Aspherical Elements: Manufacturing and Benefits

greenspun.com : LUSENET : Leica Photography : One Thread

We all know that Leica molds their asph. elements; question is: how do they get them smooth enough w/o polishing? It would seem that molding would not produce a smooth enough surface when we're dealing with light wave length dimensions. If they are polished, how do they do it? I know, for most of us it's pure speculation. Has anyone who has taken the Leica tour seen, or been explained to, the technique?

What are the advantages of asph. elements other than marketing? It seem every manufacturer makes them. Is it like turbo on a car? There are plenty of cost/benefit issues.

-- chris chen (chrischen@msn.com), March 12, 2002

Answers

Chris:

Aspheric elements can correct the inevitable aberations in light paths due to the large curvatures in spherical lenses needed to get coverage, even fields and all wavelengths to focus at the same point. The Chromatic distortion is a large part of the unsharpness at the edges of images.

I have seen metal and glass castings that are VERY smooth, but would like to hear how it can be done to such a degree without at least a final polish with a very fine abrasive/polishing compound such as lead oxide or carborundum.

Good question

-- RICHARD ILOMAKI (richardjx@hotmail.com), March 12, 2002.


Chris. The ASPH. designation IS a marketing gimmick, but it is true at the same time that aspherical lens surfaces CAN AND DO improve the performance and/or reduce the size and bulk of the lens. By the use of a non-spherical surface, it is simultaneously possible to better correct for spherical aberration and reduce the number of lens elements required to produce the same degree of overall correction. There is a technical discussion of this on Erwin's website, but as a rule of thumb, the use of 1 ASPH surface is the equivalent of two spherical surfaces in terms of the degree of correction. Of course the smaller number of lens elements means fewer potential flare- producing and contrast degrading air glass interfaces.

Also, glass moulding (a technique pioneered by Canon) is NOT the only means used by Leica to produce aspherical surfaces. In the past, they utilized hand grinding and polished (eg., in the 50/1.2 Noctilux and the 35/1.4 Summilux 11873 (1st version) lenses, both of which have two aspherical surfaces)). This procedure was too slow and costly. Now they use glass moulding and also a CNC (computer controlled)lathe to produce aspherical surfaces to the required degree of precision. I believe there is also a third procedure for producing aspherical surfaces used by leica, but I can't recall the details.

I believe the precise protocol depends on the particular lens, and especially on the diameter of the aspherical surface required (larger surfaces such as in the 90/2.0 ASPH-APO) are harder to produce with the required precision than smaller ones.

The reason that I mention that calling a lens aspherical is often a marketing gimmick, is that the use of an aspherical surface(s) is not a guarantee of a better lens (though it does guarantee better PR). This is a function of the overall lens design rather than the use of a particular element or surface. For example, many P & S cameras have lenses with 1 or more aspherical surfaces (to reduce size and bulk); as do all-in-one zooms (eg., 28-200 mm). Yet these are hardly the last word in optical performance.

-- Eliot (erosen@lij.edu), March 12, 2002.


> The reason that I mention that calling a lens aspherical is often a marketing gimmick, is that the use of an aspherical surface(s) is not a guarantee of a better lens <

My thoughts exactly Eliot. At least especially in relation to B&W photography. The aspherics kill some of the character. They maybe better suited to colour but I prefer the pre-ASPH and vintage lenses when it comes to B&W. JMHO,

-- summicron_ (summicron_@hotmail.com), March 12, 2002.


I have seen severely aspherical lenses made out of germanium for thermal imaging devices: they're turned on a lathe from bar stock, using a diamond cutter. Because of the long wavelength involved, these probably don't have to have as perfect a surface as would be needed for visible-light imaging, but the result is certainly smooth enough that no imperfections can be seen by looking at the finished lens, and i suppose they could do better if they had to. It is a pretty expensive process, since you're only making a single element at a time where grinding produces a batch together; but it's reasonably fast, automated, controlled by computer and doesn't require a whole lot of human attention.

For glass molding, I assume they would apply this technique to the mold surface to produce the proper curvature.

By using these severe aspherics in thermal imaging optics, they are able to make lenses with speeds of f/0.87, coverage of 60 degrees... with only 2 elements (granted, we're not dealing with conventional visible light imaging here, but the quality is pretty amazing considering the specs).

One of the companies doing this is ELCAN (Ernst Leitz Canada)

rick :)=

-- Rick Oleson (rick_oleson@yahoo.com), March 12, 2002.


I'm pretty sure Leica mostly uses ceramic molds for their ASPH elements these days - the lenses Eliot mentions WERE hand-ground, but haven't been made for years. They just press the glob of glass into the mold.

Ceramics can be made pretty smooth - consider the original M3 glass pressure plate or the ceramic pressure plates in some Kyocera/Contax cameras (KYOto CERAmics Corp. - get it?)

As to the exact process - I don't know if they cast a curved surface oversize and then 'cookie-cutter' out the part they need - or if the elements are cast to final size. I think all of Leica's present ASPH elements only have one ASPH surface, so the other side gets ground traditionally to a spherical surface, or flat, as needed.

You could try Erwin Puts' site - he has some comments somewhere on ASPH lens production, but I can't find the specific article - maybe one of the ASPH lens reviews.

Finally, Leica may POLISH (NOT grind) the molded ASPH surface - retaining the curvature but smoothing the surface slightly.

-- Andy Piper (apidens@denver.infi.net), March 12, 2002.



It was my understanding that current technology can also be hybrid, ie, glass-molded and final polished on machine.

The process used - molded, ground and polished, or hybrid (or other) may well depend on more factors than immediately seem aparant. Its up to production engineers to do a cost analysis to see which is best suited depending on estimated production quantity, use of existing tooling, procurement of additional tooling, possible re-use of newly procured tooling, perhaps even the labor pool or even access to co-incident local technologies...while keeping within the desired technical specification (quality). And, at times, the production engineer will suggest an alteration of technical specification to better meet avilable production capability, and/or to gain enhanced reduction in cost of production.

-- Charles (cbarcellona@telocity.com), March 12, 2002.


This is clear and simple. 5-10% quality increase for up to 500% of the cost. Is it worth it? It was for me! ......35 Lux Asph and 90 Cron Apo Asph

-- Kristian (leicashot@hotmail.com), March 12, 2002.

I recommend reading Erwin Puts' section about aspherics in his book 'Leica Lens Compendium'. There were/are three methods used by Leica. The first was simply hand ground lenses (original Noctilux), the second was a precision molding technique developed jointly by Leica, Hoya and Schott (many current Leica wide angles). The current method uses computer controlled polishing and grinding machines, i.e. Computer-based numerical control (90 APO Summicron).

-- Bud (budcook@attglobal.net), March 12, 2002.

The other benefit besides spherical correction is distortion correction. This is extremely useful in very asymmetrical designs such as the reverse telephoto(wide angle) used in SLR cameras. The differential magnification across the film plane can be reduced to near zero. Negative elements are given an aspheric shape so that they don't demagnify the image too much in the marginal areas.

-- Steve Rasmussen (srasmuss@flash.net), March 13, 2002.

Moderation questions? read the FAQ