A question about large format optics

[walter23]

What I would like to know is, is that optical affect a physical function of the large format lens, unique to that format, or merely the result of large format film's resolution power; if a medium format digital back had that power of resolution, would I get that same gradation of focus or not? That is why I am interested in understanding what the difference might be, if any, between the different formats.

I think there is a physical difference that has nothing to do with resolution, and more to do with depth of field and the optics of what is considered to be "in focus". This is a mix of intuition (which in physics can be pretty wrong, I know) and empirical observations, and facts about depth of field. A 6x6 negative with an 80mm lens at f/5.6 will have more depth of field than a 4x5 and a 180mm lens (or whatever the 80mm equiv would be) at f/5.6. That's why if you use a 35mm SLR you can achieve much more beautiful focus fall off effects than with a tiny compact digital point and shoot which has virtually infinite DOF at any moderate focal length wherein a normal lens is something like 10mm. There's a counteracting effect of enlargement (ie, with a 2/3" sensor or whatever on a tiny digital you're enlarging many fold and therefore focus differences will become more apparent) but the lens focal length differences (for a given point of view) are stronger and you see lower DOF at larger formats.

I'm not sure how this relates to the gradual nature of focus fall off, but empirically it seems like larger formats can give more gradual transitions from in focus to out of focus. If you could quantify focus (some function of the circle of confusion and spread of light or whatever) it would be cool to calculate "focus curves" showing how focus falls off in different formats as a function of distance (and lens, aperture, and focusing distance). Anybody want a programming project? I might try this out for the sake of curiosity sometime if I can find the right formulae.

[C. D. Keth]

I think you would find that physically, the focus is doing the exact same thing for a given enlargement ratio, f/stop and subject distance. The difference in larger formats, I think, is that the ability to record smaller increments of out-of-focus is present. With smaller formats (I see this daily working with video, which has comparably tiny sensors and terrible resolution to even consumer digicams) the focus all of a sudden seems to snap soft rather than transition. When I shoot 16mm it's a bit better, when I shoot 35mm it's better still.

I think it's a resolution issue, but I have not performed controlled experiments to prove it.

Edit: Something else just occured to me. LF lenses are often not of the same optical quality that top-shelf lenses for 35mm, or for cinema use are. The enlargment ratios for LF simply don't need it and they would become prohibitively expensive like cinema lenses (20-30k per prime is the going rate for the newest set of Arri/Zeiss lenses). This might make the transition a bit smoother since the in-focus areas aren't technically as in-focus as a brand new set of zeiss master primes would make it, for example. The trick is that it doesn't matter since LF will be enlarged a few times and the 35mm MP negs will be enlarged to the size of a large wall. Just a thought.

[walter23]

I think you would find that physically, the focus is doing the exact same thing for a given enlargement ratio, f/stop and subject distance.

I don't think so.

4x5 film (162mm diagonal), 180mm lens, f/5.6, 10 foot focusing distance:
Subject distance 10 ft

Depth of field
Near limit 9.52 ft
Far limit 10.5 ft
Total 1 ft

In front of subject 0.48 ft (47%)
Behind subject 0.53 ft (53%)

Hyperfocal distance 188.5 ft
Circle of confusion 0.1 mm

6x7 film (92.1mm diagonal), 100mm lens (using diagonal for equivalences), f/5.6, 10 foot focus:
Subject distance 10 ft

Depth of field
Near limit 9.3 ft
Far limit 10.8 ft
Total 1.51 ft

In front of subject 0.7 ft (46%)
Behind subject 0.81 ft (54%)

Hyperfocal distance 129.2 ft
Circle of confusion 0.045 mm

35mm (43mm diagonal), 47mm lens, f/5.6, 10 foot focus:
Subject distance 10 ft

Depth of field
Near limit 8.13 ft
Far limit 13 ft
Total 4.87 ft

In front of subject 1.87 ft (38%)
Behind subject 3 ft (62%)

Hyperfocal distance 42.9 ft
Circle of confusion 0.03 mm


I don't know exactly what goes into those DOF formulas, but I think it's based on a standard enlargement size.

This is from http://www.dofmaster.com/dofjs.html (the standard online DOF calculator tool). You can see clearly that DOF increases as format size decreases and it's not just about resolution; the decreasing focal length (for an equivalent field of view) dominates the calculation, so in 4x5 you'd have 1 foot DOF, 6x7 about 1.5 feet, and 35mm about 5 feet.

[Gordon Moat]

There are a few other ways to look at this. It is now possible, through a few companies, to place a D-SLR, or medium format digital back, onto a large format (4x5) camera. Using either a sliding adapter, or movements of the back standard on the view camera, you can use the exact same large format lens for 4x5 film, stitching a few digital back shots, or stitching many D-SLR shots. The resulting images should be very close in terms of DoF or focus fall off.

Consider that, for example, a 135mm lens on a 4x5 would still be a 135mm lens on medium format, and still an 135mm lens on 35mm (or full frame D-SLR). Without changing camera to subject distance, the DoF would actually be nearly the same on each. The main difference would be that without stitching the smaller format shots together, your FoV (field of view) would be more cropped on smaller formats. You could shoot 4x5, and cut out a 24mm by 36mm or 56mm by 72mm part of the film, though in practice hardly anyone does this.

To get a headshot (top of head to shirt collar) on 4x5 with a 135mm lens, your focus distance is 0.5m away. At f5.6 the DoF is 0.011m. A 6x7 medium format shot with the same lens (and f5.6) would only show the nose to top of head, and factoring different CoC would indicate 0.006m DoF. A 35mm (or full frame D-SLR) shot at 0.5m distance (and f5.6) would be another CoC calculation indicating 0.003m DoF, and only show a little more than one eye of the subject. However, if you wanted to get the same headshot on each format, with the 6x7 you would probably use an 80mm lens at the same distance, and perhaps a 50mm or slightly wider FoV) on 35mm (or full frame D-SLR). With the 50mm, even with different CoC, the DoF would be about 0.026m at f5.6 and 0.5m, while the 80mm on 6x7 would be about 0.020m DoF.

Playing around some more with the calculations and variations in pCAM software on my Palm, an 80mm lens on 6x7 at 0.5m distance for a headshot could be done at f2.8. In such a setting, the DoF drops to 0.010m. Using a 50mm on a 35mm camera, you could have an f1.4 aperture setting, giving a DoF of 0.006m. So that effect of focus falloff could be achieved in a few different ways. Notice I did not pick exact comparible focal lengths, but I did choose more commonly available lens choices (a 40mm f1.4 on 35mm would be another possibility)

Ciao!

Gordon Moat
A G Studio

[Mark Sawyer]

I suspect that as far as the "feel" of the lens goes, large format photographers have a much, much wider choice, and it reflects in a *lot* of the work seen here. How many vintage/classic optics or soft-focus options does the medium-format user really have to choose from?

I think the soft-focus effect is undergoing a renaissance similar to the upsurge in interest in large format, and while there are only a few modern soft focus lenses available, there are quite a few older ones still working wonderfully well. It's worth remembering that once upon a time, such lenses were an very important part of a lens-makers catalog. In 1919, Wollensak had five separate lines of soft/diffused focus lenses (Vitax, Verito, Voltas, Vesta, and Versar), but only one "sharp" general purpose line of lenses, the Velostigmat. And even at that, many Velostigmats were produced with a diffused-focus adjustment built in...

[Paul Fitzgerald]

"Let me be more precise about my example. I love the way I get focus fall off on a large format neg when I do a portrait up close: I hold the tip of the nose to the eye, and by the ear there is this beautiful transition as the focus falls off. What I would like to know is, is that optical affect a physical function of the large format lens, unique to that format, or merely the result of large format film's resolution power; if a medium format digital back had that power of resolution, would I get that same gradation of focus or not? That is why I am interested in understanding what the difference might be, if any, between the different formats."

If you took a standard portrait lens and set-up a portrait shot and changed nothing else but the film back size (5x7, 8x10, 11x14) they would each 'look' different. I guess it's a subconscious perception or a trick of the light, but it's real. It could just be the comparative 'real-estate' of the films allowing more of the 'character' of the lens to be seen at the same time.

just a thought.

[Leonard Evens]

I should specify, I am not interested in what is sharper as an abstract argument. I am interested to know whether there is a difference in optical affect between lenses of different formats.

Let me be more precise about my example. I love the way I get focus fall off on a large format neg when I do a portrait up close: I hold the tip of the nose to the eye, and by the ear there is this beautiful transition as the focus falls off. What I would like to know is, is that optical affect a physical function of the large format lens, unique to that format, or merely the result of large format film's resolution power; if a medium format digital back had that power of resolution, would I get that same gradation of focus or not? That is why I am interested in understanding what the difference might be, if any, between the different formats.

Partly this would simply be a matter of depth of field. You can certainly obtain the same DOF for a given scene and given angle of view with medium format that you do with large format, but you would have to use a larger aperture. If you concentrate on the final print, I believe the circles of confusion would behave the same way for comparable setups, where the focal lengths and apertures are appropriately adjusted.

But what you describe seems to relate to the way the image goes out of focus as you pass through the limits of DOF. I don't know for sure, but I believe this may relate to bokeh. Try en.wikipedia.org/wiki/Bokeh or do a google search. Bokeh describes how the shape of the diaphragm, lens aberrations, and other things may affect the appearence of out of focus regions. It is generally a function of the specific lens. Off hand, I see no theoretical reason why it should depend on the format, but since there are so many factors to consider and since much about bokeh is subjective, it is possible that format size is one factor affecting it. It is also possible there is some systematic bias in the design of large format lenses which might affect it, but in principle, at lteast, you ought to be able to obtain essentially the same effect with two lenses, one for large and the other for medium format.

[Struan Gray]

If you double the size of the negative you have to double the focal length to get the same field of view and stop down two stops to get the same depth of field. In principle this can be done.

However, going up in format you either need to get hold of a faster film (which will, in general, look different) or accept longer shutter speeds (which will, in general, look different).

Going down in format you need to find a faster lens, which may not be possible, or which may be dominated by aberrations at the aperture you need to use (which will, in general, look different). Failing that, you may find diffraction blur is an issue (which will, in general, look different).

LF is a regime where it is easy to have a short depth of field on large objects while the out-of-focus parts look pleasingly symmetrical and even. Tonality is super-smooth.

Small formats like 35 mm and below are a regime where it is easy to get everything in focus. Tonality is gritty and grainy.

[Leon Jester]

If you double the size of the negative you have to double the focal length to get the same field of view and stop down two stops to get the same depth of field. In principle this can be done.

However, going up in format you either need to get hold of a faster film (which will, in general, look different) or accept longer shutter speeds (which will, in general, look different).

Going down in format you need to find a faster lens, which may not be possible, or which may be dominated by aberrations at the aperture you need to use (which will, in general, look different). Failing that, you may find diffraction blur is an issue (which will, in general, look different).

LF is a regime where it is easy to have a short depth of field on large objects while the out-of-focus parts look pleasingly symmetrical and even. Tonality is super-smooth.

Small formats like 35 mm and below are a regime where it is easy to get everything in focus. Tonality is gritty and grainy.

If you use the long + short formula for a portrait lense you won't get the same depth of field, regardless.

To wit: 35mm with this formula would use a 60mm lense, 6x6 would use a 120mm lense, 4x5 would use a 9 inch (228mm) lense, 5x7 would use a 12 inch (304.8mm) lense.

As far as film resolution goes, the faster the film the grainer it is, simply because of the physical makeup of the emulsion. A portrait shot on, say, Panatomic X is going to have a vastly different appearance and tonal value than one shot on Super XX or Tri-X.

Use whatever works for you, that's going to be part of your photographic style.

[jb7]

In my opinion, there is no mystery-
the combination of image size,
coupled with choice of aperture,
will have the same optical effect-

For an original image, (disregarding magnification of that image for a moment)
depth of field will be the same for any 300mm lens,
for example,
whether shot on a digital or a 10x8.

The difference is,
not many people will consider a 24x36mm image to be the finished product.
whereas, with 8x10, it might be.

Apart from an unsuccessful bid to flog a lens,
this is my first post-
I must admit to being a little intimidated by the absolute erudition of the contributors to this site, and I find it to be an invaluable resource.

Now that I've popped my posting cherry,
maybe the next one wont be so difficult-

joseph