It appears that no one can agree on what a SEMI-Apo telescope is.
If you read any advert doublets (apo ones, ???) such as the Skywatcher ED and Takahashi FS-60 as just two of many hundreds of examples are "APOchromats", and yet there are arguments that a real APO is a triplet.
From other people
One comment is "A semi-apo tends to be a doublet (achro) with "fancy" glass. FPL-51 or FPL-53 generally.
2 wavelengths are defined for the focal length usually Red and Blue but that allows the Green to drift off from the ideal. It is the amount of this "drift" that they claim makes it semi-apo. In an ED scope this difference is small, and hopefully small enough that the eye does not detect it, so it looks CA free."
Another is
"There is no formal definition of a "semi-apochromatic.""
Another comment
"The Red and Blue are at 0 whereas the Green has a difference of about 0.005 as shown
Now if that were for a normal achro then a semi-apo would have a difference of say 0.002 or 0.003, but less then an achro, however still more then an apo which should be Red=Blue=Green, so 0 difference."
"Manufacturers claiming semi-apo are fine, they at least state what you are buying, but there a a few that stated apo for what is a doublet and so cannot be apo.. "
So are manufacturers misleading us or have we become accustomed to believing a doublet with rare glass can be an APO.
Therefore do I have one proper APO and one .......... Semi APO or two APOS
Call them doublet and triplet, stop worrying about it, relax and enjoy your scopes when the weather permits !
I agree. Many amateurs tend to be very technical and forget to enjoy the sky.
I also have not the best scopes (i.e. triplets or fluorite / FPL53), but I have a 'quick view' tabletop 80mm achromat which performs very well. And a 110mm ED doublet travel scope which performs even better, and color errors are only discernable at very bright objects like Venus and (barely) with Jupiter. On astrophotos the fringes are nonexistent.
So for me spending A$10k for an FSQ-106 is a waste of money.
I’d say there is more to it than whether it’s a doublet or a triplet. I have a 100 F/9 doublet that has better colour correction than some 127mm F/7 triplets out there. Having 1” less aperture helps with colour correction, being F/9 as opposed to F/7, being a well crafted fluorite doublet helps compared to a cheaper FPL-51 triplet.
By a formal definition an APO can have visible CA which is what blurs the line between APO and Semi-APO.
I know there are a lot of factors that need to be accounted for, but my experience with imaging with small refractors taught me that there can be a significant difference between f/7 ED doublet and f/6 triplet when it comes to RGB imaging.
There ARE definitions of an APO telescope. The "original" one, by Ernst Abbe (Zeiss), and a more modern one championed by late Thomas Back. Google it.
"An objective in which the wave aberrations do not exceed 1/4 wave optical path difference (OPD) in the spectral range from C (6563A - red) to F (4861A - blue), while the g wavelength (4358A - violet) is 1/2 wave OPD or better, has three widely spaced zero color crossings and is corrected for coma.
By whichever one you go, it is definitely possible to make an APO using just two lenses. That is, three colors crossing, and correction for spherical at two wavelengths."
What you CANNOT do, is have a fast, well corrected APO with only 2 lenses.
Here's a 6" APO using only two (extreme) types of glass with FOUR color crossings (no Fluorite needed ).
Note:
is f/14, and one of the crossings is in IR and another in near UV.
But by all definitions this doublet is better than APO.
I can give you at least half a dozen designs of doublets with 3 color crossings, some of them can be as fast as f/8. And if we reduce aperture, they can be even faster. What 3rd lens allows us is to make lenses less steep/curved (easier to make, keep collimated, and less spherochromatism), and also faster for same amount of color error.
PS you should stop reading what "other people" say in popular groups. There's plenty of proper, peer reviewed information available on the web, as well in old fashioned books (gulp). Look up Vladimir Sacek, Roger Ceragioli, Harrie Rutten & Martin van Venrooij. There you will actually find a "formal" definition of a semi-apo (based on color error). Its roots have also been established by Zeiss.
Telephoto lenses such as the Canon 300/2.8L, 400/2.8L and 600/4L have more than ten elements and then it is possible to make color-free images with such a short focal length.
Telephoto lenses such as the Canon 300/2.8L, 400/2.8L and 600/4L have more than ten elements and then it is possible to make color-free images with such a short focal length.
Apples and oranges. Telephoto lenses have to perform autofocus (FAST), do image stabilization (also fast!), and be able to stop down at a flick of a switch. And be hand-holdable. None of that applies to telescopes.
BTW, I doubt that any of the telephotos get anywhere near the correction of a wold class APO. Both color or spherical aberration wise. Try your 600mm f/4 L at 300X on Jupiter and report back (which is dead easy with a good 6" APO).
Bratislav, you’re right in one respect - the resolution of camera lenses isn’t diffraction-limited - or does it need to be - there are a whole raft of conflicting imperatives which lead to a trade-off (ie a balanced compromise) in the lens design:
- wide aperture (ie f ratio under 4 and in some cases 0.95);
- minimal vignetting and distortion over the field of view, or else correct these in software;
- flat focal plane to match the sensor;
- reasonable image quality (MTF) over the whole frame out to the corners which means chromatic and monochromatic aberrations must be controlled in a balanced way.
Hence 10 or more elements are not unusual. The issues of focusing, zooming, weight and variable aperture have no bearing on the lens design.
A telescope design becomes far simpler fundamentally because the sheer cost of much larger aperture glass dictates fewer elements and the rest - including choice of focal ratio follows from that. Or switching to a mirror.
Perkins Elmer did for example produce a 5-element lens 20cm clear aperture f/5, though in current $ you wouldn’t have much change out of $1M, and it required a crane and truck to lift/move it.
The cost and difficulty in making optical blanks of sufficient perfection is what drives the cost of lenses 20cm and larger - take a look at the APM website to see what I mean - they are astronomical, literally.
The issues of focusing, zooming, weight and variable aperture have no bearing on the lens design.
You are quite wrong. Issues of focusing, weight, image stabilization and variable aperture have a DIRECT BEARING on lens design.
I'm happy to take this in another thread as it has nothing to do with original posting, but I have designed quite a few telescopes based on lens, mirror and compound varieties and know well what can and cannot be done. For example, most telescopes do not move any part of their optics during focus (they move sensor), or if they do (like in SCT for example) they move just about the heaviest part. Telephotos have to be designed with having a cardinal plane defined where beam is narrow (so any active elements like focusing group, image stabilization group or stop down mechanism - iris - can be small in order to be able to move fast). Also, telephotos must reduce the beam as quickly as possible after entering for the same reason, so all subsequent potentially very expensive elements that are extreme dispersion glass, Fluorite and/or aspheric, can also be small and cheap. Telescopes do not have to do any of that. So we have designs like Busack/Riccardi/Wiedemann/Honders that are better corrected over larger area than those super-telephotos, and are as fast (or faster), that is have lower f/number (some can be pushed to f/2 or faster) and do not require 10 elements. But they cannot do IS, cannot quickly stop down and cannot do autofocus 20 times a second.
Trust me, if we were allowed to use same/similar size elements in a modern telephoto lens, their performance could be vastly improved without having to resort to 10+ elements. Heck, I can design a Petzval with just about 5 or so lenses at f/4 that will leave any 600/4 L (or Nikon ED) in the dust. But all lenses in it would be large and heavy, so no IS, no quick AF, body would have to be much longer and much heavier, and finally it would be a lot more expensive with several full size elements out of FPL glass or Fluorite, and others from just little bit cheaper Lanthanum glass).
Do you think it would sell?
No, it goes the other way around. They have established themselves as big players first, by having best telephoto lenses around (and bodies too!), and THEN people flocked.
If I designed 40kg, 1m long 600mm f/4 "supertelephoto" that has to be manually focused, no IS and no possibility to do shutter priority (stop down quickly) and cost better part of 30K nobody would buy it, Canon or Nikon labels notwithstanding.
Those are completely separate markets. Noone uses their Canon 600/4 to observe super close double stars, neither wildlife/sports pro's lug 6" Astrophysics or Takahashi APOs on ME-II's or 1600's around. Even if they are much sharper and much more stable.
Apples and oranges.
...
Perkins Elmer did for example produce a 5-element lens 20cm clear aperture f/5, though in current $ you wouldn’t have much change out of $1M, and it required a crane and truck to lift/move it.....
.
I have one of those . Well part of one - a 10kg or so 275mm diameter planoconvex lens element from the Perkin Elmer 916mm focal length f/4 surveillance camera used in spy/surveillance aircraft and satellites.
I have one of those . Well part of one - a 10kg or so 275mm diameter planoconvex lens element from the Perkin Elmer 916mm focal length f/4 surveillance camera used in spy/surveillance aircraft and satellites.
Just waiting for an interesting project......
Best
JA
Those old cameras were designed for film, and would be rather mediocre by today's standards.
BTW, this is the illustration on what I'm talking about. Left is the MTF (10 to 50 line pairs per mm) of SONY 135mm f/1.8 GM lens that Lens Rentals (Roger Cicala) have found to be the sharpest they have EVER tested. (and they have tested a lot, including many CaNikons )
To the right are MTF curves, also 10 to 50 lp/mm of a 450mm f/1.8 (yes, that is 250mm aperture (!)) Strange-Jones astrograph. It has only 4 elements (3 lenses and a mirror), and uses ordinary BK7 glass. It also has zero vignetting, near zero distortion, and it has perfectly flat field.
But it can't do IS, can't be stopped down instantly and can't autofocus.
Those old cameras were designed for film, and would be rather mediocre by today's standards.
Hi B,
Whilst I'm not wedded to anything (film or digital) I'd just comment that:
Yes of course they were designed for film, it was the film era. It's just that, THAT FILM was a whole lot larger than the typically used 35mm film. (121 times the surface area of normal 35mm film). I feel that despite the passage of time, that given the application to which they were applied, their precision and the expense involved in their design and manufacture that they may be of some use today. Objectively I couldn't say without any testing, but my thoughts/leanings follow.
Given that no expense was spared on these spy/surveillance systems there is every reason to expect that the film used would also have been special and certainly likely to be up to the task, certainly as good or better than commercial film stock. Given the increase in film size and that the film used would have been equal or better quality to readily available commercial film stock, that would mean a significant increase in observable detail/sharpness (resolution+contrast) on a per picture height or width basis compared with the commercial consumer cameras of the time.
Without some form of objective test it would be difficult to say by just how much the sharpness improved, but given that the film used was 9 x 18 inch surveillance film, compared with typical 35mm or smaller film. That's 121 times the area or between 9.5 to 13 times larger in height and width. Ergo from a sensor (film) standpoint one could expect something probably like 10 times the resolution on a per picture height/width basis compared with typical commercial film systems of the time based on 35mm film. Of course the lens then had to play its part.
Leaving this extra large spy film aside, I would certainly be guided by the increased resolution available on a per picture height/width basis between say a 35mm film camera system and larger format film camera systems, as demonstrating the sort of sharpness (Resolution + Contrast) improvements possible with larger sensor (film) size camera systems and that should be observable in equal sized output images (of sufficient resolution) from such systems.
Quote:
Originally Posted by bratislav
To the right are MTF curves, also 10 to 50 lp/mm of a 450mm f/1.8 (yes, that is 250mm aperture (!)) Strange-Jones astrograph. It has only 4 elements (3 lenses and a mirror), and uses ordinary BK7 glass. It also has zero vignetting, near zero distortion, and it has perfectly flat field.
But it can't do IS, can't be stopped down instantly and can't autofocus.
Do you have a link/pics etc or some further info on that design?
Strange-Jones (or Jones-Strange) is a relatively new design, just like Wiedemann/Busack/Riccardi/Honders medials (even newer). I think it was made possible with advancements of computer optimization routines and is deceptively simple. It is loosely related to Schmidt. Two large lenses, one mirror and one small lens. It can also be made into a Cassegrainian form (in which case of course can't be as fast, but still very fast, as fast as f/3.5 ). In Schmidt-like configuration it is easy to make them as large as 500mm at insane f/1.5, still diffraction limited over a huge area (much larger would not be practical as lenses are thick).
It can cover larger area much larger than 35mm (I've designed 300mm f/3 covering >100mm circle with 3 micron spots). If you factor in the resolution, and the fact that these systems work over enormous spectrum (400nm to over 1000nm), I think you will find out that they are actually capable of resolving more pixels per field than those 1M$ Perkin Elmers. That is, if we could find a CCD/CMOS chip that is large enough, with small enough pixels (and doesn't cost as much as Sydney Opera House).
I'm not aware of any web sites covering them, you have to keep up with Joneses (Ed Jones and Mike Jones that is ).
I have designed many of those in Zemax, but not aware of anyone having them made (yet).
Well it really depends on how exactly or inexactly one cares to define an apochromat:
Is it defined by
1. the number of lenses in the design or
2. by how well it corrects for aberration (typically spherical and chromatic) ?
Some would say that
1. if it has 3 zero crossings on the focus v wavelength curve (3rd order cubic function) then it's a triplet and apochromat and
2. if it has 2 zero crossings on the focus v wavelength curve (2nd order parabolic function) then it's a doublet and achromat
3. if it has 4 zero crossings on the focus v wavelength curve (4th order quartic function) then it's a superachromat
Others simply say that the apochromat corrects better than the achromat and leave that as some sort of definition.
To me the optical performance matters more than the specific design or moniker conferred on a design. "Back in the day" the achromat was taken to mean a optic without colour errors/fringes/abberation and it certainly would have appeared so in comparison with more commonly used singlet lenses of similar focal length/aperture at the time. Enter a new design that has better correction still for colour and spherical abberation and it's differentiated by calling it a apochromat, typically these were triplets.
Just because something is called an apochromat doesn't per say make it better than an achromat, even though it may be that a particular design usually confers a notional performance advantage. Only objective optical performance matters
- a 10kg or so 275mm diameter planoconvex lens element from the Perkin Elmer 916mm focal length f/4 surveillance camera used in spy/surveillance aircraft and satellites. 
) 
