lenses designed for digital or black and white capture

[Deniz Zagra]

This might be a trivial topic for a lot of people, but I'm very interested in learning more about lenses, and I couldn't find any information about these topics.

Why is every lens company manufacturing new lenses that they claim to be specifically designed for digital cameras? Apart from the ability to make the lens closer to the sensor and to include anti-aliasing coating, what other benefits do these new lenses offer? Is it a substantial difference that these companies can justify (financially) making entirely new lens designs? A few years ago, Panavision introduced their Primo V and T-series anamorphic lenses. Apparently, the T-series lenses are being used a lot, even on film cameras; however, I haven't seen many productions utilizing the Primo V lenses. I've heard from some stills shooters that lenses designed for film can introduce extra chromatic aberration into the digital frame. Am I missing something here? If you were shooting a film on digital, would you go with the original Primos or the new Primo Vs? Would there be unwanted side effects if these lenses were used for celluloid capture?

One last question: I was watching the bonus stuff on the Criterion blu ray of Barry Lyndon. There was a section where Joe Dunton talked about the lenses Kubrick owned. He mentioned some very small lenses that Kubrick used on his early black and white pictures and that these lenses were specifically designed for black and white cinematography. Any interesting facts about those types of lenses? What extra requirements/adjustments are needed for lenses designed for b&w? Are there any newly designed lenses specifically for b&w photography?

The lens in question is discussed at this time code:  6:20

 

[Dom Jaeger]

The main thing a lens specifically designed for digital sensors will be is telecentric (or more correctly  near-telecentric), which just means that the light rays coming from the rear element are fairly parallel. Film emulsion doesn't care what angle a light ray hits it, whereas a sensor photosite does. Older lenses from the film era with tiny rear elements and shallow exit pupils (often wider angles) can induce colour fringing or vignetting when used on digital sensors. 

You can quite happily use modern lenses with film, the only issues are when the rear element protrudes too close to the film plane and can hit a reflex mirror, but despite a brief dalliance with this sort of design in lenses like Angenieux's DP zooms or Panavision's Primo Vs, most manufacturers have gone back to allowing more clearance for 35mm format lenses. Large format lenses however are now often designed with shorter flange depths, since the cameras are almost all digital, and they can utilise shorter flange depth mounts like Arri's LPL or Panavision's SP70. 

The irony of companies releasing "digital" lenses is that many people prefer to use older lenses that were designed for film anyway. 

As far as lenses designed for B&W goes, I don't see what specific attributes you might want in a B&W lens that you wouldn't also want for colour. Many lenses from before the 50s were probably designed with B&W film in mind, which is possibly what Joe Dunton meant, but they could certainly still be used with colour film. As lens design progressed, you had aspheric or APO lens designs which better corrected for a larger range of wavelengths, but they will also make a B&W image sharper. 

 

[Joerg Polzfusz]

Hi!

Dom perfectly summarized the main optical aspect. But there’s another aspect: newer lenses transmit their settings (f-stop/t-stop, zoom-rate, focus distance, …) to the camera. There is special software that can then eliminate „flaws“ of the lens based on this data (e.g. DxO Optics for still images). And there’s other software that can use this data to better combine green screen scenes or cgi-scenes.

As for the „b&w only lenses“: theoretically, there’s no such thing. But in reality, there are some strange lenses with a slightly orange coating that render them useless for color films. And there are also some lenses that block certain wavelengths (which doesn’t make any difference for orthochromatic films). But this will only result in making them useless for color films. They will not result in „better“ b&w images. 

[David Mullen ASC]

Some older lenses have chromatic aberrations that were fine for b&w work but not for color.

I heard that the mythical Garutso lens had this issue which is why it disappeared. I’m surprised one hasn’t shown up in some collection though.

[Phil Rhodes]

Strictly speaking I guess you'd sort of expect that. The very earliest designs might have been designed for the peak sensitivity of orthochromatic film, which would have been bluish. Reds might look smeary. But I'm speculating wildly - did they have much capability to tweak things that finely back then?

[Simon Wyss]

Eastman-Kodak introduced their first panchromatic film in 1925. At that time the asymmetric Taylor-Hobson Cooke Opic lens was already in existence. Its six-elements Double-Gaussian design became the father of all modern systems such as the soon following Baltar by Bausch & Lomb. With six elements the optician can flatten the field, take care of even illumination, correct for almost every aberration, and make sure the entire spectrum of light colours hits the film in one plane. Kinoptik managed to design six-glass apochromats with new sorts of glass.

Full spectrum colour correction is not possible with three elements but from four on you can have very good results. An example of that type is the dialytic TTH Cooke Ivotal of f/1.4. The limitations were with the glass, if you don’t mind accepting that Joseph Petzval had only two sorts of glass to work with in 1840. Kodak employed a 1950 Petzval design with the Escort 8 camera of 1964. The Bausch & Lomb Animar 26 mm, f/1.9 is a also a modern Petzval lens.

In still photography, especially for reprographic and close-up work, symmetric designs were common. They’re relatively cheap but perform well. Again Eastman-Kodak, the f/1.9 anastigmat for Kodacolor is a symmetric dialytic four-glass lens that can still give stunning pictures. Wollensak and Elgeet have adopted it.

When Kodachrome was underway in 1934-35 the optical industry was already aware of new possibilities with the inverted telephoto or retrofocal arrangement. That had come into being for the TC three-strip beam splitting camera. The wide-angle explosion came when those patents elapsed after WW II.

[Deniz Zagra]

13 hours ago, Dom Jaeger said:

The main thing a lens specifically designed for digital sensors will be is telecentric (or more correctly  near-telecentric), which just means that the light rays coming from the rear element are fairly parallel. Film emulsion doesn't care what angle a light ray hits it, whereas a sensor photosite does.

Is it because sensors are made up of small pixels and film is just a continuous layer of silver salts? Scenes that were shot with a super wide angle lens in A Clockwork Orange (1971) had a lot of fringing on high contrast sections of the frame. Are those caused by something else?

[Dom Jaeger]

My understanding is that it’s because sensors use micro lenses above each photosite, which gather the light but can cause problems when the rays are too angled. I’m sure googling sensor design will tell you more.

Colour fringing, or chromatic aberration, is a common aberration in many lenses,  especially older wide angles. It’s a seperate thing to the colour fringing caused by photosite crosstalk on a sensor.