FORUM

Eyepieces and their effect on image contrast

by

Rodger W. Gordon

The topic of eyepieces is a "hot one" in amateur circles. This article will not deal with wide angle eyepieces except in a round about fashion. I shall confine discussion to eyepieces having an apparent field of view (afov) not exceeding 50° and no more than four or five elements and four air-glass surfaces except for comparison purposes.

The lunar and planetary observer frequently constructs or purchases optics of the highest quality to ensure maximum image contrast, but then often equips his 'scope with eyepieces having 7, 8 or even 9 elements and 8 or more air-glass surfaces.

Since each surface cannot be optically perfect and also since each lens may possess minute internal "artifacts", every additional component degrades the image slightly. Combined scatter and reflections can add up to a noticeable loss of contrast, even where all air-glass surfaces are either magnesium fluoride bloomed or multicoated. Since the contrast differences we wish to detect may be 10% to 5% and less, the fewer the number of elements and air-glass surfaces, the better the chance for seeing such low contrast detail.

Every eyepiece produces at least one double reflection and sometimes two or more. The more double reflections, the greater the chance the eyepiece will have 'ghost' images. These are not only distracting, they can markedly reduce contrast.

The number of potential double reflections can be calculated from the function R = (n-1).n/2 where R is the total potential number of double reflections and n is the number of air-glass surfaces. Cemented surfaces may have to be included if there is a large difference in the refractive index of each element, but I will ignore them.

Let's take a hypothetical eyepiece with three uncemented elements and six air-glass surfaces, labeled a,b,c,d,e,f respectively in the direction of the light path (a to f). Double reflections can arise from surfaces:

reflection paths
ab, ac, ad, ae, af
bc, bd, be bf
cd, ce, cf
de, df
ef
which totals to 15.
Using the expression (n-1).n/2
where n = 6,
R = (6-1).6/2 = 5 x 3 = 15.
If we increase n to 8,
R = (8-1).8/2 = 7 x 4 = 28.
The addition of another pair of air-glass surfaces almost doubles the number of potential double reflections!

Now consider a four air-glass type (Orthoscopic, Plossl &c)
and a two air-glass surface type (Tolles, Monocentric, Hastings triplet).
When n = 4, R = 6, and when n = 2, R = 1.

What about a binocular, including prisms, objective and wide-angle eyepiece? Typically n = 14, and hence R = (14-1).14/2 = 13 x 7 = 91. This is why a binocular, prior to the invention of mag. fluoride antireflection blooming, lost about 50% of the collected light, and the image had a hazy background and subdued contrast.

The possibilities of ghost images (including those caused by a double reflection between the last surface of the eyelens and the cornea) rises dramatically as the number of elements and associated air-glass surfaces increases. Care is taken to 'design out' ghosts because any double reflection which comes to a focus or near focus will produce a ghost image. These typically appear as either star like points to amorphous patches of light 'floating' in the fov. Ghosts may be stationary, move in the same direction as the image, or in the opposite direction., or even float around the field as the observer's head moves in the case of eye-ghosts.

Whether bloomed, multicoated, or uncoated, the observer should have the aim of keeping the number of eyepiece elements to an absolute minimum. The fewer air-glass surfaces, the less the loss of image contrast, and the better the views of low contrast detail. One has to wonder at amateurs who spend thousands of dollars on expensive apochromatic telescopes only to equip them with 7, 8 or 9 element eyepieces with 8 to 10 air-glass surfaces, resulting in unnecessary contrast losses. Ironically the loss is comparable to an achromatic's secondary spectrum losses and whatever benefit was to be had through the considerable outlay on an objective designed to cure this "problem" is thrown away.

This writer has owned some 350 - 400 different eyepieces over the past half century, and currently has 135. He has seen uncoated  Monocentrics and Hastings (three element cemented triplet with two air-glass surfaces) out perform four element, four air-glass surface eyepieces and seriously out perform 6, 7, 8 and 9 element eyepieces on image contrast and darkness of field. Of course one learns to live with a 25° - 32° afov if one desires the highest image contrast.

Multicoatings can be a help, and they can also be a hindrance when it comes to detecting low contrast detail. For low power oculars, and for oculars used in a binocular, multicoatings do improve contrast, but not at high powers due to the phenomenon of narrow angle scattering that blurrs low contrast detail.

Personally this writer prefers no coatings at all on two air-glass surface eyepieces, and only a single layer MgF2 (blooming) on 3 and 4 element, with four air-glass surfaces. Despite all the advances in eyepiece design and the introduction of peculiar and novel glasses, the 25°- 32° afov Steinheil or Zeiss Hastings triplets, or the Tolles are still the best in providing the highest image contrast. The problem is finding  them on the used equipment market. It is however perfectly feasible to fabricate a Monocentric using a commercial triplet or even simple doublets that give roughly a 20° afov.

Of the four air-glass surface eyepiece types, this writer prefers either the Zeiss Orthoskop, or the Brandon Orthoscopic made by Vernonscope (actually a reversed Plossl derivative), or the modified Clarke E-Z View supplied until recently by Apogee.

It should be noted that contrary to some popular opinion the afov of an eyepiece is not a criterion of performance. The afov is simply the apparent angle subtended by the field stop as seen from the eyepoint, and is a function of the design and field stop internal diameter.

The use of more elements, air-glass surfaces and glass types afford the designer greater degrees of freedom to correct aberrations and distortion at a greater distance off-axis. This is a legitimate goal for deep sky observers especially with an RFT where low powers are employed. However planets and lunar craters subtend small field angles, and because the observer habitually centres the object, there is no need for a wide afov.


As webmaster of this site I would like to add some personal experiences to Rodger W. Gordon's criticisms of multi-element ultra wide angle eyepiece types and their deliterious effects on image contrast.

Whilst I have not collected and used quite as many different eyepieces as Rodger "The Eyepiece King" Gordon, I have amassed a considerable collection, and have disposed of dozens in the past three decades, usually due to their poor or indifferent performance. Of all the eyepiece types I have owned and used for double star and planetary observations, by far the most superior are three humble 1929 vintage Zeiss Orthskop 6mm; 9mm and 12.5mm focal length, and only 40° afov. They are uncoated, have the appearance of dreary microscope eyepieces because they are small and 0".965 push fit. Yet they have the edge on my Monocentrics and Tolles eyepieces, and other Orthoscopics and Plossls, and "Super Plossl" types.

When it comes to wide angle and ultra wide angle eyepieces I am no fan of Naglers, or Meades. I do however use some excellent ex-MoD auctioned ultra-wide Galoc-Bertele hybrids, which have 90° afov yet only five elements in two groups. They work fine at f/10, and are ghost free. The advantage of mil-spec eyepieces lies in the superiority of the polish (the scratch-dig ratio is higher) and the coatings, which causes less light scatter, and higher image contrast. True they need adapting, but the cost of the machining is way less than the extra cost of a commercial equivalent.

On a final note, ultra-wide angle eyepieces exhibit picussion distortion. It is not uncommon to hear amateurs discussing distortion as if it were a problem encountered only in low quality wide angle designs. This is a fallacy. There are two types of distortion that manifest themselves in ultra-wide angle designs, rectilinear and angular magnification. It is impossible to correct both simultaneously at a given field angle since the former is a function of the field angle expressed in radians, and the latter a function of the tangent of the field angle. Once the afov exceeds a radian (roughly 57°) the correction becomes problematic. Angular magnification distortion is commonly corrected at the expense of astigmatism. Astronomical eyepieces are corrected for angular magnification distortion so that magnification does not vary too much with field angle. Military and binocular eyepieces are corrected for rectilinear distortion, preserving as far as practicable the linearity of features at the edge of the field. Nagler and Meade ultra-wide angle eyepieces exhibit almost 30% edge of field rectilinear distortion which I personally find quite objectionable. They also to some extent or another also exhibit spherical aberration of the exit pupil, or the "kidney bean" effect. The later type 2 and type 5 Nagler to a large extent do correct this, but it is not elliminated. Spherical aberration of the exit pupil becomes a problem at wide exit pupils, i.e. low powers.


This page was created by SimpleText2Html 1.0.2 on 28-Feb-2002.