second snaps guided by my RC6. The star is Vega. The bottom line is the unflitered version

First observation is I question the optical flatness of my Wratten 29.

The Wratten 47 seems to pass a lot of UV, but sadly a fair amount of IR too.

I can see a few features in the Vega unfiltered spectrum, but but don't know enough yet to tell if they are from the atmosphere or the star.

I tried to stretch all the spectra the same, but it is very late and I got a bit muddled.

[PeterC65]

Very interesting. I like the idea of testing out your filters using this device.

I had no idea that the Star Analyzer even existed before I read this post so have just watched the video about it. I may need to get one. What made you get the 200 rather than the 100?

 

[Ags]

I bought it second hand; the SA200 was the first one to come up. I demounted an old TS visual UHC filter I haven’t used in a decade and used the empty housing as a spacer ring between the SA200 and the 120MM.

I later tried shooting some more subs of the Pac-Man with the FMA135 as guide camera, and interestingly the guiding still worked with the SA200 still in place. The sky gods punished me for my laziness and sent over some cloud.

[PeterC65]

Having watched some more videos, this time about the RSpec software that will analyse the images obtained from an SA100/200, it seems that I now need one of these diffraction gratings! What a great thing to be able to do alongside EAA / AP.

 

[Ags]

Are you going for the SA100 or 200?

[PeterC65]

I think I'm going for the SA100 as this version seems to be designed to work with the widest range of kit. It also bends the light less which should mean a better optical performance. The calculator on their website says it should be fine with my Skymax 127, Explorer 150 and Explorer 200. It won't fit in my filter wheel but I'm thinking that I'd fit it directly to the camera nosepiece and just do spectroscopy. I have a dual scope mount so could have an EAA imaging setup alongside.

And there I was, thinking that I didn't need any more astro kit!

 

[robin_astro]

Hi,

The guy who developed the Star Analyser here 🙂  (Almost 20 years ago now, note the correct (UK) spelling, it is manufactured by Paton Hawksley Education here in the UK. For people based here, buying direct is probably the cheapest option).  If you are interested in its history, I talked about it in the first part of this BAA meeting

https://britastro.org/videos/using-low-resolution-spectroscopy-to-confirm-supernova-discoveries-2

Yes if you have the space to mount it a suitable distance away from the sensor then the original SA100 is the better option. I developed the SA200 later for people who wanted to try it mounted in filter wheels, which are generally too close to the sensor for the SA100.  The SA200  also has a lower profile so it fits in most wheels. There is a page on my website explaining the difference

http://www.threehillsobservatory.co.uk/astro/spectroscopy_16.htm

Cheers

Robin

[robin_astro]

1 hour ago, PeterC65 said:

The calculator on their website

Just to clarify,  that website is for RSpec software, not specifically for the Star Analyser. Tom Field sells the Star Analysers in the US along with his software but there are many other resellers around the globe. (The calculations behind  the calculator are mine and Tom then built the web version and hosts it on his website). There are a lot of good tips and examples on its use there on the Rspec website though.   The website of the manufacturer of the Star Analyser is here

https://www.patonhawksley.com/

Cheers

Robin

 

 

[robin_astro]

You might also find these pages on Christian Buil's website of interest (He is probably the most experienced amateur spectroscopist in the world. The idea for the Star analyser originally came from what he was doing with simple gratings)  

http://www.astrosurf.com/buil/staranalyser/obs.htm

http://www.astrosurf.com/buil/staranalyser2/evaluation_en.htm

http://www.astrosurf.com/aras/staranalyser/userguide.htm

(The additional wedge prism he mentions there improves the resolution a bit and can be bought as an accessory but makes things more complicated so I generally recommend starting without one)

The rest of his website is also a goldmine of information on the subject

 

Cheers

Robin

 

[PeterC65]

@robin_astro I'm planning to fit the SA100 to the filter thread at the end of my cameras nosepiece. The camera is a Player One Uranus-C with 12.5mm of back focus and the nosepiece is 25mm long. I do also use a ZWO electronic filter wheel for EAA and I fit the camera to it via a Baader ultra low profile clamp, so with the filter wheel the filters are about 5mm further away from the sensor.

I assume that the SA200 is intended for filter wheel setups where the camera is screw fitted to the filter wheel, and therefore the filters are closer to the cameras sensor?

The Star Analyser together with the RSpec software sound like they would open up a whole new avenue for my EAA related observing. Normally I don't much bother with stars but this kit would make them a whole lot more interesting to me. One question that you may be able to answer ... I assume that when calibrating the spectrum curve from pixels to angstroms in RSpec that the calibration remains good as long as you keep the same sensor to grating distance, even if you change scopes?

 

[robin_astro]

You can start with pretty much any camera but once you get into it, you will find that a mono camera has many advantages over colour  if you have that option.

The spacing is not critical, you just need enough to spread the light out sufficiently to see the features (The spectrum is effectively a series of images of the star at different wavelengths so the smaller the star image and the longer the spectrum, the more detail you can potentially see. Make sure you include the zero order (star image) in the field though as that is the zero reference point for your calibration. (There is a limit to resolution though as the simple setup produces aberrations and field curvature which get worse as the spectrum gets longer and the spectrum gets dimmer. (spectroscopy needs a lot of light, exposures with the Star Analyser are typically 1-200x longer than for the same well exposed star).  The Star Analyser is really about learning about spectroscopy with a minimal outlay  so I encourage experimenting and if the bug bites you can move on to more sophisticated (and expensive) equipment.

Yes the dispersion only depends on the spacing but it is easy to measure, typically using a hot star  with clear Hydrogen Balmer lines. This can also be used to focus at the start of the night and used as a reference to correct the spectrum for the response of the instrument and the wavelength dependent absorption of  the atmosphere. If you are interested in a general overview of the whole process from measurement to final calibrated spectrum, David Boyd's presentation " Scientific Analysis of Amateur Spectra" top link on the BAA spectroscopy resources page here is a good one (with more sophisticated instruments but the principles are the same for all astronomical spectroscopy)

https://britastro.org/section_information_/equipment-and-techniques-section-overview/spectroscopy

and for a taste of what can be revealed about stars even with the simple Star Analyser here are some examples on my BAA page where I post some of my my interesting observations (mostly spectroscopy using various instruments) 

https://britastro.org/observations/observation.php?id=20201216_234948_8cabda965bfe692f

https://britastro.org/observations/observation.php?id=20210406_144443_9e1c6a4cf219d14d

https://britastro.org/observations/observation.php?id=20230523_183229_5116a1a27f78a1ea

 

Cheers

Robin

 

 

[Ags]

Now I am intrigued by processing spectrums, so I have started playing with this. Here is my Vega spectrum simply plotted by intensity.

Next step would be normalize the spectrum (create a synth flat for it) and then calibrate it (maybe autocalibrate by some kind of "plate solving" routine?).

[lunator]

HI Ags

The Vega spectra looks interesting, Lots of dips  😀

I have an SA100 and now I have a permanent set up I plan to get my head around taking spectra.

If I get any good at it I hope to be able to take spectra of bright doubles. 😀

Cheers 

Ian

[Ags]

Things to buy: ASI678MM - been meaning to buy this for Solar/Lunar imaging anyway. 

[Ags]

Ok, my app does input spectrum >> intensity graph >> denoised intensity graph.

[Ags]

Ok, now I do: input > raw intensity > denoised > flat > normalised.

[robin_astro]

3 hours ago, Ags said:

Now I am intrigued by processing spectrums, so I have started playing with this. Here is my Vega spectrum simply plotted by intensity.

Next step would be normalize the spectrum (create a synth flat for it) and then calibrate it (maybe autocalibrate by some kind of "plate solving" routine?).

You can normalise (rectify) the spectrum by dividing by the continuum but you lose the information about the continuum shape. This can be preserved by measuring the response of the instrument and atmosphere using a standard reference star and using this to response correct (calibrate in relative flux) the target spectrum.  The wavelength calibration can also be done using a reference star with obvious lines. This calibration can then be transferred to any unknown star using the zero order as a reference.

As well as David Boyd's presentation linked above, see my presentation " low resolution slitless spectroscopy" here as an example of this (slides 17-33)

https://britastro.org/document_folder/baa-document-store/past-talks/spectroscopy-workshop-at-the-nlo-oct-2015

and for more information on how to correct for instrument and atmosphere see this document on my website

http://www.threehillsobservatory.co.uk/astro/Relative_flux_calibration_20221222.pdf

Cheers

Robin

Edited August 25 by robin_astro
added range of relevant slides from presentation

[Ags]

No idea if the spectrum corresponds to reality, but here is my colorized spectrum:

[PeterC65]

2 hours ago, Ags said:

Things to buy: ASI678MM - been meaning to buy this for Solar/Lunar imaging anyway. 

This camera has been on my 'kit to buy list' for its improved sensitivity over my colour camera, but so far I've resisted as I'm not sure I would like monochrome. It's also occurred to me that it would work nicely for spectroscopy however.

 

[Ags]

The small pixels of the 678 would work well with my FMA135 and zero amp glow is an attractive feature. It would also work well with a reducer when imaging the Sun.

[vlaiv]

3 hours ago, Ags said:

No idea if the spectrum corresponds to reality, but here is my colorized spectrum:

That color is quite a bit away from actual spectrum colors (which are impossible to reproduce accurately on computer screen). Best you can do is try to pick color that is the most similar to visual spectrum colors - and then full spectrum looks like this:

This has been normalized by intensity but rendered by "human eye/brain response". You can see that far ends of spectrum are very dark - this is because we start lacking sensitivity in those parts of spectrum. You can get a feel for what those colors look like if you look at the rainbow - but do be careful, rainbows are superimposed on bright sky in the background and that shifts color somewhat - for comparison, here is recording of a rainbow:

(colors look somewhat brighter because of background sky brightness).

[Ags]

Yes, I know my rainbow is defective, it is a simple approximation using sine curves. Maybe I will just load the colorization from a bitmap.

[robin_astro]

11 hours ago, Ags said:

No idea if the spectrum corresponds to reality, but here is my colorized spectrum:

 

You can approximately colourise the spectrum (the various spectrum processing programs eg Visual Spec, RSpec, BASS project etc can so this for you) but you first need to calibrate your spectrum in wavelength. Your colours are way off, here is your spectrum with some lines identified.

 

The broad bands beyond H alpha are from our atmosphere and are in the IR so cannot be represented.  H alpha is in the deep red and H beta is at the boundary between green and blue. 

Calibrating Star Analyser spectra is straightforward. It is linear so you just need the zero order and one identified line. The dispersion in this image is ~7 Angstrom/pixel. As a guide to what lines are present in Vega take a look at Christian Buil's annotated atlas of the Vega spectrum bottom of the page here

http://www.astrosurf.com/buil/us/vatlas/vatlas.htm

The spectrum of a star like Vega (spectral type A0v) and many others can also be called up from a library (The Pickles library) which is included in various spectrum processing programs. Here is an A0v star in Visual Spec wit the Hydrogen Balmer lines marked in green

These programs  have  the tools you need to calibrate your spectra in wavelength and intensity and identify the lines.

 

Cheers

Robin

Edited August 26 by robin_astro

[robin_astro]

10 minutes ago, robin_astro said:

The spectrum of a star like Vega (spectral type A0v) and many others can also be called up from a library (The Pickles library) which is included in various spectrum processing programs. Here is an A0v star in Visual Spec

Here is the part of the spectrum covering the visible region,  rectified (normalised to the continuum) and colourised using Visual Spec 

[robin_astro]

You seem to have very low sensitivity in the blue/green part of the spectrum for some reason. With this mono camera, the H beta line would normally be near the highest point in the raw spectrum with several other hydrogen lines visible towards the blue end.  Something like this from my presentation I mentioned earlier in the thread

Do you have any filters in the system?  If not, one thing to check is that you are looking at the right spectrum. The grating produces a number of spectra either side of the zero order image. The spectrum you want is the  brightest of the two  nearest the zero order.  If you have not done yet you can also download the Star Analyser manuals from here where I have included some useful tips 

https://www.patonhawksley.com/resources

Cheers

Robin