Definition of Infinity?

giordano · February 13, 2012

Advertisement (gone after registration)

When looking at Mtf curves for macro lenses Zeiss gives two or more distances, in the case of the Contax 645 120mm F4 Makro it gives: M1:1, M1:2 and this label:

Unless otherwise indicated, the

performance data refer to large object

distances, for which normal photographic

lenses are primarily used.

What distance is Zeiss using? And is this not their (practical) definition of infinity for testing purposes? I would be willing to submit it is a lot less than fx1000 otherwise they would be shooting at a target impossible to construct and need to test on another planet in order to eliminate atmospheric haze.

120mm Makro: http://applications.zeiss.com/C12578620052CA69/0/F57F3589D022C8D3C125786800233778/$file/apo-makro-planar4_120mm_e.pdf

One of us doesn't understand arithmetic, and I don't think it's me. f is the focal length, e.g. 50mm. 1000 times 50mm is 50m (roughly 165 feet). Or for a 600mm lens, f *1000 is 600m, well under half a mile.

The quote from Zeiss refers to the distance or range of distances at which the lens was designed to give its best performance. It's been amply demonstrated in this thread that no lens is ever used to photograph subjects an infinite distance away, so there's no point computing a lens for general photography that's at its best at "infinity"; instead, a somewhat closer and more typical distance should be used. Maybe this is where the "f * 1000" comes from. Michael, can you clarify?

February 13, 2012

Hi giordano,

Take a look here Definition of Infinity?. I'm sure you'll find what you were looking for!

pgk · February 13, 2012

But as far as the marking on the distance scale is concerned, it is supposed to be exact (within tolerances) and to conform to the definition.

Michael. My point was that the last marked distance on the scale was once apparently/supposedly half the distance at which the lens is "for all practical intents and purposes" focused at infinity. Not that infinity on the lens was incorrectly placed on the lens - there is a difference - but in the context of this discussion this might have been at times a useful distance to bear in mind. As it happens a quick look at my recent M lenses shows that this is no longer the case with final distances probably marked more for cosmetic reasoning than for any useful or logical function.

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

darylgo · February 13, 2012

instead, a somewhat closer and more typical distance should be used. Maybe this is where the "f * 1000" comes from. Michael, can you clarify?

Exactly, that was the point I was trying to make, theory is theory but lens focus is real world and the f x 1000 may have greater significance than the larger values referenced in this thread.

giordano · February 13, 2012

Exactly, that was the point I was trying to make, theory is theory but lens focus is real world and the f x 1000 may have greater significance than the larger values referenced in this thread.

Only if it's clear that 50 metres or 1000 * f is not equivalent to "infinity" when it comes to collimating a lens or calibrating a focusing mount

mjh · February 13, 2012

What distance is Zeiss using? And is this not their (practical) definition of infinity for testing purposes?

In the good old times before the advent of fast computers, lenses were optimized for infinity focus. Not because anyone believed that infinity was what lenses were most often focused at, but just to simplify the already quite complex calculations and to reduce the number of those calculations. These days this isn’t necessary anymore; one can optimize lenses for more typical subject distances which for many lenses might be 2 to 5 meters, say, rather than infinity. But of course the typical subject distance depends on the lens and will differ between a portrait lens and an extreme telephoto lens used by wildlife photographers. In the case of the lenses for the S2, Leica claims they optimize performance nor for one but for all distances, although there will always be a distance where performance reaches a maximum and one would strive to achieve this peak performance at a subject distance typical for that lens.

Just as an aside: Reading this thread I got the expression that infinity is seen as something that can only be approximated. To quote from the Hitchhiker’s Guide to the Galaxy: “INFINITE: Bigger than the biggest thing ever and then some. Much bigger than that in fact, really amazingly immense, a totally stunning size, real ‘wow, that’s big,’ time. Infinity is just so big that by comparison, bigness itself looks real titchy. Gigantic multiplied by colossal multiplied by staggeringly huge is the sort of concept we’re trying to get across here.” But as far as focusing lenses is concerned, there is no need to approximate. When the distance between the lens (or actually its rear principal plane) and the sensor (or film) equals the focal length, the lens is focused at infinity, i.e. parallel light rays hitting the lens will converge in the sensor/film plane. That’s what the infinity mark indicates: at that mark, the image distance is equal to the focal length. If a lens manufacturer wants to test for this, they wouldn’t bother with something at 1000 x f, a belfry in the next village, or a bright star – they will use a collimated laser beam. Parallel light rays from a subject at infinity are created on a lab table. Only the photographer who is typically lacking such equipment would ever need to rely on approximations. The factor by which to multiply the focal length depends on our resolution requirements, i.e. which CoC is capturing our notion of acceptable sharpness. 1000 x f was a useful rule of thumb for moderately fast lenses (not for a Noctilux) when a 30 µm CoC was deemed adequate, but not anymore (as we have seen).

Edited February 13, 2012 by mjh

SJP · February 14, 2012

<snip>To quote from the Hitchhiker’s Guide to the Galaxy: “INFINITE: Bigger than the biggest thing ever and then some. Much bigger than that in fact, really amazingly immense, a totally stunning size, real ‘wow, that’s big,’ time. Infinity is just so big that by comparison, bigness itself looks real titchy. Gigantic multiplied by colossal multiplied by staggeringly huge is the sort of concept we’re trying to get across here.”<snip>

February 15, 2012

Advertisement (gone after registration)

In the good old times before the advent of fast computers, lenses were optimized for infinity focus. Not because anyone believed that infinity was what lenses were most often focused at, but just to simplify the already quite complex calculations and to reduce the number of those calculations. These days this isn’t necessary anymore; one can optimize lenses for more typical subject distances which for many lenses might be 2 to 5 meters, say, rather than infinity...

...do you have a reference for this? What was optimized for infinity?

nick.edwards1 · February 15, 2012

According to Gunter Osterloh's book, Leica use a value for infinity in their lens design calculations.

I think that this figure is probably what would be termed 'Leica's definition of infinity'.

This value is quoted as 1,701,411,834,605 x 1035 meters.

Ref. Leica M Advanced Photo School 2011 edition P137.

It is stated that this value is used as that number is the largest that an be utilised in the computer - but maybe times have changed.

Best regards,

Nick.

Lindolfi · February 15, 2012

This also surprises me, that it is stated that camera lenses would be optimised for the infinite object plane. In "Optical Aberration Coefficients" by H.A. Buchdahl (1954), formula's are offered at other object planes for the monochromatic and chromatic aberrations. (for a glimpse of that: click)

February 15, 2012

I think it is typical, both in the "good old times" and more recently, since some things can only be perfect at a single conjugate distance, infinity is chosen.

Rick in CO · February 15, 2012

Or more precisely, what is Leica's definition of infinity as applied to M9? On my M9, there is a need to focus slightly less than infinity at distances I would consider infinity when using a DSLR. Would 200 meters be considered infinity?

? Wouldn't this just be the difference in depth of field markings between two different lens manufacturers? Maybe you're noticing differences between lens originally manufactured for film vs flat digital sensors? I have never considered there to be a single discrete infinity focus - focus varies for any given distance. Isn't the infinity mark on a lens for depth of field considerations, which does not necessarily yield sharp focus (for a pretty decent review see the recent Leica Mag article). I have wrestled with this issue setting up new lenses for cam-coupled focusing on my Linhof Technikas. You basically determine several optimal near distance focus points and let infinity fall where it may - it's not an absolutely discrete measurement in practice; in theory, maybe.

jaapv · February 15, 2012

Not to mention the temperature of the lens mount..

pico · February 15, 2012

I have wrestled with this issue setting up new lenses for cam-coupled focusing on my Linhof Technikas. You basically determine several optimal near distance focus points and let infinity fall where it may - it's not an absolutely discrete measurement in practice; in theory, maybe.

Why are you calibrating the rangefinder on the Super Technika? Is it broken?

I know that the Graflex rangefinder requires four, possibly three distance settings to calibrate, and the Linhof finder appears to be similar but for the Linhof Super Technika the cams do the work (at least for 4x5 and larger) and are carefully machined to a curve that covers all distances - and of course there is the helpful adjustable infinity stop. The Super Technika's cams will be serial numbered to match the lens.

Rick in CO · February 16, 2012

Why are you calibrating the rangefinder on the Super Technika? Is it broken?

I know that the Graflex rangefinder requires four, possibly three distance settings to calibrate, and the Linhof finder appears to be similar but for the Linhof Super Technika the cams do the work (at least for 4x5 and larger) and are carefully machined to a curve that covers all distances - and of course there is the helpful adjustable infinity stop. The Super Technika's cams will be serial numbered to match the lens.

Yes, that is all true. The 4x5 Linhof Technika V and the 6x9 Technika 70 and 6x9 Super Technika V are manufactured to such tolerances that one can switch cam/lens combinations between these (and later) cameras provided one has the matching serial numbered lens. This cannot be done with the earlier cameras, which have both the serial number from the camera and the lens inscribed on the cam. Linhof even provided instructions in a pamphlet (included was a focusing target!) that allowed the user to properly set this up on their camera without having to send the camera in for service. It is not well known that even the Graphic cameras required the cams to be precisely ground by the service department to focus accurately, and that even the focusing scales (both Linhof and Graflex) required this custom service! My comment on letting the infinity focus fall "where it may" was in reference to the placement of the infinity stop - its position was determined in relation to finding accurate focus for the nearer distances. For a more in depth discussion see my web page fotoduo.com and click on Ricks Classics, Linhof Technika 70.