Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Lots of noise at red end of spectrum?

  • Please log in to reply
8 replies to this topic

#1 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1264
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 21 October 2019 - 07:13 PM

Hello All,
I had a great night last night. I recreated an IR curve using Vega and captured the spectrum for 6 stars. I was able to get a very close match for 5. But I had two problems:

 

1)  I seem to have a ton of noise at the red-end of many of the spectra. I was capturing data in SharpCap and then transferring a single frame to RSpec. I think it is because I had the exposure was too low and consequently had to crank up the gain too much but it seems that the extra noise was much worse at the longer wavelengths. Is there an explanation for this? I think my spectra were better after I realized this and corrected it but not quite sure. Please see the image of Scheat.

 

2) While 5 out of 6 spectra were very close to the reference, my spectrum of DenebAlOkab was way off.  It is supposed to be a K2 star but my spectrum is much closer to an A or F. I thought I must have simply captured the wrong star but it looked like I was aiming correctly???

 

For all, I captured with my Evolution 8 at f10 with a 178 mono camera and the SA-100. I used a spacer to get to 3.65 A/pix.

 

Thanks for any hints.

 

 

Attached Thumbnails

  • Scheat.png
  • DenbAlOkab.png


#2 Brett Waller

Brett Waller

    Mariner 2

  • *****
  • Posts: 279
  • Joined: 18 Aug 2006
  • Loc: Staunton, VA

Posted 21 October 2019 - 08:54 PM

The quantum efficiency of the IMX178 chip drops below 20% and wavelengths greater than 8000A, so in that region, you might expect the SNR to drop and the apparent noise level to noticeably increase. By 9000A, the QE is below 10% which would double the apparent noise.  I suspect, this may be the nature of the problem. In addition, you may have contributing issues from the reflectivity of the mirror coating and/or absorption is the glass elements.

 

I don't think you provided enough information for me to determine what the problem was with the spectrum of Deneb al Okab. I might suggest collecting data from stars on both sides of K2 to see if you can notice any trends that might help you unravel the problem.

 

Brett



#3 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1264
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 21 October 2019 - 09:02 PM

The quantum efficiency of the IMX178 chip drops below 20% and wavelengths greater than 8000A, so in that region, you might expect the SNR to drop and the apparent noise level to noticeably increase. By 9000A, the QE is below 10% which would double the apparent noise.  I suspect, this may be the nature of the problem. In addition, you may have contributing issues from the reflectivity of the mirror coating and/or absorption is the glass elements.

 

I don't think you provided enough information for me to determine what the problem was with the spectrum of Deneb al Okab. I might suggest collecting data from stars on both sides of K2 to see if you can notice any trends that might help you unravel the problem.

 

Brett

Hey Brett,

Thanks for the explanation. That makes alot of sense. I am guessing that I just captured the wrong star for Deneb al Okab. I will try it again next clear night and see if the same thing happens.



#4 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 745
  • Joined: 18 Dec 2005

Posted 22 October 2019 - 05:23 PM

Hi Dave,

 

Yes the noise will follow the inverse of the instrument response.(not just the camera remember, the grating also has a significant response, optimised by blazing for the middle of the visible spectrum). Where the response is low (at the blue and red ends) the instrument response correction effectively has to turn up the gain in those regions so the noise get worse.  You can see an example of this here on my website

http://www.threehill.../spectra_30.htm

(The repeat spectra 2/3 down the page)

 

Also are you are taking just a single 8 bit video frame? If so you only have 256 levels to play with so when you multiply these up through the instrument response correction it generates digitisation noise.  It helps with both these issues if you can take several well exposed exposures and stack them to produce a 16 bit fit file.

 

Yes it looks like the last one is not the star you thought it was. (The otherwise featureless blue continuum with Balmer lines show it is a much hotter star)

 

Cheers

Robin


Edited by robin_astro, 22 October 2019 - 05:37 PM.


#5 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 745
  • Joined: 18 Dec 2005

Posted 22 October 2019 - 05:39 PM

Dave,

 

I had to look up Deneb el Okab :-)

 

According to Wikipedia it can refer to either Zeta Aquila (A0v) or Epsilon Aquilla (K1iii)  Perhaps this explains the confusion ?

 

Robin



#6 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1264
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 22 October 2019 - 05:52 PM

Hi Dave,

 

Yes the noise will follow the inverse of the instrument response.(not just the camera remember, the grating also has a significant response, optimised by blazing for the middle of the visible spectrum). Where the response is low (at the blue and red ends) the instrument response correction effectively has to turn up the gain in those regions so the noise get worse.  You can see an example of this here on my website

http://www.threehill.../spectra_30.htm

(The repeat spectra 2/3 down the page)

 

Also are you are taking just a single 8 bit video frame? If so you only have 256 levels to play with so when you multiply these up through the instrument response correction it generates digitisation noise.  It helps with both these issues if you can take several well exposed exposures and stack them to produce a 16 bit fit file.

 

 

Thanks Robin. Very good to know. Yes, I am using 16 bits but to a png file. I haven't attempted to stack yet but that is on my list.



#7 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 745
  • Joined: 18 Dec 2005

Posted 22 October 2019 - 05:52 PM

 

the grating also has a significant response, optimised by blazing for the middle of the visible spectrum). 

You can see the characteristic shape of a blazed grating here for example 

http://astro.if.ufrg...redes/chap9.htm

It looks similar to the typical camera senstivity but this is a coincidence. The camera and grating response multiply together to give an instrument response curve which tails off even steeper at the blue an red ends. Add to that the atmosphere which absorbs at the blue end and it starts to become really tough to get a good signal/noise ratio at the blue end compared with the middle wavelengths

 

Cheers

Robin



#8 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1264
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 22 October 2019 - 05:55 PM

Dave,

 

I had to look up Deneb el Okab :-)

 

According to Wikipedia it can refer to either Zeta Aquila (A0v) or Epsilon Aquilla (K1iii)  Perhaps this explains the confusion ?

 

Robin

Yep, that is it! I just saw the K1iii star and didn't look any further. My spectrum definitely looks like a A-type star. Very cool. I couldn't figure out how I got the wrong one...


  • robin_astro likes this

#9 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 745
  • Joined: 18 Dec 2005

Posted 22 October 2019 - 05:59 PM

Yep, that is it! I just saw the K1iii star and didn't look any further. My spectrum definitely looks like a A-type star. Very cool. I couldn't figure out how I got the wrong one...

That's a good example of the power of spectroscopy !  




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics