ICEINSPACE
Moon Phase
CURRENT MOON
New Moon 0.3%
|
|
23-11-2013, 08:43 AM
|
|
Registered User
|
|
Join Date: Dec 2005
Location: Sydney
Posts: 1,879
|
|
Quote:
Originally Posted by alpal
|
Thats nice advertising about what Ion milling is capable of in a research situation .
Have a look at the link on that page showing final ion milled results of an RC 24" - P-V is 1/10 wave - this would probably have been around 1/4 wave if only rotational symmetric polishing had been carried out ( easy enough to do with pitch polishing )
|
23-11-2013, 08:45 AM
|
|
Registered User
|
|
Join Date: Dec 2005
Location: Sydney
Posts: 1,879
|
|
Quote:
Originally Posted by Peter Ward
So I guess I'm one of those deluded nutters who waited years and parted with some serious bucks...
|
Nice one
Playing the devils advocate are there any comparison shots available with the competing gear that show spending all that extra money made a difference to the final quality of your image? Just asking
Last edited by Satchmo; 23-11-2013 at 09:18 AM.
|
23-11-2013, 09:09 AM
|
|
Registered User
|
|
Join Date: Jan 2012
Location: Melbourne
Posts: 3,695
|
|
Quote:
Originally Posted by Satchmo
Thats nice advertising about what Ion milling is capable of in a research situation .
Have a look at the link on that page showing final ion milled results of an RC 24" - P-V is 1/10 wave - this would probably have been around 1/4 wave if only rotational symmetric polishing had been carried out ( easy enough to do with pitch polishing )
|
Actually here:
http://www.rcopticalsystems.com/tele...ogram_JN07.pdf
it says:
RMS = 0.024 ( = 1/42 wave )
Peak to valley = 0.1 wave
Remembering that the reflected wave is only half that value.
Still - I think it's a pity that RCOS went out of business.
Maybe ion milled optic telescopes will become collectors items?
|
23-11-2013, 09:25 AM
|
|
Registered User
|
|
Join Date: Dec 2005
Location: Sydney
Posts: 1,879
|
|
Al - I suspected when you were throwing the figure of 1/20 and 1/100 wave around that you weren't talking RMS
I'm not sure what you meant about reflected beam being half that .
Those figures would be final focus wavefront error.
There is nothing particularly unusual about a 1/10 PV - 1/40 RMS surface for what I would consider a diffraction limited optic , but on a large RC with highly aspheric surfaces thats a very good result. I think Star Instruments surfaces were typically 1/4 PV wavfront .
|
23-11-2013, 09:40 AM
|
|
Registered User
|
|
Join Date: Jan 2012
Location: Melbourne
Posts: 3,695
|
|
Quote:
Originally Posted by Satchmo
Al - I suspected when you were throwing the figure of 1/20 and 1/100 wave around that you weren't talking RMS
I'm not sure what you meant about reflected beam being half that .
Those figures would be final focus wavefront error.
There is nothing particularly unusual about a 1/10 PV - 1/40 RMS surface for what I would consider a diffraction limited optic , but on a large RC with highly aspheric surfaces thats a very good result. I think Star Instruments surfaces were typically 1/4 PV wavfront .
|
It's not clear that they mean -
"final focus wave front error"
If you have a 1/20 wave mirror then it reflects at 1/10 wave accuracy.
If it then hits a 1/20th wave secondary then the final wave front is 1/5th accuracy.
|
23-11-2013, 10:38 AM
|
|
Registered User
|
|
Join Date: Dec 2005
Location: Sydney
Posts: 1,879
|
|
An interferogram analysis of the quality of an imaging optical system is always supposed to be presented as wavefront error unless otherwise specified . The setting in the analysis software of the 'waves per fringe spacing' determines that the results are correctly presented and quantified .
In the case of this RC opticsal auto-collimation tested system the spacing would have been 0.5 waves per fringe. If you ordered a flat you would expect the results to be expressed as a surface error. It is possible for an optician to fiddle this aspect by setting a less sensitive wave per fringe factor than is technically correct for the testing set up , thus making the optics appear to be better than it really is . It happens.
|
23-11-2013, 11:07 AM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
Quote:
Originally Posted by Peter Ward
Well...
I look at it this way....if an optical system is, to coin an old chestnut, is barely "diffraction limited" then *even the slightest* seeing distortion, by definition makes the images it produces, poorer than that limit.
If on the other hand, the optical system is better than "diffraction limited"
(yes... this is possible) it can suck up, for example, a 1/6th wave of atmospheric distortion before *total aberrations* take it into the 1/4 wave "diffraction limit"
Hope that makes sense
|
Peter,
it would be relatively easy to demonstrate the currency of your argument (that the AP RH) produces an intrinsically sharper image at focus compared to a Newtonian or SC, by providing a planetary image taken with it so we can do a real world comparison.
-c
|
23-11-2013, 11:53 AM
|
|
avandonk
|
|
Join Date: Aug 2005
Location: Melbourne
Posts: 4,786
|
|
Completely different scenario. Planetary imagers are just recording the tiny peaks of the Airy Disc. They are recording quite bright images that can be temporally separated.
At f40 the Airy disc is about 50 micron. Please explain!
Bert
Quote:
Originally Posted by Satchmo
To be fair the error on Hubble mirror was 1/2 wave RMS - much greater than you will find on a lower end typical consumer scope .
There has been a thread on this before - but planetary imagers need focal lengths of 5 to 15 meters to literally oversample the airy pattern with pixels to exploit every bit of quality in their optics and detail possible in the image. Its all to do with pixel scale versus airy disc size vs. seeing . I think there would be few deep sky imaging rigs exploiting optical quality fully , but I do feel that the better the optics , the less havoc tilt and defocus of the wavefront due to seeing , will have on the FWHM. There is no doubt that the guys with known good optics that are well mounted and guided are coming up with the premiun images.
|
|
23-11-2013, 12:20 PM
|
|
Registered User
|
|
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
|
|
Hi Bert
at f25 the FWHM of the Airy pattern is 14microns - this sort of focal number is used with 5.6 micron pixels for planetary imaging
at f19 the FWHM is 10.7 microns, used with 3.75 micron pixels
basically, planetary imaging uses ~Nyquist sampling on the optics PSF, not the atmospheric PSF as used for DSO - that is the main difference.
Regards Ray
Last edited by Shiraz; 23-11-2013 at 02:59 PM.
|
23-11-2013, 02:52 PM
|
|
Registered User
|
|
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
|
|
Quote:
Originally Posted by gregbradley
I agree with Peter. He has expressed better than me but he certainly is pointing out the phenomena I have observed which was my point.
His RHA image says it all. You can't get those tiny stars with other gear. The other scope I have seen tiny stars like that is the infamous ASA 12. When it works its a tight machine.
There is apart from seeing and distortions the factor of scatter.
If your argument were true then there would be no point in getting those Ion Milled RCOS's that essentially had perfect figure. Yet have a look at those Rob Gendler images taken on a 14.5 inch RCOS with Ion Milled optics in WA on his astropic site.
They have very tight stars, show detail in commonly imaged southern objects you just don't see on images from this site except Martin's from Sierra Nevada.
So there is obviously another factor or factors that come into play than simply seeing limited, 1/4 wave limited etc. Scatter is one, sharp focus as Peter points out is another. Having had many scopes some definitely snap to focus better than others. Some have sharp focus all across the field and others super sharp in the centre and not so much in the corners etc. Addded to that would be deformations in the performance of the mirror/lens with different temperatures (Paul Haese and Bird can tell you about the importance of cooled optics) as well as deformations caused by stress from a lack of proper mirror support in the scope (as Rick Hedrick from Planewave was telling me, your 1/10th wave mirror once mounted may be way less than you think). So mechanical issues as above become very important factors in getting that sublime optical system which is quite a rare instrument. I guess that is why the high end stuff often seems overengineered like Tak scopes, AP scopes, etc.
Greg.
|
hi again Greg.
Tight stars do not of themselves indicate high resolution. Sure the optics must be good enough for seeing limited performance, but the star tightness is then mainly a function of the sampling. I think that Peter gets about 1.63 arcsecond pixels when he uses a 16803 with the short focal length RH, which is undersampling in most conditions. A star that would cover say 4 pixels in Peter's system (at 1.63 arcsec) will cover about 25 pixels in your CDK/16803 system (sampled at your system scale of about 0.64 arc sec). With the 694 on your CDK, the same star will cover about 100 pixels - so Peter's stars will look tiny compared to yours, as they should - not because of any optics magic, but because of sampling.
One can certainly get tighter looking stars (and a huge increase in sensitivity) by undersampling, but at the cost of not recording the full detail in extended objects. Undersampling is certainly a valid strategy though if you want to image large faint targets - eg as Bert does.
agree that mechanical quality appears to be a common characteristic of larger high end astrographs - and I would guess that it is very often the difference between an optical system being seeing limited or being a poor performer.
Regards Ray
Last edited by Shiraz; 23-11-2013 at 09:24 PM.
|
23-11-2013, 11:50 PM
|
|
Registered User
|
|
Join Date: Feb 2006
Location: Sydney
Posts: 18,032
|
|
Quote:
Originally Posted by Shiraz
hi again Greg.
Tight stars do not of themselves indicate high resolution. Sure the optics must be good enough for seeing limited performance, but the star tightness is then mainly a function of the sampling. I think that Peter gets about 1.63 arcsecond pixels when he uses a 16803 with the short focal length RH, which is undersampling in most conditions. A star that would cover say 4 pixels in Peter's system (at 1.63 arcsec) will cover about 25 pixels in your CDK/16803 system (sampled at your system scale of about 0.64 arc sec). With the 694 on your CDK, the same star will cover about 100 pixels - so Peter's stars will look tiny compared to yours, as they should - not because of any optics magic, but because of sampling.
One can certainly get tighter looking stars (and a huge increase in sensitivity) by undersampling, but at the cost of not recording the full detail in extended objects. Undersampling is certainly a valid strategy though if you want to image large faint targets - eg as Bert does.
agree that mechanical quality appears to be a common characteristic of larger high end astrographs - and I would guess that it is very often the difference between an optical system being seeing limited or being a poor performer.
Regards Ray
|
I see. I can see the sampling being part of that. But if I imaged that same scene with my high quality APOs of around the same focal length I don't think I would get stars that small. So the accuracy of the optics and that optical system is having some effect there.
Greg.
|
24-11-2013, 10:29 AM
|
|
Galaxy hitchhiking guide
|
|
Join Date: Dec 2007
Location: The Shire
Posts: 8,335
|
|
Quote:
Originally Posted by Shiraz
hi again Greg.
Tight stars do not of themselves indicate high resolution. Sure the optics must be good enough for seeing limited performance, but the star tightness is then mainly a function of the sampling......
Regards Ray
|
You conveniently ignore the physical size of the airy disk with this approach.
Airy disk size (assuming perfect optics) is purely a function of F-ratio.
At F8 the light is being spread over 10 microns. At F3.8 this contracts to 4.7 microns.
Not withstanding sampling, assuming the same aperture, you are concentrating the same flux into fewer pixels....
|
24-11-2013, 11:01 AM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
I think it also worth adding to the discussion that field correction is a distinct issue separate from optical quality... if the component added to the image blur circle size by virtue of optical quality is below the sampling limit (on axis) then it (optical quality defects) will swell the image blur circle more or less the same at the edge of field.... the caveat to this is the usually unstated assumption that the location of the optical defect is coincident with the field stop. Unless you are talking about full aperture correctors located at the centre of curvature of the primary (lichtenecker, baker schmidt etc) then this is pretty much a given.
on somewhat of a tangent... but not totally unrelated... if you put the aperture stop at the centre of curvature of the primary mirror (ie, at twice the focal length) and leave the primary spherical, the system will have no coma or astigmatism...at all! The only issues with it would be spherical aberration which would limit the speed to about f8 for a 200mm aperture, you would have field curvature (1600mm radius) and the system would be physically enormous.
It is however a very useful hypothetical instrument to help you understand the true nature of coma and astigmatism in geometric optics....
c
|
24-11-2013, 11:10 AM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
Quote:
Originally Posted by Peter Ward
You conveniently ignore the physical size of the airy disk with this approach.
Airy disk size (assuming perfect optics) is purely a function of F-ratio.
At F8 the light is being spread over 10 microns. At F3.8 this contracts to 4.7 microns.
Not withstanding sampling, assuming the same aperture, you are concentrating the same flux into fewer pixels....
|
Yep... but you would have to increase the system focal ratio to around f24 (with a typical ccd ) before you are sampling the image at a level where diffraction becomes a sensible limit... this is why you can throw 50 percent of the collected energy into the diffraction rings without suffering any loss of detail at f4.... whether that is in the form of large central obstruction or a moderate amount of sa.... it really doesn't matter as long as the distribution curve is the same.
The design parameters of the hst are an interesting case in point.... when the goal is ultimate deep sky resolution, f24 (you would assume) is ideal coupled with a small secondary obstruction (15%) that doesn't throw a high percentage of its light out of the airy disk. The design parameters of the hst however, don't include an atmospheric component that would otherwise impose a different limit and therefore a different optical configuration optimised for those specific conditions.
Last edited by clive milne; 24-11-2013 at 12:01 PM.
|
24-11-2013, 12:10 PM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
Quote:
Originally Posted by Peter Ward
Airy disk size (assuming perfect optics) is purely a function of F-ratio...
|
At the risk of sounding like pedant.. the airy disk size is proportional to f ratio and the wavelength of light concerned. It also decreases in size as you increase the size of the secondary obstruction.. if, however, the ca gets very large the amount of light concentrated in the airy disk reduces to the point where it is no longer meaningful to ignore the first diffraction ring as being a more sensible measure of the system performance..
|
24-11-2013, 12:23 PM
|
|
Registered User
|
|
Join Date: Jan 2012
Location: Melbourne
Posts: 3,695
|
|
Hi Clive,
what do you think about the manufacturers claim from "Hypergraph"
that it has a spot size 8 times smaller than a Ritchey Chretien?
|
24-11-2013, 02:00 PM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
Quote:
Originally Posted by alpal
Hi Clive,
what do you think about the manufacturers claim from "Hypergraph"
that it has a spot size 8 times smaller than a Ritchey Chretien?
|
It is my opinion that the claim is most likely factually correct however, if it is implied that the hypergraph is 8 times better than an equivalent rc, 'ipso facto' by virtue of its spot diagrams I would suggest that the claim is specious at best....perhaps even disingenuous.
It would be equivalent to me putting a ridiculously high gear ratio into my honda accord and advertising it as being set up to go twice as fast as an Aston Martin at its rev limit.
Theoretical ray tracing is a valuable tool for the optical industry... but context is required before making assumptions about how relevant that information is to any given real world situation. It is important to view the (potential) performance of a telescope in terms of it being just one component of a larger system designed to perform a specific task... ie) convert the light flux from a distant object into a series of numbers proportional to its intensity and hopefully encoding spatial & wavelength data (colour and resolution) in the most cost effective means possible... ie) for a given investment you would hope to achieve the best signal with the least noise... if that isn't the merit function that is applied, then it probably reduces to an exercise in ego gratification... that is perfectly fine as an end in itself but it is better to be honest with oneself if that is actually the case imho.... the choices you make will be different.
if maximising s/n is your goal then the best bang for your buck is always going to be had from a permanently set up telescope at a dark site using a scripted routine on a robotic mount... A simple well made 14" newtonian on a paramount will likely produce more and better data under a dark sky and good seeing than you will ever hope to achieve with a hypergraph from central Melbourne. .. or even one that you have to drag out to the bush 5 nights a year.
best
-c
Last edited by clive milne; 24-11-2013 at 02:18 PM.
|
24-11-2013, 02:55 PM
|
Registered User
|
|
Join Date: May 2011
Location: Freo WA
Posts: 1,443
|
|
On a philosophical note... in matters of opinion solicited in a public forum, you are pretty much guaranteed to come across mutually exclusive answers. A useful mental tool to apply to resolve conflicting conclusions is to rank them in terms of defeasibility.
|
25-11-2013, 08:47 AM
|
|
Registered User
|
|
Join Date: Jan 2012
Location: Melbourne
Posts: 3,695
|
|
Quote:
Originally Posted by clive milne
It is my opinion that the claim is most likely factually correct however, if it is implied that the hypergraph is 8 times better than an equivalent rc, 'ipso facto' by virtue of its spot diagrams I would suggest that the claim is specious at best....perhaps even disingenuous.
It would be equivalent to me putting a ridiculously high gear ratio into my honda accord and advertising it as being set up to go twice as fast as an Aston Martin at its rev limit.
Theoretical ray tracing is a valuable tool for the optical industry... but context is required before making assumptions about how relevant that information is to any given real world situation. It is important to view the (potential) performance of a telescope in terms of it being just one component of a larger system designed to perform a specific task... ie) convert the light flux from a distant object into a series of numbers proportional to its intensity and hopefully encoding spatial & wavelength data (colour and resolution) in the most cost effective means possible... ie) for a given investment you would hope to achieve the best signal with the least noise... if that isn't the merit function that is applied, then it probably reduces to an exercise in ego gratification... that is perfectly fine as an end in itself but it is better to be honest with oneself if that is actually the case imho.... the choices you make will be different.
if maximising s/n is your goal then the best bang for your buck is always going to be had from a permanently set up telescope at a dark site using a scripted routine on a robotic mount... A simple well made 14" newtonian on a paramount will likely produce more and better data under a dark sky and good seeing than you will ever hope to achieve with a hypergraph from central Melbourne. .. or even one that you have to drag out to the bush 5 nights a year.
best
-c
|
I kind of came to the same decision right at the start.
I use a Newt. as the most cost effective way to do imaging.
I have not managed to take any images for over 6 months due to
every moonless night being cloudy.
I am hoping that I'll get a whole string of good nights one day soon
& I'm prepared to travel to get a dark site.
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
All times are GMT +10. The time is now 09:22 AM.
|
|