First light with Star Analyser 100!

[tooth_dr]

It was really cloudy, and the only star I could really see was Betelgeuse.  The first problem I encountered was the star and spectrum not fitting on the chip at the same time.  I took two separate frames and combined them PS (this could be the problem)

 

I've attached the graph, and the image I took.  I was using the 10", 1200mm FL, I will try again with the ED80 at 600mm.  I should have started with an A-class star from what I can now see!  M-class not ideal lol!  I was using the trial version of R-Spec.

 

Any input welcomed !

 

[Paz]

Wow that looks interesting. Can you tell different class stars visually or is it only through computer analysis that the different types show up?

[andrew s]

Hi, Looks a bit strange to me see here for an example http://iainpetrie.typepad.com/the_four_ages_of_sand/2011/01/the-joy-of-spectroscopy-finally-first-ever-spectrum-of-star-light.html

It is best if you get the spectrum horizontal on the chip and can change the length of the spectrum by varying the grating chip spacing. I think the R-Spec site has a calculator.

Nice to get your first spectrum under your belt.

Regards Andrew

[andrew s]

16 minutes ago, Paz said:

Wow that looks interesting. Can you tell different class stars visually or is it only through computer analysis that the different types show up?

You can see the broad spectral classes by eye but you need some analysis to get a the finer points of the classification. This is an introduction http://astro.unl.edu/naap/hr/hr_background1.html

Regards Andrew

[Gasman]

You really need the star and the spectrum in the same image . They both need to be on the chip to get the calibration , the star is at position zero on the x axis as you’ve shown. Might have to move the SA nearer the chip and yes better to start with an A type star like Vega which has very clear emission lines!

HTH

Steve

[Merlin66]

Keeping the star image and spectrum horizontal across the chip introduces less artifacts.

onwards and upwards.

[tooth_dr]

Thanks guys.  I’ll shorten the nosepiece a bit in that case see if it helps. 

[Glasspusher]

Congratulations on your first spectrum....glad you decided to have ago! Be careful you might get hooked.

John

[ejp1684]

Congratulation! My StarAnalyser still gets a lot of use even though I'm moving on to higher resolutions. It's a fascinating aspect of astronomy. Even on bright moonlit nights you can still capture spectra.

Zero image on left, spectra on right, and as level as you can get it.

Eric.

[robin_astro]

Congratulations on first light!  Which camera are you using ?  Looking at your equipment list one of the mono ones would be better as you get a much more even spectral response (The large humps in your spectrum are due to the differences in sensitivity of the different colour pixels and the fine sawtooth pattern is due to uneven coverage of the R,G,B pixels and the effect of the spectrum being diagonal) 

You should still be able to see visually the molecular bands in Betelgeuse though. I suggest trying a bit shorter exposure (under expose rather than risk overexposure)  You should at least be able to see something like this similar type star delta Virginis on my website here (Which is also a bit of history  as it is the first spectrum ever taken with a Star Analyser :-)  )

http://www.threehillsobservatory.co.uk/astro/spectra_14.htm

Cheers

Robin

[tooth_dr]

Thanks again everyone! I’m really keen to get out again now.

 

@robin_astro I will use my Skynix 2.0M (mono) next time, I just had the Skyris 618c handy so used it.  Will shortening my nosepiece adapter, bringing the grating closet to the sensor, help for it all in?

I used a 2s exposure with a low gain. What would you suggest? I’m pretty sure it’s over-exposed and smudged the image.

Thanks for the link, interesting that it is your first spectrum too! I’ll have a read through it. 

[robin_astro]

48 minutes ago, tooth_dr said:

 I will use my Skynix 2.0M (mono) next time, I just had the Skyris 618c handy so used it.  Will shortening my nosepiece adapter, bringing the grating closet to the sensor, help for it all in?

I used a 2s exposure with a low gain. What would you suggest? I’m pretty sure it’s over-exposed and smudged the image.

Thanks for the link, interesting that it is your first spectrum too! I’ll have a read through it. 

Not my first spectrum but the first spectrum ever taken by anyone using Star Analyser (I developed the Star Analyser and persuaded Paton Hawksley to manufacture it. I took the spectrum using the production prototype.)

Yes the closer the grating, the shorter the spectrum. Too short though and you start to lose resolution. You can use the calculator here to work out the best spacing 

http://www.patonhawksley.co.uk/calculator/

2 sec is rather long for such a bright star. The sharpness of the spectrum depends on good focus and seeing so on bright objects you can use the same techniques as used for planetary imaging. (short exposures, selecting, aligning and stacking the best. No post processing/sharpening though)  Focus is critical and best done first on a main sequence A type star like Delta Cas which show clear Balmer lines or Be stars like Gamma Cas which shows clear H alpha emission. Examples of what they should look like here  (recorded for the Sky at Night programme back in 2005 the week before I pitched the idea of the Star Analyser to Paton Hawksley)

http://www.threehillsobservatory.co.uk/astro/spectra_12.htm

Cheers

Robin

 

[robin_astro]

31 minutes ago, robin_astro said:

the closer the grating, the shorter the spectrum. Too short though and you start to lose resolution. You can use the calculator here to work out the best spacing 

http://www.patonhawksley.co.uk/calculator/

Looking at your image I dont think you will have any problem fitting the useful part of the spectrum horizontally in the field. The white/purple part of the spectrum is in the Infra-Red which is not very interesting so it does not matter if that part is off the edge of the field of view (All the RGB filters on the pixels become transparent in the IR so the spectrum appears approximately white.) 

Robin

[tooth_dr]

@robin_astro Thanks Robin, your explanations have really helped me.  I just tried to get it all in on the screen.  I've learnt so much about the practical aspects of spectroscopy today.

 

I'm really lucky to be going to see Prof Carl Murray talk about Cassini-Huygens this evening, and it gives clear skies from 10pm onwards so I'll take all the advice on board and give it another go tonight.

[tooth_dr]

Taking all the advice onboard, I have tried to image horizontally, I used a mono camera, and I under-exposed.

I cant seem to get the spectrum to stack through...

 

Here is Gamma Ursae Majoris from last night, feedback welcomed:

 

 

 

 

 

 

[andrew s]

That looks fine to me, much improved. While you don't want to over expose I.e. saturate the spectrum, there is no harm in getting it as high as you can. 

Regards Andrew 

[tooth_dr]

29 minutes ago, andrew s said:

That looks fine to me, much improved. While you don't want to over expose I.e. saturate the spectrum, there is no harm in getting it as high as you can. 

Regards Andrew 

Thanks Andrew.  I didn't read up on exposure settings.  I varied the exposure time and gain settings until I could make out absorption lines on the monitor, and recorded at that.  There is probably a scientific was to do this?!

[tooth_dr]

I’ve now realised that I was wrong to use a Bhatinov mask to get the star in focus as this means the spectrum is not!

[Merlin66]

Yes, focus on the spectrum not the zero order star image......

It does get easier with practise ;-)

[tooth_dr]

Thanks @Merlin66 It's very rewarding trying this out.  The last night I was out, I had about an hour at it, then it started raining.  But I was satisfied I had a productive night.  If I had have been doing deep sky imaging, I'll have barely got setup, and  taking the first sub, and would have had to abandon it!  Plus I'm learning a bit of science about the stars, bringing back knowledge I learnt on the open university course I complete a few years ago

[robin_astro]

That's looking good but I am a bit surprised how fast the spectrum is dropping off at the red end. I would expect a more gradual drop with the spectrum continuing to at least 7500A, showing  H alpha and the Telluric bands more clearly. From the plot it looks like the low brightness values are being clipped to zero. (We need to be able to see right down to the sky background) Is this the raw spectrum without background subtraction?  

Robin

[robin_astro]

6 minutes ago, robin_astro said:

That's looking good but I am a bit surprised how fast the spectrum is dropping off at the red end. I would expect a more gradual drop with the spectrum continuing to at least 7500A, showing  H alpha and the Telluric bands more clearly. From the plot it looks like the low brightness values are being clipped to zero.

More like these  for example (The blue plot)

http://www.threehillsobservatory.co.uk/astro/spectra_5.htm

http://www.threehillsobservatory.co.uk/astro/spectra_12.htm

[tooth_dr]

17 minutes ago, robin_astro said:

That's looking good but I am a bit surprised how fast the spectrum is dropping off at the red end. I would expect a more gradual drop with the spectrum continuing to at least 7500A, showing  H alpha and the Telluric bands more clearly. From the plot it looks like the low brightness values are being clipped to zero. (We need to be able to see right down to the sky background) Is this the raw spectrum without background subtraction?  

Robin

I clicked the background subtraction box, can try again without it ticked!

My Lumenera LU070 mono camera has a glass IR filter in front of the sensor. Should I remove it?

 

Thsnks for your help.

Adam. 

[robin_astro]

Hi Adam,

yes remove the IR block to give you better sensitivity at the far red (and possibly violet) end.

The spectrum and sky background regions may not be well set up by default.  I forget exactly how it is done in Rspec but you can manually set up the zones that the spectrum and sky background are measured in. It is important not to get any cross contamination. You need to include the full height of  the spectrum in the spectrum binning region and only the sky background in the sky subtraction regions.  Turn up the brightness in the image to see this more clearly. You can see an example of the selected regions here using ISIS for example. There should be something similar in RSpec  (A faint star in a crowded field, yours will be easier !) 

Cheers

Robin

[tooth_dr]

Hi Robin, I understand that all now, thanks for explaining it clearly. I had a look at ISIS earlier.  I'll try to nail the focus and get a few simple stars under my belt next time it's clear.

 

Thanks again

Adam.