You don't need aperture for imaging?

[MartinB]

Sorry I've come to this a bit late, been busy with other things! Focal length determines image scale - how big the image looks and the FOV. Exposure time is determined by the F ratio of the scope so aperture is a factor but only in relation to the focal ratio. For a given focal length increasing the aperture will shorten the exposure time.

Resolution is determined by a number of factors but most notably pixel size of the camera and focal length of the scope, and for long exposure DSO imaging is limited by seeing conditions. The smaller the pixel size or the longer the focal length the smaller the sampling rate measured in arc seconds per pixel. Traditionally a sampling rate of 2-3 arc seconds per pixel has been thought of as being ideal. You need good seeing to drop below this. In superb seeing it is apparently possible to image effectively with a sampling rate of less that 1 arc second per pixel, in poor seeing 5 arc seconds per pixel could be your limit. So if you use a ZS66 and an SXV H9 you will achieve better resolution than with an SBIG ST2000 because the H9 has smaller pixels. However if you were to use the same cameras with a C11 the SBIG2000 would probably do a better job because it's larger pixels would sample the highly magnified image more effectively.

The other important issue is camera chip size. If you are using a modded web cam you could probably get M42 using a ZS66 and a 0.5 F/R. You will however be "under sampling" - too many arc seconds per pixel for fine resolution. Using an SBIG1100 you could probably get M42 in the field of view with an 1800mm focal length scope with around 1 arc second per pixel i.e very high resolution if the seeing allowed.

However, large chips need a flat field to avoid distortions at the periphery so you are then talking Tak FSQs, Richey Cretiens ££££ Coma correctors can apparently do a good job with reflectors for much less.

So aperture is an issue for exposure time but not resolution. Assuming you have one imaging camera what is really needed is a carefully chosen mix of scopes able to work with a focal reducer giving you a range of focal lengths from 300 to 2000mm. For imaging scopes are really the equivalent of eye pieces for visual.

[Kaptain Klevtsov]

For imaging scopes are really the equivalent of eye pieces for visual.

Ah ha! So, how many eyepieces do you visual guys own, on average? I'm guessing at six, or thereabouts, so I need (counts on fingers) at least another 'scope. to even have half a chance. So its official, can't wait to tell her.

Joking aside, I've been looking for some hard data WRT central obstruction and stuff, and came across loads of references to the Dawes Criterion. This states that the maximum resolution theoretically possible is equal to (4.56/ Aperture in inches) in arc seconds of resolution. This would be 116/mm in metric. So the ED80 would have a resolution of 1.45 arc seconds, the 200mm Newt would do better at 0.58 arc esconds.

I've got a lot to learn, and I'm off to read Damian Peach's (he the man!) site here. http://www.damianpeach.com/simulation.htm

Captain Chaos

[MartinB]

But Damian Peach is a planetary imager using fast exposures much less influenced by the seeing. Your Atik and 8" F5 reflector will give a sampling rate of 1.32 arc seconds per pixel. You can probably achieve that on nights of especially good seeing.

[FLO]

I attended Damien Peach's talk at Astrofest where he demonstrated very clearly the link between aperture and resolution, using graphs, maths and pictures! The link is undeniable, what is interesting though is whether or not it is necessary to have more than 3-4" aperture when imaging objects other than the Moon and planets?

[GordonCopestake]

Martin, so what is the "ideal" for a dSLR? To give the best resolution?

[Kaptain Klevtsov]

But Damian Peach is a planetary imager using fast exposures much less influenced by the seeing. Your Atik and 8" F5 reflector will give a sampling rate of 1.32 arc seconds per pixel. You can probably achieve that on nights of especially good seeing.

So what this means in real life terms is that the 'scope resolution with good seeing is better than the camera can do, so no detail is lost by using two camera pixels for one star (with a Golb. for example). Allowing for seeing, the combination should be ok. Is that correct martin?

Captain Chaos

[MartinB]

So what this means in real life terms is that the 'scope resolution with good seeing is better than the camera can do, so no detail is lost by using two camera pixels for one star (with a Golb. for example). Allowing for seeing, the combination should be ok. Is that correct martin?

Captain Chaos

That's certainly true with long focal lengths and moderate seeing CC plus the extra sensitivity of course.

Steve, aperture makes a major difference for planetary imaging because the seeing is less of an issue (although still significant). Planetary cameras also have very small pixels so you can get down to the Dawes criterion (and beyond for planets where Dawes doesn't always apply) esp in Barbados!

Gordon, it will still apply with DSLRs. The sampling rate formula is 206 x pixel siize in microns/focal length in mm. You just need to find out the pixel size of your camera

[GazOC]

I know Damien Peach has imaged lunar features that theoretically his scope can't resolve.

[GordonCopestake]

The D70 I have has a pixel size of 7.8 microns (I think) so the forumula gives:

206 x 7.8 (D70) / 1500mm (for my C6) giving 1.0712 arcseconds at f/10

Using a 0.63 focal reducer gives a focal length of 945mm, so

206 x 7.8 / 945 = 1.7 arcseconds at f/6.3

and a 3.3 focal reducer gives 495mm, so

206 x 7.8 / 495 = 3.246 arcseconds at f/3.3

Using these numbers as a rough guide it seems that the longer the focal length the better the resolution? Which kinda makes sense, although that doesn't take into account FOV. That being said a 6" scope has a theorectical resolution of 0.77 arcseconds which means to hit that resolution on my D70 I would need a focal length of 2086.755874! or f/14

[Astroman]

Thanks MartinB for that excellent analysis! It was just after reading another, nearly identical treatise that I decided to mainly stay with visual, since I didn't understand a word of it. :shock: Then, I decided to "do REAL science", and was forced to grapple with the beast.

Nicely explained.

[Kaptain Klevtsov]

I've been trying to sort this out and came up with an analogy that seems to work, maybe a few comments from you guys would tighten up my grasp on it.

The light coming into the 'scope arrives as photons. These I imagine as tiny paintballs, which splat onto the CCD of the camera. The 'scope model (in my analogy) has magnets round the end of the tube (it looks like a refractor, as that's the simplest to think about, but has a central obstruction like an SCT). The paintballs are also magnetised so that any that pass close to the magnets round the rim of the 'scope get repelled and aim for the focus point on the CCD. The closer they are to the magnets, the more they get repelled and bent towards the focus point. Any paintballs that enter through the middle get repelled equally from all the rim magnets and so carry on without deviation. These also hit the focus point on the CCD, or would do if the CO wasn't there.

WRT resolution, its obviously pointless to have a camera with pixels so small that the splatted paintball covers them all, so the ideal is to have the splat from a paintball just cover one pixel.

The splat size is determined by the aperture of the 'scope, but the central obstruction messes with the rim magnets as all the best paintballs (the ones which don't need to be deflected), hit the CO and are wasted. The paintballs which just miss the CO are wrongly deflected as the CO is metal and the paintballs are magnetised.

The splat size is the point spread function, and the unwanted deflection is diffraction in the jargon.

Is any of this making sense?

Captain Chaos

[MartinB]

Phew! So that simplifies things eh :shock: Not sure about the splat size depending on aperture but the number of splats will be greater with more aperture. Splat size is PSF - yes (heavy duty stuff) and the deflection thing would do for diffraction but not sure where this is leading. I think your original view about CO not being an issue is correct. RC scopes have more CO than any other type and are the Rolls Royce imaging scopes. Not great for visual because of the CO apparently. Other than reducing effective aperture a small amount I doubt it makes any significant difference when imaging.

[GordonCopestake]

I rediscovered a link in my favorites that discusses the effect of a central obstruction (a little off topic, but we were discussing refractors and newts, with aperture vs no central obstruction)

http://legault.club.fr/obstruction.html

[GordonCopestake]

And this one is a little more on topic:

http://legault.club.fr/which.html

A good read CC if you havn't come across it before.

[Kaptain Klevtsov]

Thanks Gordon, I'll peruse those shortly.

Captain Chaos

[Robfal]

I know this is off subject however CC (and anyone else interested)

I use this free download to work out how an image will look on a CCD for a given scope. Download the extra photo's from the link. This program give a visual image of how the DSO should look

http://www.newastro.com/newastro/book_new/camera_app.asp

I'll post this in the imaging processing section aswell (where is should be)

Rob

[GordonCopestake]

Sounds good Rob, i'll take a look

[Kaptain Klevtsov]

Now that I've read the two pages that you pointed at Gordon, I'm more confused about the CO than ever. The first one shows data that suggests that my 200mm Newt. degrades, by virtue of the CO, to a 130mm APO. Now I'm not currently in a position to spare any body parts (most of mine are decidedly second hand anyway) so I can't see me getting a big APO anytime soon in exchange for the Newt. That bit cheered me up a bit to be honest.

The second one, in the very last bit, switched on a lightbulb for me. I've been using the IR block filter since I got it as it gets rid of the unfocussed IR light which messes up the image. Well not with a Newt. it doesn't, because there is no refraction, only reflection, so all the colours go where they ought to. Maybe I need to try without the blocking filter and see what happens.

That's something else I've learnt here in this thread. Great stuff. Just need Kain to make another box so that he can lock his clouds up next to his hairy spiders and I can have another crack at it.

Captain Chaos

[GordonCopestake]

Hehe, glad it was some use CC! I wont deny that a lot of it went over my head, but it looked to be talking about the right sort of thing.

[FLO]

The first one shows data that suggests that my 200mm Newt. degrades, by virtue of the CO, to a 130mm APO.

What is degraded? Light, resolution or contrast?

[themos]

Well not with a Newt. it doesn't, because there is no refraction, only reflection, so all the colours go where they ought to.

Except if you put a Barlow or similar in the way, of course. But you are into DSO's, not planetary imaging, is that right?

[GazOC]

The first one shows data that suggests that my 200mm Newt. degrades, by virtue of the CO, to a 130mm APO.

What is degraded? Light, resolution or contrast?

Contrast, resolution for a 200mm Newt is given as a ~190mm refractor

[GazOC]

This is the maths in a nutshell (from Hartmut Frommerts website)

1. the image definition (contrast, sharpness) equivalent to that of an unobstructed instrument with effective aperture D_eff:

D_eff = D - d

2. the light-gathering power of an unobstructed telescope with aperture D_lg (very straightforward):

D_lg = sqrt( D^2 - d^2 )

So

Reflector diameter

33 % obstruction 300 250 225 200 150 110

Equivalent diameter of refractor

about image definition 200 170 150 130 100 73

Equivalent diameter of refractor

about light gathering 280 235 210 190 140 100

[Kaptain Klevtsov]

Well not with a Newt. it doesn't, because there is no refraction, only reflection, so all the colours go where they ought to.

Except if you put a Barlow or similar in the way, of course. But you are into DSO's, not planetary imaging, is that right?

Of course you are right there, I only use the webcam for planets and that involves some heavy duty Barlowing, better keep the IR blocker on then.

Captain Chaos

[FLO]

Talking of aperture and resolution - we could see the swirls the Whirlpool galaxy's dust lanes thru the 12" Skyliner 300P at the Star Party 8)