FORUM

Apochromatic vs Achromatic colour correction

by

Chris Lord



A comparison of standard colour correction curves for triplet apochromats, doublet semi-apochromats and achromatic doublets
by Chris Lord


spectral comparison @ 80% strehl ratio

Ref: Mardina Clark: comparative colour correction of apochromats and achromatcs
ATMOS simulated diffraction image


































simulated visual image TEC ED140simulated diffraction pattern 6-inch f/15 doublet

A comparison of standard colour correction curves for triplet apochromats, doublet semi-apochromats and achromatic doublets

Speaking as the proud owner of the best mid-aperture apochromatic refractor in its class, the TEC140APO, I am occasionally irritated by apo users who have never used an achromatic refractor yet denounce achromatics for their poor colour correction. I have engaged in arguments where the proponents of apochromats claim wildly exaggerated reductions in longitudinal colour shift compared to a standard achromatic doublet. Reductions in colour shift of a million fold between a 6-inch triplet apo & a 6-inch doublet achromatic. It is however a fact that most of the improved colour correction achieved by adding an additional lens or employing extra low dispersion glass, is used to reduce the objective focal length. The resulting colour shift is only marginally shorter than that of a standard achromatic.


The usual deceipt resorted to by some apo makers is to make a direct comparison. Take for instance this ATMOS comparison between a 100mm f/10 triplet apo and a 100mm f/10 standard flint-crown doublet. Another deceipt is to make the comparison between the r (deep red) and g (violet) lines rather than the more favourable visual limits C (H-alpha) and F (H-beta) lines.

chromatic colour shift comparison
The C-F colour shift of a standard doublet is 0.05%f compared with 0.02%f for an oiled triplet apo. Since we are comparing the visual appearance of secondary spectrum (unfocused colours), the colour error beyond normal visual limits is of no significance. The ability to see colour requires photopic vision. Visual sensitivities relative to peak sensitivity @ Å5500 at the following Fraunhofer lines are as follows (extracted from Southall's Physiological Optics):
C Å6563 8%
e Å5460 99%
F Å4861 14%
g Å4360 0.6%

and to quote from Roger Ceraglioli's primer, 'A Survey of Refractive Systems for Astronomical Telescopes', Ch.2, para 29, "The violet line represents the g-spectral line at 0.436 micron ... marks the approximate limit of human vision at shorter wavelengths."

What this means in practice is that a 5.5-inch f/15 Fraunhofer achromatic doublet exhibits a colour error only 2.5 times greater than my TEC140 f/7 oiled triplet. The ratio over C-g is 0.24%f to 0.03%f or 8:1. The fact that you cannot possibly see starlight so deep in the violet seems to be immaterial to the purblind devotees of apos.

It is a pity no one nowadays is prepared to make a triplet apo with a similar focal ratio to a standard doublet, because if they did they just might achieve a similar specification to the best apochromatic triplet I have ever looked through, the 6-inch f/18 Taylor-Cooke at Calton Hill Observatory, Edinburgh. Dennis Taylor published his experiments in Monthly Notices 54,2,67 (also Grubb's publication - The Adjustment & Testing of Telescope Objectives - 4th Ed. 1946) & MN54,5,328. The CA for the 6-inch f/18 triplet is @ C fr = +0.00002fy; @ F fb = +0.000013fy, which is an order of magnitude superior to the TEC ED140 f/7 oiled triplet. Also Dennis Taylor's apochromatic objective was a true apochromat, not a mere aporoximation.

I can hear the howls of protestation as I write this sentence. Unless an objective conforms to Abbe's definition of apochromatism, it is not a true apochromat. Abbe defined an apochromatic objective as being one corrected parfocally at three widely separated wavelengths across the visual spectrum whilst also being corrected for spherical aberration and coma at two widely separated visual wavelengths. In fact he went further, and stated that one of the crossing points for colour correction should also coincide with one of the crossing points for spherical aberration and coma. Ideally for a visual objective that crossing point should lie as close as feasible to the e-line. It is a tall order, and not a single modern so-called apochromatic objective comes anywhere close to meeting it.

The late Tom Back provides his reasons for deviating from the Abbe criterion for apochromatism [defining apochromatism]. His reasoning seems sensible bearing in mind manufacturing practicalities and market realities (few nowadays want a long telescope because it isn't portable, gets buffetted unduly by the wind, and so needs a far heavier mount). I however find his arguments a tad self serving. What he is basically arguing is that Abbe's criterion is too strict, and the object glasses he (& others) make are just as good, if not better. His argument might hold if it weren't for the irksome precedent of Dennis Taylor's apochromatic objective which did meet Abbe's criterion.

Here are three typical apo colour curves, one for an oiled triplet apo, one for an air-spaced triplet apo, and one for an air-spaced doublet Super-Apo ("Super APO" as first coined by James G. Baker is any lens with four colour crossings. The doublet would be better described as a "Semi-APO" which according to Zeiss' original useage was a lens with half the correction of a standard achromatic, hence the Zeiss "Halb Apochromat" AS lens, contemporaneous with H. Dennis Taylor's true apochromats. Sadly terms like "Super APO" & "Semi-APO" have been appropriated and mis-applied to the point where they are yet another example of manufacturer's advertising gobbledygook and both now presumably mean whatever the maker wants them to mean).

TEC ED140 colour error diagmegrez80 colour error diag






















Notice if you will both triplets achieve only two crossings at the e-line and the doublet only one. Yuri Petrunin commented in May 2007 that the Takahashi TOA series achieved apochromatism without a single crossing. Takahashi do not publish colour error diagrams but I have no reason to doubt Yuri. Even with the benefits of ED & Fluorocrown glasses, it is simply not practicable to obtain three crossing points in the red, green and blue, whilst keeping spherochromatism within ±1/4 wave at two wavenlengths. The only practical way of designing a fast OG (f/5 to f/7) that meets the Abbe criterion is to use four elements, a design invented by Zeiss and used by Hassleblad in their hyper-achromatic telephoto lenses, or to do as Roland Christen has done and use a fluortie element in an EDF f/7 triplet AP Starfire160EDF f/7 triplet.

To summarise, what we have here is a re-writing of telescope making history. Companies intentionally manipulating accepted optical standards and definitions. According to the apophyles who covet these telescopes, all refractors prior to recent ED & Fluorocrown triplets suffer from unacceptable secondary spectrum, whereas Ed & Fluorocrown triplets exhibit essentially zero colour error. Yet what do we find? The apo's they so fervently champion are mostly not true apo's at all, and the secondary spectrum exhibited by a standard doublet is only 2.5 times that of a fast triplet "apo". In their over-ebullient enthusiasm for their respective manufacturing heroes, all perspective is lost. The principal advantage of an apo like my TEC ED140 is its portability. A 6-inch f/15 refractor is a big, and necessarily heavy telescope. It needs a permanent mounting and an observatory. A TEC ED140 does not. To claim that no useful observations can be made with a 6-inch f/15 achromat as many apophyles do is a canard, as any proud Unitron owner will testify. And when I hear neophytes deriding owners of such marvellous refractors as "old school" my response is simple, "time for you to get back to school, you have a lot to learn."






This page was created by SimpleText2Html 1.0.2 on 09-Dec-2007.