Chromatic Aberration in Black and White

[Ed Richards]

I am curious about the effects of CA in black and white. In particular, its role in the look of older lenses. I will confess that my objective is to pick some old lenses for 35mm, but no one in that world thinks of aberrations in a positive light.:-) I want some lenses to take the edge off my digital work. For example, this is a shot with an old 55mm f1.2 Nikon lens, at 1.2. Spherical aberration adds to the look of this image.

http://www.epr-art.com/galleries/b4a-la-ca-churches/photos/_DSC9986.jpg

[Bruce Watson]

It's hard to make a lens for photography that doesn't exhibit some level of chromatic aberration. The best lenses are apochromatic, meaning they can focus three different wave lengths at the same point. Perhaps most lens are "just" achromatic, meaning they can focus two wave lengths at the same point (typically red and blue IIRC).

Either way, there are some wave lengths that will be focused better then others. The effect is that some wave lengths will be sharper than others in the final print. Can you see it? I imagine if you were looking for it and were meticulous you'd find it eventually. I suspect that much depends on the depth of focus (as opposed to the depth of field). Since the depth of focus for small formats is smaller (if I'm wrong about that someone please correct me) I would think this would be of greater concern there than with LF. I think this because people are getting great images with older, less well corrected lenses on LF all the time.

In any case, what you'd see in a B&W print would be an inconsistency of focus that you couldn't easily correlate (because the color information is gone). It would be mildly disconcerting. Probably not in a pleasing way.

Would this work for you? IDK. I suspect you are looking for a more broadly applied effect like spherical aberration which comes across as a more general softening. Again, IDK.

[walter23]

I had a neat old 58mm screw-mount Takumar lens with a heliar-influenced design that made for some pretty cool images because of the aberration. Bokeh was a bit weird though ("chunky"). I sold it because I couldn't get a proper adapter and had to kind of cobble it onto my digital camera (it was an older Asahiflex screw-mount, not M42, and adapters aren't readily available for that).

[Dan Fromm]

Bruce, the Boyer Opale -- read about it, the company, and most of its other lenses at http://www.galerie-photo.com/boyer-lens-optic.html -- is a soft focus lens that got the effect by chromatic aberration. It is an anachromatized tessar type, was discontinued when color film came in. Berthiot's f/4 Color, their Series I d, is an other soft focus anachromat. Entirely different formula from the Opale. You can read about it here: http://www.collection-appareils.fr/accesnotices/html/lire_repertoire?repert=som_berthiot&marque=Som%20berthiot&modele=Catalogue&PHPSESSID=333965b3c0b5e4652c9d3802b532215d

Opales turn up from time to time on one eBay or another but I've never seen a Serie I d Color offered. I have the impression that most of the lenses Berthiot made for large format are now very uncommon, have no idea how many of the various types were made.

[Ed Richards]

Bruce,

Since I think (and am sure I will be corrected) that the CA is relative to wavelength, it would be the same in absolute terms for all lenses. Thus a much bigger pain in 35mm, and a monster in smaller sensors unless almost perfectly corrected in glass, or digitally corrected in camera. It should be mostly unnoticable in LF except in small prints or extreme cases - fine wire window screen in a picture might do it.

[BetterSense]

It's been my experience, or at least my superstition, that colored filters increase sharpness in B&W due to removal of CA.

[ondebanks]

It's been my experience, or at least my superstition, that colored filters increase sharpness in B&W due to removal of CA.

Not a superstition - this is absolutely true. Companies such as Lumicon supplied astrophotographers with filters like "Minus Violet" and "Hydrogen alpha", with the promise of "pinpoint star images" by cutting off selected regions of the spectrum. They work like a charm.

When hypersensitized Kodak Tech Pan reigned supreme, even modest 35mm camera lenses and a deep red (H alpha) filter delivered exquisitely sharp and detailed results on nebulae and starfields. The trick was to test for the best focus point with the filter in place. Usually it was about 2/3 of the way from the infinity mark to the infrared mark on the lens barrel. The filter also eliminated the nasty sky glow from light pollution wavelengths and natural atmospheric airglow.

Ray

[Brian Ellis]

It's been my experience, or at least my superstition, that colored filters increase sharpness in B&W due to removal of CA.

Never heard that one before. Which doesn't mean it isn't true, just that I've never heard of it and am dubious (or at least wonder whether any benefit from the removal of CA isn't more than offset by deterioration in other aspects of the image such as increased flare).

[Emmanuel BIGLER]

It's been my experience, or at least my superstition, that colored filters increase sharpness in B&W due to removal of CA.

Well, it depends on the degree of 'sharpness" that you are expecting from your lens.
If you criterion is "400 cycles/mm, not less", then you have to seriously think about using filters for B&W work ;)
When I was a student in the 80's, our lab acquired for nothing (except transportation and re-installation charges) a complete photomask copying bench system that was just about to be discarded by another French institution.
The bench was fitted with a superb CERCO process lens, operating in visible light, this was a f/4 diffraction-limited top-class lens ; it had to be permanently fitted with a built-in green filter.
I have no idea about how this lens would perform for general photography without its dedicated filter...

Jörg Krusche has lent to me a series of technical documentations for apo-ronar lenses ; the MTF charts in those old data sheets are not plotted like in modern lens data sheet, but various MTF curves are shown for various kinds of light sources. So at the time, Rodenstock did care for the influence on image performance of the color of the light source.
see for example, apo ronar 240, on page 2 (sorry, this is the French version), MTF charts differ for CYAN, MAGENTA or YELLOW filtering (http://www.cijoint.fr/cjlink.php?file=cj201101/cijMDP9AHE.pdf)
However I assume that hose curves are computed; and not measured; hence, I have no idea on the real, visible influence on image quality in real "white-light" life...


for those interested in the history of photomask fabrication, the system with this CERCO lens was already obsolete in 1980, at the time, photomasks for the so-called "0.8 micron" technology were no longer made by making a photo of a big drawing created by a plotter on a big sheet of paper, but by a direct optical pattern generator (rotating slit machines) illuminated by a mercury arc lamp. Nowadays fabrication of photomasks relies on direct laser or e-beam writing.

[Jack Dahlgren]

Never heard that one before. Which doesn't mean it isn't true, just that I've never heard of it and am dubious (or at least wonder whether any benefit from the removal of CA isn't more than offset by deterioration in other aspects of the image such as increased flare).

I've heard of it. Must have been in a book I read long ago, but I'm sure that lens designers have always tried to correct for CA (unless they were designing it in).

The more interesting thing is that the original poster is talking about doing this for a camera with a color sensor. In that case CA can be completely eliminated by just selecting one channel, or it can be present in full form and it should take just a few minutes of experimentation to evaluate. Further, much of the software out there allows correction of CA via lens profiles - meant to eliminate CA, but perhaps applying the wrong profile to a lens could create CA. CA being angular varies geometrically from the center of the lens so you would look for it more near the edges. I think that it would not be that difficult to create an algorithm which creates CA, running through a number of wavelengths and shifting them proportionally.

But why all the fuss when a smear of vaseline on a filter will achieve more dramatic results?

[Ed Richards]

> But why all the fuss when a smear of vaseline on a filter will achieve more dramatic results?

Because subtle is what makes great images.:-) I do not want to ignite a film/digital war, but I am working on sorting out how much of what I like about older images on 35mm is due to less than perfect lenses versus film rather than digital. I posted here because of the irony of how many LF folks spend their time using old lenses to make their images look like nostalgic shots done in the 19th century. Who better to opine on what crappy lens design and construction can do to make a beautiful image that does not depend on being sharp or high contrast. The lens used on the picture is 40+ year old design for one of the first ultra high speed lenses - lots of aberrations, but also sharp.

[cowanw]

Ray filters (yellow) were common in the early 20 th century for this reason.

In THE SOFT-FOCUS LENS AND ANGLO-AMERICAN
PICTORIALISM By
William Russell Young there is an excellant discussion of this; for example.

"The single element lenses were premised on both Wollaston and Watzek’s
researches and nearly always took a meniscus form, plano-convex and bi-convex having
crucial shortcomings. Cheap to manufacture, they were difficult to use, “such a lens
possessing all possible errors, and giving, as a result of its optical defects, a very soft
quality of definition.”12 The chemical and visual focus did not coincide thus an
adjustment had to be made after the visual focus. Because a simple meniscus has
significant chromatic aberration, the “soft quality” was exaggerated after panchromatic
emulsions came into common use, unless, however, “a ray filter {yellow equivalent to a
modern K-2} and color sensitive plates are used, the lens is rendered for all practical
purposes completely achromatic and no correction need be made after focusing"

"4. Dallmeyer-Bergheim (1896) which used two elements, neither corrected
separately nor as a pair corrected, with variable separation which both changed the focal length and the amount of aberrations. Unlike later soft focus lenses, it was largely
dependent on chromatic aberration (rather than spherical aberration) for its effect"

[Jack Dahlgren]

> But why all the fuss when a smear of vaseline on a filter will achieve more dramatic results?

Because subtle is what makes great images.:-) I do not want to ignite a film/digital war, but I am working on sorting out how much of what I like about older images on 35mm is due to less than perfect lenses versus film rather than digital. I posted here because of the irony of how many LF folks spend their time using old lenses to make their images look like nostalgic shots done in the 19th century. Who better to opine on what crappy lens design and construction can do to make a beautiful image that does not depend on being sharp or high contrast. The lens used on the picture is 40+ year old design for one of the first ultra high speed lenses - lots of aberrations, but also sharp.

I have one a 55mm f/1.2 I agree it is a great lens. Wide open or at f/8.
I just don't think that there is enough CA in any modern (<40 year old) Nikkor's to give a hint of that old look through CA alone.

Your comment about "irony" of LF users making pictorial shots seems to come from an assumption that large format exists only to achieve high levels of detail. I think that may be one of the goals that LF users have, but ultimately, LF continues to exist because of the flexibility. Flexibility to tilt, shift, rise, swing, focus very close or far, flexibility to use modern and ancient lenses, different size backs, flexibility to create extremely sharp or extremely soft images, ability to use wet plate, film, paper, whatever... So it would only be ironic if everyone using LF used it in the same way. Fortunately, that is not the case. I think that is what makes it interesting.

[Ed Richards]

Irony because 90% of the posts about lenses are whether this incredibly sharp lens is a little sharper than some other incredibly sharp lens.:-)

Some of the fast old Nikkor have enough spherical and chromatic aberration and wide open to give a nice look. I have an old 35mm f1.4 on the way. The 135 f2 AI might be another good candidate, and your lens is also nice wide open.