#1  
Old 23-04-2012, 06:58 AM
dvj's Avatar
dvj (John)
Registered User

dvj is offline
 
Join Date: Oct 2011
Location: U.S.A
Posts: 755
Which One is Faster? F/3 or F/6

So here is a question to you tech theory types. An 8-inch F/3 or an 8-inch F/6. Which system is faster for astrophotography?

Looking at Nyquist sampeling information:

Let's match the pixel micron size to our FL's

8" f/3 with a CCD camera of ~2 micron pixels
8" f/6 with a CCD camera of ~4 micron pixels

Assume that both telescopes are of the same optical design and our CCD cameras have the same QE.

Aperature is the same, and we have optimized our pixel size for our focal length for coverate of at least 2 pixels by the Airy disc. Which would be a "faster" optical system for CCD Imaging?

Although the field of view is different, whould I actually gather more nebula in a shorter period of time with the F/3? Me thinks not.
Reply With Quote
  #2  
Old 23-04-2012, 08:03 AM
mplanet62's Avatar
mplanet62 (Michael)
Registered User

mplanet62 is offline
 
Join Date: Oct 2010
Location: Melbourne
Posts: 86
This is a very popular mix-up in terminology. While f number creates different perception of the picture brightness, light gathering capacity of a telescope depends chiefly on it's aperture. 8" f3 and 8"f6 have the same aperture, hence same light gathering capacity. With f6 you will just see lesser portion of a nebula and surrounding sky, and possibly more fine detail. But you will not see "more nebula" in any of those telescopes - as they grabbed and transmitted the same number of photons. The picture will look sharper, brighter and more distorted with f3 though - due to different light distribution inside the telescope.
Reply With Quote
  #3  
Old 23-04-2012, 08:03 AM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
You're operating at the same magnification in both cases and you'll have the same FOV if both cameras have the same pixel geometry (same number of rows and columns).

Both configurations will have the same effective speed. You're spreading the same number of photons over the same number of pixels in both cases - the aperture and FOV are the same.

Comparing f/ratio between scopes only makes sense if you're talking about the same sensor in both cases.

Cheers,
Rick.

Last edited by RickS; 23-04-2012 at 08:06 AM. Reason: Second post beat my reply by seconds. Removed comment that didn't make sense with an intervening post....
Reply With Quote
  #4  
Old 23-04-2012, 10:28 AM
avandonk's Avatar
avandonk
avandonk

avandonk is offline
 
Join Date: Aug 2005
Location: Melbourne
Posts: 4,786
The f ratio myth is real. You cannot beat real aperture. An f3 200mm aperture telescope will collect far more photons than an an F3 lens of 50mm FL say.

You did mention Niquist theory so obviously you understand sampling theory.


It is a complex balance of sampling and inherent brightness.

It is absolute heresy to say that an equivalent aperture at f/8 is as good as one at f/3 for dim extended objects!

By the way for star intensities only aperture rules. For nebulosity only focal ratio is the only arbiter.


One has to balance the the pixel size of the detector to the optic. The sampling here is important for final resolution.

What was the question again?

Bert
Reply With Quote
  #5  
Old 23-04-2012, 12:07 PM
rally
Registered User

rally is offline
 
Join Date: Sep 2007
Location: Australia
Posts: 896
Not really sure what the underlying question in your mind is here

However in your example - both systems will end up with the identical image scale.
If both cameras have the same number of pixels then both cameras will also have the same total field of view

However the telescopes image circle may or may not be matched to the sensors (since you said they were to be the same) and the Well depth of the 2uM sensor is going to be much smaller and certainly inferior to the 4um sensor and therefore likely have considerably more noise as a percentage of signal - ie Signal to Noise Ratio will be worse.

eg Read noise will be much the same for each and Dark noise too, but total signal will be much less in the smaller pixel sized CCD - based on area but also on the fact that there is some lost sensor area overhead in each pixel too - space for the internal wiring and the boundary edges of the silicon structures and any other etched circuitry.
So the signal to noise ratio will be significantly different - thus bit depth might end up being something like 7/8 Vs 10/11 (I'd need to work it out properly if you're interested)

The point source objects (stars) are still going to saturate the pixels in each system in similarly quick times, but you wont necessarily have the well depth to capture the faint details in the extended objects.
And you wont be able to increase the exposure time on the smaller camera to compensate because it will not have the well depth and hence dynamic range to collect any more photos/electrons.

So having the larger sized CCD is going to make all the difference in your example.
That will provide you with the best result.

Of course this is all moot if the Seeing conditions, light pollution and local conditions are not able to support it and nor is your system choice going to matter much if you cannot track and guide on the object to a high degree
So removing tracking, guiding, seeing errors, low periodic error etc are likely to make a much bigger difference - but I diverge from your posted question as stated, but i do think that is a missing element in what you are hoping to understand.

Cheers

Rally
Reply With Quote
  #6  
Old 23-04-2012, 04:59 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
I think its not that cut and dry.

Depends on the target. Some targets are widefield with bright and dim areas.

If you zoom in (F6) you may be imaging part of the dim area. If you zoom out (F3) you would be taking in more of the bright areas and therefore more signal and it would expose faster (not the dim areas though).

Of course you are assuming even performance of 2 sensors one with half the size pixels of the other and we know that is rarely the case.

9 micron pixels seems to be the allround winner in cameras. Theoretically smaller pixels should work better in a faster system but Kodak chips seem to perform best around the 9 micron pixel size. Perhaps a characteristic of their architectural design?

Kodak went to 5.5 micron pixels with their true sense sensors. These sensors only now seem to be coming out in cameras despite being around for quite some time. It takes the market a long time to accept new products like that.

Of course there are no popular cameras with either 4 or 2 micron pixels out there. The smallest Kodak is 5.4microns. Not sure about Sony - they may have some in the 4+ micron band.

Greg.
Reply With Quote
  #7  
Old 24-04-2012, 06:38 AM
dvj's Avatar
dvj (John)
Registered User

dvj is offline
 
Join Date: Oct 2011
Location: U.S.A
Posts: 755
My intent with the question is to simply understand the f-ratio myth better and how to explain it.

There are no 2um or 4um pixel cameras that I am aware of, but I wanted to keep some constants in the model. 200mm aperture, pixel size matched to the focal length as per Nyquist theory.
Reply With Quote
  #8  
Old 24-04-2012, 03:19 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Quote:
Originally Posted by dvj View Post
My intent with the question is to simply understand the f-ratio myth better and how to explain it.

There are no 2um or 4um pixel cameras that I am aware of, but I wanted to keep some constants in the model. 200mm aperture, pixel size matched to the focal length as per Nyquist theory.
Best to get it out of theory and into practise.

You have 2 approx 12 inch F3.6/8 scopes being used on this site - An APRHA by Peter Ward and a 12 inch Orion Optics UK F3.8.

They both are producing images that have not been seen with similar aperture but longer focal length.

So there is something that occurs there. Perhap the little variables of pixel matching, less effect of seeing, simpler optical paths, less prone to tube currents, efficiencies of larger pixels add up to a noticeable difference.

Greg.
Reply With Quote
  #9  
Old 24-04-2012, 09:00 PM
rally
Registered User

rally is offline
 
Join Date: Sep 2007
Location: Australia
Posts: 896
Greg,

Not sure what you mean by this - image scale is image scale.
Run some examples through to verify this.

In the example provided by the OP - The image scale is the same !
You will get EXACTLY the same image from both of them - if both cameras have the same number of pixels.

Both image scales are EXACTLY 0.69 arcsecs/pix, 7.3 x 7.3 arc mins fov on a 640x640 array.
The f6 system is concentrating its light onto larger CCD, the f3 onto a smaller one !

I didn't assume the performance of the two CCDs was the same - I said they were very different and why.

9uM pixels may be ideal for some situations, but definitely not all of them.
Sometimes 24uM pixels are best and other times 2uM pixels are best.
Its a question of selecting the ideal sampling rates and image scale for the OTA and the intended purpose. Not to mention CCD specs such as spectral response, Qe, budget etc
So yes a different choice of CCDs on the same OTA is necessary for Planetary or DSO or Widefield Vs Narrowfield etc, but that is not what was being asked.

I think the OP's choice of pixel sizes used was purely hypothetical to illustrate problem, but Lumenera have a number of 4um cameras for Astro use as do others and there are probably dozens of point and shoots in the 2um size range.

Rally



Quote:
Originally Posted by gregbradley View Post
I think its not that cut and dry.

Depends on the target. Some targets are widefield with bright and dim areas.

If you zoom in (F6) you may be imaging part of the dim area. If you zoom out (F3) you would be taking in more of the bright areas and therefore more signal and it would expose faster (not the dim areas though).

Of course you are assuming even performance of 2 sensors one with half the size pixels of the other and we know that is rarely the case.

9 micron pixels seems to be the allround winner in cameras. Theoretically smaller pixels should work better in a faster system but Kodak chips seem to perform best around the 9 micron pixel size. Perhaps a characteristic of their architectural design?

Kodak went to 5.5 micron pixels with their true sense sensors. These sensors only now seem to be coming out in cameras despite being around for quite some time. It takes the market a long time to accept new products like that.

Of course there are no popular cameras with either 4 or 2 micron pixels out there. The smallest Kodak is 5.4microns. Not sure about Sony - they may have some in the 4+ micron band.

Greg.
Reply With Quote
  #10  
Old 25-04-2012, 06:53 AM
avandonk's Avatar
avandonk
avandonk

avandonk is offline
 
Join Date: Aug 2005
Location: Melbourne
Posts: 4,786
Have a look here John

http://www.clarkvision.com/articles/...yth/index.html


This bloke Roger Clark knows what he is talking about and explains things very clearly. In fact his whole site is a mine of information.

One of the references he gives is also worth a look.

http://www.stanmooreastro.com/f_ratio_myth.htm



It seems the f-ratio myth was really due to the reciprocity failure of film. It just does not apply as much to modern detectors that have a far wider integration time than film.


Bert

Last edited by avandonk; 25-04-2012 at 07:16 AM.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 08:25 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement