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

FWHM in arcsec is normalized?

  • Please log in to reply
7 replies to this topic

#1 sbharrat

sbharrat

    Apollo

  • -----
  • topic starter
  • Posts: 1,209
  • Joined: 28 Nov 2020
  • Loc: NJ, USA

Posted 25 November 2021 - 10:18 AM

On various threads about FWHM, I have seen comments to effect that FWHM in arc-sec is a normalized form that allows comparison among different systems. In particular, it made possible certain reasoning about 5" FWHM being "bad" and 2" FWHM being "good" etc. (Yes, I understand this all depends on the seeing.) With this in hand, I have learned that I average about 3.5" FWHM. This is with 1.61"/px image scale, and about 3"/px guide scale. 

 

I recently changed from flattener to 0.79x reducer for some targets. I assumed that since I was doing FWHM in arc-sec, and my mount and guiding setup had not changed, I should expect the same FWHM. But I have enough nights now to know that it is not. I consistently average around 4.7 which is suspiciously close to my original scaled by the change in image scale. (The new image scale is 2.03"/px so not very undersampled.)

 

Am I reasoning about this correctly? Should I expect the same FWHM in this situation where guiding/mount hasn't changed?



#2 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 9,814
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 25 November 2021 - 10:55 AM

A couple of things.

 

I find when auto focusing that I get  down to around 2 pixels no matter which one of my refractors I use. So, with that in mind as my focal length shortens, I get a different average FWHM. I don't use reducers and, of course, the scopes are different but it's been that way for a while. When I autofocus I get down to around 2 pixels on the shorter scopes and around 3 pixels on the long (1100mm) scope.

 

I noticed when I changed cameras from my ASI071  to a QHY268 with smaller pixels the FWHM in arc seconds went down on all three scopes. 

 

Having recently been told by a master optician that I know nothing about optics I won't speculate as to the cause, but I think you're seeing the same thing. When I measure individual stars in the images, they are much closer. The medium focal length scope has plenty of 1.5 FWHM stars. When you use most analysis software you're seeing an average. 

 

Rgrds-Ross


  • sbharrat likes this

#3 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,766
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 25 November 2021 - 12:17 PM

On various threads about FWHM, I have seen comments to effect that FWHM in arc-sec is a normalized form that allows comparison among different systems. In particular, it made possible certain reasoning about 5" FWHM being "bad" and 2" FWHM being "good" etc. (Yes, I understand this all depends on the seeing.) With this in hand, I have learned that I average about 3.5" FWHM. This is with 1.61"/px image scale, and about 3"/px guide scale. 

 

I recently changed from flattener to 0.79x reducer for some targets. I assumed that since I was doing FWHM in arc-sec, and my mount and guiding setup had not changed, I should expect the same FWHM. But I have enough nights now to know that it is not. I consistently average around 4.7 which is suspiciously close to my original scaled by the change in image scale. (The new image scale is 2.03"/px so not very undersampled.)

 

Am I reasoning about this correctly? Should I expect the same FWHM in this situation where guiding/mount hasn't changed?

What software are you using to measure FWHM?  Is it reporting in pixels or arc seconds?  When you changed to the reducer, did you change the settings in the software so it knows the new pixel scale, or effective focal length?

 

-Dan


  • ks__observer and sbharrat like this

#4 sbharrat

sbharrat

    Apollo

  • -----
  • topic starter
  • Posts: 1,209
  • Joined: 28 Nov 2020
  • Loc: NJ, USA

Posted 25 November 2021 - 02:02 PM

What software are you using to measure FWHM?  Is it reporting in pixels or arc seconds?  When you changed to the reducer, did you change the settings in the software so it knows the new pixel scale, or effective focal length?

 

-Dan

PixInsight. Yes, I have change the image scale (what SFS uses) to reflect the reducer. 


  • ks__observer likes this

#5 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,766
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 25 November 2021 - 02:17 PM

Hmmm ... i was just watching a video on this yesterday.  The presenter, who seemed very knowledgable and used lots of optical engineering formulae (above my pay grade :-), showed that FWHM changes relative to f/ratio.  Apparently, an f/10 scope should produce a larger FWHM compared to, say an f/5 scope, all else equal.  

 

Since your reducer lowered your f/ratio, it would seem like your FWHM should get smaller, but it did the opposite.  Not sure what's going on.

 

-Dan


  • sbharrat likes this

#6 imtl

imtl

    Skylab

  • *****
  • Moderators
  • Posts: 4,270
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 25 November 2021 - 02:53 PM

Why do you think there is anything wrong?

 

In a very crude way you can estimate the FWHM as SQRT(Seeing^2 + Dawes^2 + GuideRMS^2 + ImageScale^2)

 

In other words, when you have a spot (star blurred by seeing and guiding errors) in the focal plane, you sample that spot with your sensor array. Your sampling will effect the further blurring of the spot because your sensor has a finite pixel size. Pixel size matters or if you want, image scale.


  • sbharrat likes this

#7 sbharrat

sbharrat

    Apollo

  • -----
  • topic starter
  • Posts: 1,209
  • Joined: 28 Nov 2020
  • Loc: NJ, USA

Posted 26 November 2021 - 10:24 AM

Why do you think there is anything wrong?

 

In a very crude way you can estimate the FWHM as SQRT(Seeing^2 + Dawes^2 + GuideRMS^2 + ImageScale^2)

 

In other words, when you have a spot (star blurred by seeing and guiding errors) in the focal plane, you sample that spot with your sensor array. Your sampling will effect the further blurring of the spot because your sensor has a finite pixel size. Pixel size matters or if you want, image scale.

Ok. Thanks for the crude formula. I don't completely understand it because the units of ImageScale are "/px. But if makes some sense the imagescale sets some "lower bound" on FWHM. 

 

So then my thinking about what a "normalized" measurement means is completely off. I *thought* that FWHM in arc-sec would be invariant to differences in image scale in particular. This is the only way for statements like "with average seeing, FWHM above 5" is bad and 2" is good". (Which I have seen multiple times on this forum.) You might as well just keep FWHM in pixels knowing that you can't compare it to what anyone else is reporting. 



#8 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 9,814
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 26 November 2021 - 11:51 AM

The size of the airy disk gets smaller as the F number decreases. Effectively fast scopes certeris paribus have better resolution than slow scopes. Here's a link. 

 

However, it you look at the formula imtl put up, it's obvious that seeing will dominate with any well guided telescope larger than about 4".

 

I wonder where the formula comes from but it's totally consistent with what I've seen over the years. 

 

I use 5 telescopes 2.8", 5", 12.5" and 24" in diameter. If I were to graph FWHM in arc seconds with those scopes it matches the formula pretty closely. All of the systems that carry those scopes have excellent guiding - under .5 arc seconds RMS for long exposures. They are within 100 feet of each other. The key variants on a given night are the seeing and the image scale.

 

The 24 and 12.5 are both F8 and back to back it's all just seeing related - results in terms of FWHM (at similar image scales) are essentially whatever the seeing allows. 

 

Rgrds-Ross


  • Midnight Dan likes this


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