DOF where is it?

[Ororaro]

If it does, then obviously both aperture and focal length remain constant and the DOF markings are also true.

This is correct.

The problem with the dof and coc talk is everyone answers the question they want to answer.

 

[rxmd]

If it does, then obviously both aperture and focal length remain constant and the DOF markings are also true.

Unfortunately this is not true, because the DOF markings are made for a 24x36 sensor. The formulas upon which DOF markings are based mostly assume a certain relationship between the focal length and the size of the recording area; that's why you can't strap a smaller sensor behind the lens and expect the DOF markings to mean the same as with a larger sensor. (The reason why this is assumed in the formulas has a lot to do with acceptable circle of confusion, viewing distances and enlargement factors; there is a lot of scary math behind it, but we don't need to get into all that here.)

This used to be much better known at the time when it was more common to use, say, lenses designed for medium format on a 35mm body. Crop factors are nothing new, but mixing lenses for different formats has been out of fashion for a while so that people have forgotten what they used to know 30 years ago, or so it seems. So a 90/f2.8 lens for medium format stays a 90/f2.8 lens if you use it on 35mm, aperture and focal length are constant, and the position of the film plane with regard to the lens is the same, too, but the DOF markings nevertheless don't coincide. People used to be aware of this when it was more common to reuse medium format lenses on 35mm bodies, but nowadays it looks like this has been forgotten, maybe because few people have ever looked at a medium format lens.

Here's a couple of examples from real-world lenses that I have here on my table. Let's look at lenses in the 90mm range first. The first (MF) is from a Vega-12 90/f2.8 lens which has DOF markings for a film area of 56x56mm, the second (35mm) from a Jupiter-9 85/f2 lens which has markings for 24x36mm. Assume both are focused at 4m. The DOF scale gives me:

f/4: ca 3.5-4.8m (MF) - ca 3.7-4.5m (35mm)
f/11: ca 2.8-6.5m (MF) - ca 3.2-5.5m (35mm)
f/22: ca 2.3-20m (MF) - ca. 2.7-9m (35mm)

As we see, the DOF markings on the lens designed for the larger medium format film are considerably more generous than those on the lens designed for the smaller 35mm format. Both are Russian lenses with their scales based on comparable mathematical assumptions; while the focal lengths are not 100% identical, the 35mm lens is marginally wider which, in theory, should give some more generous DOF, but apparently the difference in format is more important and the DOF markings on the 35mm lens are shallower. If I mount the MF lens on a 35mm body, it will work just fine; but if I go by the DOF markings on the Vega lens, I will find that in fact on the smaller film area, DOF is shallower than the markings suggest. So it's best to err on the side of caution and give a stop or two extra.

(EDIT: Obviously the visual impression of the two lenses on their native formats is vastly different; a 90 on medium format is practically a normal lens, while a 85 lens on 35mm is a short tele. However, when you have the 90 on a 35mm camera, it will work as a short tele, too, and it will not be evident to the user that it was "designed" as a normal lens. So when you use it on a smaller format, you can safely forget all about the fact that it's "originally" a normal lens, because that's not what you use it as.)

Here's another example from the 50mm range, in MF the Zeiss Flektogon 50/f4 mentioned above, in 35mm a Jupiter-3 50/f1.5. Assume both lenses are focused at 2m. The DOF scales give me:

f/5.6: 1.6-2.7m (MF) - 1.8-2.4m (35mm)
f/11: 1.3-4m (MF) - 1.6-2.9m (35mm)
f/22: 1m-beyond infinity (MF) - 1.25m-5.5m (35mm)

Again, both lenses have the same focal length and we are looking at the same aperture, but the DOF markings on the 50mm lens designed for a 56x56mm film area are considerably more generous than those on a 50mm lens designed for 24x36mm. So if I were to use the Flektogon on a 35mm body, it would be a 50mm lens all the same, but the DOF markings on the lens would be too generous; in fact they are consistently off by about 1.5 stops. These 1.5 stops come about because of the difference in film format.

(EDIT: Obviously the visual impression of the two lenses on their native formats is vastly different; a 50 is a normal lens on 35mm, while on medium format it is a strong wideangle. In fact this is why the DOF scale is more generous. However, when using the medium-format 50 on 35mm all notions of wideangle can be forgotten, because it will work as a normal lens like any other; the information that it's "originally" a wideangle then has no practical significance at all. Same thing, incidentally, about using, say, a 35 on an M8; for all practical purposes, including looking through the viewfinder and DOF, the 35 no longer is a wideangle, but now it is a normal lens.)

The first objection will be that these are apples and oranges because there is talk about medium format. The answer is obviously that this is no fundamental difference; either way we are talking about using lenses on a camera with a smaller sensor than its DOF scale assumes. In one scale the DOF scale assumes a "sensor" of 56x56mm and we are comparing it with a "sensor" of 24x36mm; in the other case (the one you're concerned with) the DOF scale on Leica lenses assumes a "sensor" of 24x36mm and you are using it on a sensor of 18x27mm. So fundamentally this is a similar situation. The precise numbers will be different, of course.

The second objection will be that these are Soviet and GDR lenses and that those can't be trusted. However, these happen to have the most precisely readable DOF markings on the lenses that I have lying around. Whoever raises this objection is invited to contribute similar observations from Hasselblad and Leica lenses or whatever; the general trend will be the same. You can actually verify precise numbers with any online DOF calculator that allows you to specify a sensor format.

All in all, the DOF markings on any lens are printed there with some assumptions about the sensor size in mind. These examples should have shown that it does make a difference what size the sensor has. Hence if you use a camera with a smaller sensor size you can't assume that the DOF scale has the same meaning; even if the focal length and aperture are the same, the "sensor" size is different, and hence the scale needs to be adapted. The difference isn't that big between 35mm film and the M8's 1.3 crop sensor, so I'd probably just give it an extra stop for good measure; when shooting at f/8, use the DOF markings for f/5.6 and you should be in safe waters.

The position of the sensor in relation to the prime focal point of the lens (often given with a white line on top of the camera) is different than, say on an M6. Am I right or am I wrong, guys?

Well *here* you're definitely wrong and there's unlikely to be disagreement; the distance from the film plane to the lens in a film M Leica is the same as from the sensor to the lens in an M8 (27.95mm from film/sensor to bayonet in both cases, to be precise). This is because the focusing mechanism of the lens assumes a constant distance here; if the distance was different, the rangefinder mechanism would be off and all the distance information on the lens would be wrong.

Philipp

 

[stephenpicken]

from sp%69cken itself

from sp%69cken itself

I was amused to find the reference to my magnum opus on DOF & perspective here, meanwhile the excel file also contains a universal IR focus guestimator.

If the hyperlinks do not work please use the ones pasted below to avoid the sp%69ck-filter. NOTE however that the files have moved to google as my previous internet provider 'planet' fouled up in a major way and has been abandoned for good.

enjoy
Stephen

http://sp%69cken59.googlepages.com/DOFresolutionandperspective.pdf
http://sp%69cken59.googlepages.com/resolutionandDOF.xls

INDEED THE EMBEDDED LINKS DO NOT WORK - NO ***** ALLOWED