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

Different F-Ratio scopes, same camera-How to find Exposure Time

  • Please log in to reply
14 replies to this topic

#1 carlosgib

carlosgib

    Sputnik

  • -----
  • topic starter
  • Posts: 40
  • Joined: 15 May 2013
  • Loc: Southern Spain

Posted 12 August 2020 - 03:41 AM

This is an interesting topic, and there are a lot of different opinions on the subject. One of the more often expressed opinions is that aperture doesn't really matter for astrophotography, and that only f-ratio really matters. I am personally of the opinion that both does matter. I read quite a bit in different forums, and the conversation take a very intensive options that the element of finding a solution seems to dissipate alone the way.

This is an hypothetical example, but one that I which to find a reasonable solution.
System A  : RASA F2.2, 11” (279.4mm), Focal Length 620mm, camera is a QHY (CMOS)268C with a 3.76um pixel size and 26megapixels
System B : Stellarvue 130mm, F7, Focal Length 920mm, same camera.
So, if we use the expression (F2.2/F7)^2 = nearly 10%

The f/2.2 scope will have the shorter exposure times.  The required exposure time will be 10% of what the f/7 scope requires for the same image brightness.

So, if system A took an image of NGC7023 with a total integration time of 11.6hours, and if we apply the above expression using system B the total integration time will be 104 hours which is unrealistic.

The other factor is the angular magnification of both scopes, one is 620mm while the other is 920mm, (920/620 = 1.48) , there is a 1.5x differences and the area of sky covered of 2.25 ((920/620) ^2 ) times larger for B scope, also I believe that we need to take into consideration that there is a differences in F-Ratio.

I am just pointing out the different factors that I need to take into consideration to be able to come with the correct exposure time, this is my objective, finding a formula that I can rely on, I understand that the observation location with be different from one person to another and as we know the skyglow is also an important factor.

Taking into consideration that I have an excellent Polar Alignment and mount and without entering into an intensive discussion what would be a reasonable formula to calculated the exposure time based my points above, or if I wanted to experiment if I see an image from a fellow astrophotographer using the same camera??

Thanks



#2 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 04:24 AM

AP is about collecting the most photons possible for a given integration.

SNR = sqrt(target photons)

F-ratio indicates the rate at which your system will collect photons.

Extended object comparison = (f1/f2)^2

Point Source (stars) = (f2/f1)^2 * (Aperture1/Ap2)^2

Camera exposure time: Look up swamp read noise on CN.


Edited by ks__observer, 12 August 2020 - 09:58 AM.

  • kathyastro likes this

#3 kathyastro

kathyastro

    Vanguard

  • *****
  • Posts: 2,174
  • Joined: 23 Dec 2016
  • Loc: Nova Scotia

Posted 12 August 2020 - 04:26 AM

So, if system A took an image of NGC7023 with a total integration time of 11.6hours, and if we apply the above expression using system B the total integration time will be 104 hours which is unrealistic.

Why is it unrealistic?  The reason people get a RASA is so they can take exposures that would be too long with another scope in a shorter time.  You would only take an 11.6 hour integration with the RASA if you were after the same kind of signal-to-noise ratio that you would get with 104 hours at f/7.  If you didn't need that kind of SNR, you wouldn't shoot such long exposures on either scope.


Edited by kathyastro, 12 August 2020 - 04:27 AM.

  • ks__observer likes this

#4 sharkmelley

sharkmelley

    Soyuz

  • *****
  • Posts: 3,583
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 12 August 2020 - 05:35 AM

It all depends on what you are trying to achieve.  If your aim is to get the best image of NGC7023, you might weigh up system A and system B in the following alternative ways:

  • System A has the larger aperture and will collect photons from NGC7023 4x quicker than system B.  This could be a huge advantage.
  • System B has the longer focal length which offers the possibility of greater resolution in your final image even though it takes 10x longer to achieve the same pixel level SNR.

Mark



#5 imtl

imtl

    Viking 1

  • *****
  • Posts: 820
  • Joined: 07 Jun 2016

Posted 12 August 2020 - 05:42 AM

AP is about collecting the most photons possible for a given integration.

For a given image scale.

 

Resolution is a big part of AP as well. If that was not the case then we would all be using a Canon/Nikon F/2 lenses with no need for a telescope. 

 

So I do agree with you about what you wrote but it is not the complete picture.

 

Eyal



#6 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 07:13 AM

For a given image scale.

 

Resolution is a big part of AP as well. If that was not the case then we would all be using a Canon/Nikon F/2 lenses with no need for a telescope. 

 

So I do agree with you about what you wrote but it is not the complete picture.

 

Eyal

Everyone should be using f/2 lenses if doing broadband and the FOV captures your target -- narrowband you may want to stop down to f/4 because the of narrowband filter offset.

Camera speed = (aperture * pix size / focal length)^2

Obviously you cannot increase pix size so much that it degrades image quality.

If you look, however, at most AP images, you can zoom in quite a bit.

So the issue is not too little resolution, but too much resolution. 

i think most people can decrease resolution to increase SNR.



#7 imtl

imtl

    Viking 1

  • *****
  • Posts: 820
  • Joined: 07 Jun 2016

Posted 12 August 2020 - 08:09 AM

Everyone should be using f/2 lenses if doing broadband and the FOV captures your target -- narrowband you may want to stop down to f/4 because the of narrowband filter offset.

Camera speed = (aperture * pix size / focal length)^2

Obviously you cannot increase pix size so much that it degrades image quality.

If you look, however, at most AP images, you can zoom in quite a bit.

So the issue is not too little resolution, but too much resolution. 

i think most people can decrease resolution to increase SNR.

This is all fine when dealing with widefield. Not with small faint fuzzies. Over there you need speed and resolution. Having a small galaxy spread over 100 pixels is not to my taste as AP image. And that is what I get at the moment since I only have a widefield setup.

 

I think there is a balance (which is not straightforward) to reach between speed and resolution which depends greatly on your average local conditions. Nothing here is as simple as just go for the fastest lens. Fast lenses have their disadvantages as well.

 

So, I agree with you on most parts. Especially when it comes to widefield.

 

Eyal



#8 descott12

descott12

    Vendor - Solar Live View

  • *****
  • Vendors
  • Posts: 2,529
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 12 August 2020 - 08:40 AM

 

Extended object comparison = (f1/f2)^2

Point Source (stars) = (f1/f2)^2 + (Aperture1/Ap2)^2

 

I find it interesting that point sources have an aperture component. Any idea which one is most prominent in real-world situations?



#9 kathyastro

kathyastro

    Vanguard

  • *****
  • Posts: 2,174
  • Joined: 23 Dec 2016
  • Loc: Nova Scotia

Posted 12 August 2020 - 08:50 AM

I find it interesting that point sources have an aperture component. Any idea which one is most prominent in real-world situations?

That depends on your target.  An open cluster is all stars, so only the point source formula applies.  A nebula is an extended source, but there are obviously point source stars in the image which are not part of the target.  So one formula applies to the nebula and the other applies to the stars in the same frame.


  • ks__observer likes this

#10 sharkmelley

sharkmelley

    Soyuz

  • *****
  • Posts: 3,583
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 12 August 2020 - 09:06 AM

Point Source (stars) = (f1/f2)^2 + (Aperture1/Ap2)^2

This formula makes no sense to me because if you have 2 lenses with the same focal ratio and the same aperture then f1=f2 and aperture1=aperture2, so the formula gives:

Point Source (stars) = 2

which is plainly wrong

 

Can you either explain the formula or provide a reference for it?

 

Mark


Edited by sharkmelley, 12 August 2020 - 09:36 AM.


#11 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 09:37 AM

This formula makes no sense to me because if you have 2 lenses with the same focal ratio and the same aperture then f1=f2 and aperture1=aperture2, so the formula gives:

Point Source (stars) = 2

which is plainly wrong

 

Can you either explain the formula or point to a reference for it?

 

Mark

I honestly cannot explain it.

The formula was stated by one of the CN gurus.

In a private message he supplied me with the following citation, which honestly is a bit over my head.  See link below.

The formula, however, clearly shows that star brightness is a function of aperture^2.

https://www.dropbox....6PBnC8rNKa?dl=0


Edited by ks__observer, 12 August 2020 - 09:39 AM.


#12 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 09:42 AM

I find it interesting that point sources have an aperture component. Any idea which one is most prominent in real-world situations?

As Kathy, said it depends on the target.

Globular clusters would also benefit from larger aperture.



#13 kathyastro

kathyastro

    Vanguard

  • *****
  • Posts: 2,174
  • Joined: 23 Dec 2016
  • Loc: Nova Scotia

Posted 12 August 2020 - 09:44 AM

I don't think the "+" sign is intended to be taken literally.  The formula is trying to say that, for point sources, both the focal ratio squared and the aperture ratio squared play a role. 

 

In theory, point sources should be proportional only to aperture ratio squared, since a theoretical point source does not spread its light out on the sensor.  However, real world point sources do spread out their light due to diffraction, seeing, poor collimation, poor focus, etc.  But not as much as extended sources do, because stars start out as point sources.  So the actual pattern on the sensor is a mixture of point source effects and extended object effects.  Hence both appear in the formula.

 

Obviously, it is not a straight addition.  It will be a portion of one plus a portion of the other.  What those proportions are will depend on all those other factors.



#14 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 09:50 AM

For a given image scale.

 

Resolution is a big part of AP as well. If that was not the case then we would all be using a Canon/Nikon F/2 lenses with no need for a telescope. 

 

So I do agree with you about what you wrote but it is not the complete picture.

 

Eyal

I agree -- you need a minimum resolution to achieve sufficient image quality.  

Obviously, the smaller the target, the more resolution is needed.

I have found that in my location, Long Island, NY, my best nights with my 8in f/4 Newt, using an ASI-1600, at 1arc-sec/pixel -- and even using my 9.25 SCT, my very best FWHM is around 2.0 arc-sec -- but average FWHM is between 2.5 to 3.0 arc-sec.

If I need to sample the FWHM at 2.5 pixels for maximum resolution, then there is not much point where I live to imaging below 1arc-sec/pixel -- though from small objects I can shoot with the SCT at 0.5 arc-sec/pixel and then downsamaple/bin to 1.0 arc-sec/pixel to get to 2.5 arc-sec/pixel.


Edited by ks__observer, 12 August 2020 - 09:51 AM.

  • imtl likes this

#15 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,381
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 12 August 2020 - 09:54 AM

Apologies, meant to say:

Point Source (stars) = (f2/f1)^2 * (Aperture1/Ap2)^2

Or function of (Aperture / f-ratio)^2


Edited by ks__observer, 12 August 2020 - 12:54 PM.



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