impact of aperture, f-ratio etc for imaging

[rmuhlack (Richard)]

I've been trying to get my head around the influence that different imaging equipment properties have on the final resulting image.

In part this comes from looking at some of the amazing images in the deep space forum and trying to figure out how I might be able to convert their acquisition details with the gear that they used to an equivalent for the gear that I have so that I can then benchmark my images against theirs.

So, lets take a hypothetical where we have two scope:
Scope 1 is a 200mm aperture f5
Scope 2 is a 300mm aperture f5

Obviously focal length is different for each, so lets assume we select a camera for each scope that gives each scope/camera combo an image scale of 1 arcsec per pixel. QE, well depth, read noise, dark noise and operating temperature for both cameras in this hypothetical are the same.

Each scope and camera combo is used under dark skies where seeing is limited to 2 arcsecs.

So - in what respect will the images from these two setups be different?

[rmuhlack (Richard)]

if there is no difference, does this mean that if I see someone's shots in the deep sky section or on astrobin etc and would like to try the same target, then provided image scale is similar, I can scale exposure times by f-ratio and QE to get a ball park estimate of an equivalent exposure to what they used?

[sgazer (Lee)]

quickly without too much thought, I think the camera on the 300mm f5 scope will have larger pixels in order to maintain the same arcsec/pixel with the longer focal length, but despite this it's performance is stated as being the same as its smaller pixel sibling, so overall, I don't think there would be any difference.

However normally, the larger pixel camera (all else being equal) would have lower noice and better efficiency, so the camera itself would be better.......


I think !!

[rmuhlack (Richard)]

an interesting point - thanks.

I guess a related side question is under these hypothetical conditions, what advantage is gained from more aperture (if any)

[Terry B]

Originally Posted by rmuhlack

an interesting point - thanks.

I guess a related side question is under these hypothetical conditions, what advantage is gained from more aperture (if any)

Resolution= aperture/wavelength.
The larger the scope the higher the resolution.
This is the theory.
The practice is that for large scopes (ie not 60mm refractors etc) other problems tend to limit resolution like the "seeing", tracking etc.

[rmuhlack (Richard)]

Originally Posted by Terry B

Resolution= aperture/wavelength.
The larger the scope the higher the resolution.
This is the theory.
The practice is that for large scopes (ie not 60mm refractors etc) other problems tend to limit resolution like the "seeing", tracking etc.

that is my understanding also. However for the hypothetical equipment setup I described, where both scopes are using cameras which give 1 arcsec per pixel, and both are operated under skies where seeing is limited to 2 arcsecs per pixel, is there any practical resolution gain from the larger aperture scope?

In fact on further thought, if the image scale is 1 arcsec per pixel for both, does it even matter what the seeing is? Even if the seeing was at 0.5 arcsec per pixel, if both scopes are using a camera which has an image scale of 1arcsec per pixel, then surely that is the resolution which is recorded ?

[Merlin66 (Ken)]

The size of the Airy disk is:
2.44 x lambda x f ratio (in microns) where lambda is the wavelength.
So although the larger scope will gather more light it (due to the longer focal length) is spread over a larger linear disk.
The resolution will be defined by the seeing and the pixel size (Nyquist sampling) - so if each scope/ camera combo has the same arc sec/ pixel then under similar seeing conditions they will have the same effective resolution.

[rmuhlack (Richard)]

thanks Ken. just to clarify, what are the wavelength units in that formula - microns or nm ?

[Merlin66 (Ken)]

Lambda in nm.

[Terry B]

Originally Posted by rmuhlack

that is my understanding also. However for the hypothetical equipment setup I described, where both scopes are using cameras which give 1 arcsec per pixel, and both are operated under skies where seeing is limited to 2 arcsecs per pixel, is there any practical resolution gain from the larger aperture scope?

In fact on further thought, if the image scale is 1 arcsec per pixel for both, does it even matter what the seeing is? Even if the seeing was at 0.5 arcsec per pixel, if both scopes are using a camera which has an image scale of 1arcsec per pixel, then surely that is the resolution which is recorded ?

The image scale isn't the defining issue. As Ken states the size of the airy disc is. For your example probably the airy disc is smaller than the seeing but if you used an extreme example like a 30mm diam scope then the airy disc will be much greater than the seeing. If you still have 1 arcsec/pixel it just means you are oversampled and this isn't the resolution.

[rmuhlack (Richard)]

i see. so if the image scale was greater than the size of the airy disk in both cases, there is nothing to be gained by using the 300mm over the 200mm ? but if that is not the case, then (assuming perfect seeing) the 300mm will deliver greater resolution?

And I take it that in my example where the f-ratio, and camera specs such as QE, noise etc are held constant, that exposure remains constant as well across both scope/camera combos....?

[rogerg (Roger)]

Originally Posted by rmuhlack

And I take it that in my example where the f-ratio, and camera specs such as QE, noise etc are held constant, that exposure remains constant as well across both scope/camera combos....?

I think this question is irrelevant because you've placed so many unrealistic fixed constraints on your example. You say "camera specs such as QE, noise etc are held constant" yet you say both systems have 1"/pixel at different focal lengths.

Realistically your pixel size will have changed in order to have 1"/pixel at both focal lengths, affecting QE and probably other aspects of the camera (full well depth?), noise characteristics of the camera?, etc.

[rmuhlack (Richard)]

I would expect that a "stronger signal" is produced/recorded (sorry, that is probably the wrong terminology) for faint stars and diffuse nebula etc with the larger 300mm scope in my example. Or is that governed by the f ratio as well?

[rmuhlack (Richard)]

Originally Posted by rogerg

I think this question is irrelevant because you've placed so many unrealistic fixed constraints on your example. You say "camera specs such as QE, noise etc are held constant" yet you say both systems have 1"/pixel at different focal lengths.

Realistically your pixel size will have changed in order to have 1"/pixel at both focal lengths, affecting QE and probably other aspects of the camera (full well depth?), noise characteristics of the camera?, etc.

I take your point. It is of course an unrealistic scenario. I posed the hypothetical like that simply to try and tease out )in my own mind as much as anything) how these different properties would impact on data acquisition. This is firstly so that I might have some clue about how I can scale the equipment specs used for images that I see others post here to get vague ballpark equivalent capture details with my existing equipment. And second, that if I were to upgrade my camera and/or scope, what specs to focus on within a finite budget to optimise the performance:price relationship.

[rmuhlack (Richard)]

apologies if am asking too many questions

[Merlin66 (Ken)]

No, there's no "stronger" signal - for stellar images...due to the enlarging of the Airy disk with focal length. Your constraint of the same f ratio mean longer focal lengths with larger apertures...

[rogerg (Roger)]

Originally Posted by rmuhlack

I take your point. It is of course an unrealistic scenario. I posed the hypothetical like that simply to try and tease out )in my own mind as much as anything) how these different properties would impact on data acquisition. This is firstly so that I might have some clue about how I can scale the equipment specs used for images that I see others post here to get vague ballpark equivalent capture details with my existing equipment. And second, that if I were to upgrade my camera and/or scope, what specs to focus on within a finite budget to optimise the performance:price relationship.

yeap, worth doing I just think you might have too many unrealistic fixed constraints to actually work through a scenario that will make sense This stuff does my head in sometimes.

Re your other comment about "stronger single" ... you're saying you have 1"/pixel in both setups and you have f/5 in both setups so I don't think you have a "stronger" signal in one or the other. Again, I think the scenario is unrealistic

I could be wrong.

[rmuhlack (Richard)]

Okay then, I'll give an example of the sort of calculation I'm thinking. An extreme example and most likely unachieveable for me, but its always good to challenge yourself, and as they say "imitation is the sincerest form of flattery".

Anyway...We have Mike S's sensational Virgo Galaxy photo here. He has a total integration time of 240mins, using his AG12 at f3.8 with a KAF16803 which has an image scale of 1.63 arcsecs per pixel. The size of the Airy disk with his scope is 1.09 arcsecs (according to this calculator here). I'll use an averaged QE of 55%

My scope is a VC200L with focal reducer (200mm aperture, f6.4), and a lowly (comparatively laughable) Canon 400D. The image scale is 0.92 arcsecs per pixel. The Airy disc with this scope is 1.64 arcsecs. I will optimistically assume a QE similar to the 40D listed here of 33%

So 240mins x (f6.4/f3.8)^2 = 680mins. To account for the difference in QE, 680mins *0.55/0.33 = 1134mins. The full well and read noise of the KAF16803 is 100,000 and 9e- respectively, while for the 400D at ISO800 it is a measly 5611 and 5.2e- respectively. The dynamic range of both is 11111 for the KAF16803 and 1079 for the Canon. Not quite sure how to factor this one in - perhaps it means that I should scale my subs by a factor of 1079/11111 if I want to preserve the dynamic range that Mike has, although that would probably mean that Im only taking subs of about 1min (assuming he was using say 10min subs) which means i'm probably missing out on the faint stuff. As far as theoretical resolution goes, possibly we are similar - he is limited by his image scale of 1.6", where as I am limited by the airy disk of 1.6" also...?

If I were to upgrade to say a KAF6303, that would change my image scale to 1.45" (so still limited by the Airy disk) however my QE would go up to something like 65% and with a full well of 100,000 and read noise of 11e- my dynamic range is now at 9090. Because of the increase in QE my exposure has dropped from 1134 to 576mins. Also, I can now possibly take ~9min subs to Mike's 10mins and maintain a similar dynamic range. Resolution as I understand it would still similar at about 1.6" in both cases.


I could have made some mistakes in the above and I apologise if that is the case, but it hopefully gives a practical example of why I've been asking these questions.



(btw - I realise the FOV will be different for the two scopes above. I'm also aware that the KAF16803 is mono and my existing camera is OSC and that in itself will result in a resolution cost. Obviously the above ignores seeing effects as well)

[Bassnut (Fred)]

When you say "impact" of all these technical issues, in the end it's all about the "impact" of the image really.

Aesthetically anyway, anything under 2m FL is rubbish, so your in trouble there, and the only things that count camera wise are QE and well depth, the 6303 is an excellent choice right there

[rmuhlack (Richard)]

Originally Posted by Bassnut

When you say "impact" of all these technical issues, in the end it's all about the "impact" of the image really.

Aesthetically anyway, anything under 2m FL is rubbish, so your in trouble there, and the only things that count camera wise are QE and well depth, the 6303 is an excellent choice right there

I assume that as Mr "Narrowfield rules" that your statement of "anything under 2m FL is rubbish" is just hyperbole. Surely FOV comes into "image impact" as well. I mean, if we were to use an Orion Optics ODK12, that has a FL of 2040mm. A KAF16803 and a KAF1603 both have similar well depth and QE of 100,000 and 55-60% and the same pixel size of 9micron, yet the 1603 on that scope will give a FOV of 23.5'x15.6' where as the KAF16803 will give more than 10x the area at a FOV of 62'x62'