Mirror Quality for Imaging?

[Bassnut (Fred)]

Originally Posted by alpal

Do you have some results that you could post please?

That would be only stretched examples - no other processing & at full scale.

Well no, not really. Id have to plough through large numbers of subs from the LX200R to find the best and I couldnt now tell if the seeing was the same as the subs taken with the RCOS many months apart.

Perhaps I should have said I didnt spend too much time keeping the LX200R in perfect focus either so the number of samples that matched the RCOS were pretty small.

Seeing for both at the same urban site was usually around 6 arc secs, the best was 4 arc secs and that was rare.

With megadata on the same object for both OTAs with 6 arcsec seeing at the same exposure times and with my usual overprocessing, the 12" actually often gave better results on dim objects (less noise due to less stretching with more apature).

Now that I am in dark skies at typical 2 arcsecs and some times better seeing, I dont think the LX200R would compete with the RCOS, you wouldnt expect it to.

[clive milne]

Originally Posted by alpal

So - you think that it 's all bragging rights then?

I didn't say that.
What I did imply though was that people have been known to optimise one particular component of a system at great expense without doing a cost/benefit analysis with respect to how that impacts the performance of the system as a whole.

I think if I searched the internet I'd find comparisons that would disprove your theory.

Bearing in mind that the op was specifically asking about the impact of mirror quality in newtonians it would be better to limit the comparison to images taken through newtonians, or at the very least, telescopes with apertures and focal ratios typical of newtonians.

A much easier way to see the impact of 1/3 wave on prime focus imaging would be to find someone with a newtonian of known quality and ask them to take a few images for you. With and without a 50% central obstruction (which could be circle cut out of cardboard) The diffraction effects of the larger central obstruction will be of similar properties and magnitude as 1/3 of SA.

There are many images on the net (even this forum) taken by instruments with 50% central obstruction that show no degradation of imaging performance by virtue of their central obstruction, so implicitly, 1/3 wave is good enough for such an instrument.

[clive milne]

Further to that, it would be instructive to increase the system focal ratio to the point where the focal plane is no longer under sampled ie) f24 and repeat the experiment. You will find that 1/3 wave is no longer good enough.

[clive milne]

On the subject of 1/100 wave optics. Some questions I would ask:
Are these optics mechanically supported in such a way that preserves the original figure of revolution at all alt angles?
Is the mirror substrate temperature actively controlled to within 0.1 degree of ambient? how is boundary layer turbulence managed? Is the primary mirror baffle flushed to disrupt the chimney effect?

I hope you see my point. ...

[alpal]

Originally Posted by clive milne

I didn't say that.
What I did imply though was that people have been known to optimise one particular component of a system at great expense without doing a cost/benefit analysis with respect to how that impacts the performance of the system as a whole.



Bearing in mind that the op was specifically asking about the impact of mirror quality in newtonians it would be better to limit the comparison to images taken through newtonians, or at the very least, telescopes with apertures and focal ratios typical of newtonians.

A much easier way to see the impact of 1/3 wave on prime focus imaging would be to find someone with a newtonian of known quality and ask them to take a few images for you. With and without a 50% central obstruction (which could be circle cut out of cardboard) The diffraction effects of the larger central obstruction will be of similar properties and magnitude as 1/3 of SA.

There are many images on the net (even this forum) taken by instruments with 50% central obstruction that show no degradation of imaging performance by virtue of their central obstruction, so implicitly, 1/3 wave is good enough for such an instrument.

I still reckon you need a 1/16 wave for both primary & secondary.
At least you get a theoretical 1/4 wave at the eyepiece or camera.
I have a 27% obstruction on my 8" Newt.
I can't find my old program to tell me the loss of contrast with that obstruction.

cheers
Allan

[clive milne]

With all due respect Allan, an unwillingness to accept a logical and empirically derived argument is not actually a valid defence against it.

[alpal]

Originally Posted by clive milne

With all due respect Allen, an unwillingness to accept a logical and empirically derived argument is not a valid defence against it.

I appeal to authority:

"So why were some of the ( now no longer produced ) RCOS scopes ion milled to 1/100th wave accuracy?"

You have to convince me that leaders in Govt. & scientific & engineering
astronomy businesses have all made invalid choices by paying a lot of extra money
to have ion milling performed on their telescopes.

1/100th wave is way better than anything we're talking about.

[Rac (Raymond)]

Haha maybe that's because they didn't have to pay for it out of their own pocket.

[clive milne]

Originally Posted by alpal

I appeal to authority:

I think 'invoking' is a more accurate term.

You have to convince me that leaders in Govt. & scientific & engineering
astronomy businesses have all made invalid choices by paying a lot of extra money
to have ion milling performed on their telescopes.

The measure of a logically consistent argument is not determined by my ability to convince anyone of anything. Nor is it determined by the spending habits of a government department. If that were the case, a tradesman would conclude that hammers cost US$364, toilet seats would be US $640 and coffee machines would be US $7600 each.
http://en.m.wikipedia.org/wiki/Proje...ment_Oversight

[alpal]

I post the above.

[alpal]

http://www.rcopticalsystems.com/tele...n_milling.html

"Post figuring is a common procedure in modern glass optics. The Keck telescope mirror segments, for example, were polished to the l/20 (optical) level and then post figured to the l/100 level by ion milling". See:4. Allen, L.N., Keim, R.E., Lewis, T.S. & Ullom, J. "Surface Error Correction Of A Keck 10m Telescope Primary Mirror Segment By Ion Milling", Proc. SPIE 1531, 195-204 (1991) .

Ion Beam Figuring (IBF) was originally developed by the Eastman Kodak Company in 1988 and became operational in 1990. IBF is an excellent complement to conventional figuring. The optics are first polished (ground) conventionally and then final figure is milled by IBF. Ion Beam Figuring works on a molecular (atomic) level.

Without question, Ion Milling is the most important breakthrough ever in optical manufacturing!

RCOS offers the HIGHEST QUALITY and most economical ION MILLED OPTICS possible in the 0.37m to 1.00m sizes. We have supplied hundreds of telescopes with ION Milled Optics.

A Ritchey-Chrétien 16 inch primary mirror already figured and polished to 1/6 wave P-V is mapped by a special camera. The optics are then inserted into a high-vacuum chamber facing down. The IBF then directs a beam of argon ions upward to the glass. The glass is removed on a molecular level. The beam itself is translated across the optical surface through a computer controlled mask, removing figuring errors and surface roughness leftover from conventional polishing.

When the Ion Beam-Figuring is finished, the chamber is purged and the process is complete. This completed 16 inch Ritchey-Chrétien primary mirror required only 3 iterations with the Ion-Beam.

[glend (Glen)]

And this is an affordable process for amateurs, ie people without government budgets? And where in Australia can I send my mirror to get this done? As the orignal posted, i'd like to remind the debating team that what I raised was mirrors for imaging newtonans, and if quality improvements would yield better images - something I am now convinced is not justifed - on a retirees income.

[alpal]

Originally Posted by glend

And this is an affordable process for amateurs, ie people without government budgets? And where in Australia can I send my mirror to get this done? As the orignal posted, i'd like to remind the debating team that what I raised was mirrors for imaging newtonans, and if quality improvements would yield better images - something I am now convinced is not justifed - on a retirees income.

How much do you want to spend?

[clive milne]

None of which is relevant to prime focus newtonian primary mirror accuracy requirements.

[Satchmo]

Alpal : The modification to the airy pattern caused by central obstruction or spherical aberration are not quite the same although parallel points can be found at similar Strehl ratios . When you combine spherical aberration and central obstruction the effects are more complex and I would think under different kinds of seeing , it is not so easy to model the outcome. Conversations I have had with Peter Read from SDM who often compares telescopes of known quality together visually , is that better optics seem to fair much better in some kinds of slower period inclement seeing than marginal ones- which in the immediate instance seems counter intuitive. I think it just comes from having more headroom for transverse error in a geometric analysis

I looked at this with Aberrator a while back and I dont think it is a simple as just allowing spherical aberration effect to be as large as the central obstruction degradation and assuming it will just mask it. In geometric terms your ray bundle is already 3 times the diameter of the airy disc with an unobstructed system.

I think getting the best images possible is achieved by trying to keep all aberrations to a minimum . You are going to have seeing , optical mounting problems , mis-collimation and and local thermal issues all in the `wobbly stack ' from star to image plane, even assuming perfect optics . But I don't think letting optical aberrations loose to the theoretical limits of your obstruction will always yield maximum quality - I think it is prudent to try and build some headroom into all aspects of the telescope.

Professional instruments with large obstructions and sites with no better than 1 or 0.5 arc second could probably specify their optics to a number of waves if they were to let aberrations slide to equal the best case blur spot for a given seeing condition , but they generally try to achieve better than 1/20 wave because they can and it is a form of future proofing.

Most of the best imaging around is done by amateurs who have paid attention to every aspect of their setup to try and keep their signal to noise as good as possible . They are generally not taken with cheap optics , cheap mounts etc . Obviously chasing or paying for 1/100 wave optics would be a hollow pursuit , but I think good headroom means optics of 1/8 wave front or better where the geometrical ray bundle meets at about half the diameter of the theretical airy disc. It would be hard to pinpoint exactly how all the improvements are made , but I'm guessing its the some total of a lot of subtle gains in every aspect of the instrument .The anecdotal evidence I have seen from imagers using known quality optics suggest to me that quality it is not an empty thing to strive for in the pursuit of fine images. Headroom in optical quality is important too.

[Shiraz (Ray)]

systems like Keck need very high precision mirrors, because they can operate with true deformable mirror adaptive optics, which allows them to cut through seeing and get close to the diffraction limit - so the optics must be diffraction limited.

A more useful comparison to our imaging systems is possibly the VISTA survey scope on Paranal. This has a FWHM of 0.51 arcsec (from a 4m mirror!!), so has been designed with optics that are good enough for the seeing (which can be a bit less than 1 arc sec), but are nothing like diffraction limited. In Australia, we get seeing about 1/2 that good (ref), so we can get by with a 1 arc sec FWHM optical system. An ordinary 8 inch GSO Newtonian will do better than that.

ref: http://203.15.109.22/astro/manuals/obsguide/node5.html fig 1.6

[alpal]

Originally Posted by Satchmo

Alpal : The modification to the airy pattern caused by central obstruction or spherical aberration are not quite the same although parallel points can be found at similar Strehl ratios . When you combine spherical aberration and central obstruction the effects are more complex and I would think under different kinds of seeing , it is not so easy to model the outcome. Conversations I have had with Peter Read from SDM who often compares telescopes of known quality together visually , is that better optics seem to fair much better in some kinds of slower period inclement seeing than marginal ones- which in the immediate instance seems counter intuitive. I think it just comes from having more headroom for transverse error in a geometric analysis

I looked at this with Aberrator a while back and I dont think it is a simple as just allowing spherical aberration effect to be as large as the central obstruction degradation and assuming it will just mask it. In geometric terms your ray bundle is already 3 times the diameter of the airy disc with an unobstructed system.

I think getting the best images possible is achieved by trying to keep all aberrations to a minimum . You are going to have seeing , optical mounting problems , mis-collimation and and local thermal issues all in the `wobbly stack ' from star to image plane, even assuming perfect optics . But I don't think letting optical aberrations loose to the theoretical limits of your obstruction will always yield maximum quality - I think it is prudent to try and build some headroom into all aspects of the telescope.

Professional instruments with large obstructions and sites with no better than 1 or 0.5 arc second could probably specify their optics to a number of waves if they were to let aberrations slide to equal the best case blur spot for a given seeing condition , but they generally try to achieve better than 1/20 wave because they can and it is a form of future proofing.

Most of the best imaging around is done by amateurs who have paid attention to every aspect of their setup to try and keep their signal to noise as good as possible . They are generally not taken with cheap optics , cheap mounts etc . Obviously chasing or paying for 1/100 wave optics would be a hollow pursuit , but I think good headroom means optics of 1/8 wave front or better where the geometrical ray bundle meets at about half the diameter of the theretical airy disc. It would be hard to pinpoint exactly how all the improvements are made , but I'm guessing its the some total of a lot of subtle gains in every aspect of the instrument .The anecdotal evidence I have seen from imagers using known quality optics suggest to me that quality it is not an empty thing to strive for in the pursuit of fine images. Headroom in optical quality is important too.

What is the program that allows you to compare the obstruction
with the theoretical Dawes limit or Rayleigh criterion?

[Satchmo]

http://www.iceinspace.com.au/forum/s...uality+imaging

Glend,Alpal, al the simulations you need to see are presented in this thread frpm 2012 on just this topic, with major contributions from all the usual suspects. :-) The end of the thread is probably the most useful.

Regardless I note that the best images on this forum are taken by guys like Rolf, Sidoneo, Fitz Henry, all using known high quality optics. Whether it is because they treat all aspects of the train with equal respect I dont know. I am a visual guy, and I know for sure that everything helps there, and also for the planetary guys doing short exposures .

Glend: you would probably benefit most from upgrading your mount and tracking capability than worrying about the improvemnts of ion milled optics.

[glend (Glen)]

Originally Posted by Satchmo

http://www.iceinspace.com.au/forum/s...uality+imaging

Glend,Alpal, al the simulations you need to see are presented in this thread frpm 2012 on just this topic, with major contributions from all the usual suspects. :-) The end of the thread is probably the most useful.

Regardless I note that the best images on this forum are taken by guys like Rolf, Sidoneo, Fitz Henry, all using known high quality optics. Whether it is because they treat all aspects of the train with equal respect I dont know. I am a visual guy, and I know for sure that everything helps there, and also for the planetary guys doing short exposures .

Glend: you would probably benefit most from upgrading your mount and tracking capability than worrying about the improvemnts of ion milled optics.

Thanks for that Mark. I will confess that I am not really trying to produce images that compete with Rolf, Sidoneo, Fitz Henry. I am just trying to get the best performance out of my admittedly budget imaging system as possible. Imaging newts are cost effective and can produce good results, certainly I am happy with my images so far and they are getting better - but not to the point where I would want to post them up here on IIS and run the gauntlet of the critics. So much of imaging lies in the post-capture processing side and that's where I need more work.

I didn't know I had a mount problem. I run it on a new NEQ6Pro, and the combined imaging load is well under the mount payload capacity. I keep the legs short (not extended at all) and it sits on indented pavers buried in the ground. I always make sure to run the Polar alignment routine in the new Synscan V3.35 and it is correct to less than a few seconds. I run wide dovetail bars top and bottom of the scope secured to the cast rings spaced 13" apart to stiffen up the tube support structure. The scope and gear is all balanced on the mount in both axis. All of my imaging is guided now, using an ZWO ASI130MM camera on an Orion Miniguider scope set in the middle of the top dovetail bar. I use Metaguide for guiding and it works great with the ASI camera, it will hold on the target star very tightly for hours at a time. My 8" GSO imaging newt is an f5 (picked to reduce coma effects), with 1000mm focal length, with upgrade mirror springs, flocking, a light shield, and the collimation seems absolutely stable in all the extreme angles of operation (I've checked that). I run a GSO Coma Corrector with spacers to hit my camera sensor at the right distance, and I have no trouble with focus (use a Bathtinov Mask). The imaging train on the focuser does not flex, or seem to, and I keep it as sort as possible. I do run the stock GSO Crayford style focuser(which is lockable) but plan to move to a Moonlight in the future. Not sure what more I can do with a fairly simple Canon 450D DSLR setup.
Which is why I originally asked about the mirror.

[bird (Anthony Wesley)]

I presume most of the imaging being discussed here is low power / wide field - ie I can't see anyone above talking about the optics needed for high-res planetary work...

I have a 16" f/4 newtonian and optical quality matters a lot to me :-) See this image for what is possible in good seeing, you won't get that with low quality optics...

http://www.acquerra.com.au/astro/gal...-174959utc.png

Jupiter was about 50 arcsec in diameter here, you're welcome to measure the image to see what resolving power it has. At the time this was taken I had a 14.5" f/5 mirror in my newtonian, and central obstruction about 28%.

cheers, Bird