View Full Version : Depth of Field + Lens Size + Enlargement Factor

Ken Lee
16-Jul-2004, 11:04
We often read that longer lenses give less depth of field.

When considering large format lenses, we also need to consider the fact that enlargement is correspondingly less.

So, even though a 300mm lens gives less depth of field than a 50mm lens, what if we compare an 8x10 contact print to an 8x10 print enlarged from a 35mm image taken with a 50mm lens.. Would the 8x10 enlargement still appear to have more depth of field than the 8x10 contact print ?

Robert J Cardon
16-Jul-2004, 11:43
Sharpness or DOF? It really depends on the subject/picture being taken - how the plane of focus corresponds to the elements in the shot. And what does it matter? Who's shooting 35mm?

Content, content, content!


16-Jul-2004, 11:49
Depth of field is a brain bender for me, so whenever I am trying to figure out which format/lens combination to use when I want to maintain or maximize depth of field, I run calculations with online depth of field calculators (or DOF Master on my PDA) to see what hyperfocal distance I will be dealing with. Both:

http://dfleming.ameranet.com/dofjs.html and http://www.silverlight.co.uk/resources/dof_calc.html

allow you to select your film size to work out near/far hyperfocal distances for a given distance of focus. For example if you plug in: 35mm format, 50mm lens, f22 aperture, 25ft focus distance and compare those numbers to numbers generated by using: 8x10 format, 300mm lens, f22 aperture, 25ft focus distance; you will see that DOF shrinks considerably. Cheers,

Jay DeFehr
16-Jul-2004, 12:04
Ken, while your question is a valid one, you've left some important variable out of the equation, maybe intentionally. lenses for 35mm formats become diffraction limited beyond f22, while a 300mm lens for 8x10 format can stop down to f 90 or beyond. More important are view camera movements which allow manipulation of the plane of focus. These factors combined with the enlargement factor cited in your question combine to eliminate most of the dof advantages enjoyed by smaller formats, and that says nothing of the other, equally important advantages enjoyed by larger formats.

Alan Davenport
16-Jul-2004, 12:08
Even if you allow for different size circle of confusion, the 35mm will have greater DOF.

For a subject 20 feet away (focused at that distance:)

35mm format, 50mm lens, f/16, CoC=0.025mm: DOF extends from about 10 feet out to 620 feet.

8x10 format, 300mm lens, f/64, CoC 0.188mm: DOF about 11 feet out to 88 feet.

Trouble is, this is all apples to oranges. With the 8x10 camera, you may be able to use a judicious tilt or two to make the apparent DOF seem infinite. And don't forget the subjective impact of a contact print vs an 8X enlargement...

Bob Wheeler wrote an interesting article dealing with the differences in DOF in small formats vs large, it can be seen at:

Ken Lee
16-Jul-2004, 12:20
Let's try to rule out any extraneous considerations, like view camera movements, apertures, etc.

I am not trying to determine what is better or easier, but just understand if the additional depth of field derived from shorter lenses, is compensated for, in larger formats, by an equivalently lower degree of enlargement. Forget about grain, lens quality, etc. I'm just talking about the optical principles themselves... I think. ;-)

Perhaps Professor Evens can throw some light on this one.

Jay DeFehr
16-Jul-2004, 13:33
Okay, if you want to make a straight comparison with the only variables being enlargement factor and focal lengths, ie same aperture for both lenses, both focussed at infinity, which Circle Of least Confusion figure would you use? The one consistent with contact printing, or the one required for enlarging? You see, there's really no way to make the comparison you want to make without biasing the results one way or the other, which is why all of the applicable factors should be considered.

Ken Lee
16-Jul-2004, 14:06
Let's say we set up both cameras along a picket fence, focus on one spot, and shoot both images at the same f-stop. Comparing the 8x10 contact print with the 8x10 enlargement, will there be an obvious difference that an average viewer would notice ? The difference between a 50mm and a 30mm lens is quite a bit - but is that difference ameliorated by the 8x enlargement ?

Ken Lee
16-Jul-2004, 14:07
Sorry - I meant to say that the difference between a 50 mm and 300 mm lens is quite a bit.

16-Jul-2004, 14:23
Ken, the shallow depth of field of the 300mm lens on 8x10 would make most of the picket fence out of focus, which on a contact print would look far far worse than the loss of sharpness by enlarging a 35mm frame 8x, (assuming that the fence is mostly in focus due to the greater depth of field from a 50mm lens in the smaller format). This is assuming that both lenses are stopped down to say, f16. Aperture will always play a part. If both lenses are wide open, the difference between formats will be less.

16-Jul-2004, 14:39
Let me take the picket fence mental picture. I think it's helpful to think of this in terms of how the relationships change as things scale. Let me elaborate:

Imagine a 35mm camera with a 50mm lens set up to take a picture along the picket fence. Now scale things up to the large format case. Let's assume that all the linear measurements just get multiplied by a factor of 6: we're just giant-sizing the world. The lens becomes a 300mm lens. The film grows to be something along the lines of 5x7. We'll even permit a circle of confusion that is 6 times larger. Everything is exactly the same as before, except the fence.

Were this to be a complete scaling, we'd have to enlarge the world by a factor of 6 as well. At that point, the resulting image would be the same as in the small format case (ignoring factors like diffraction, I suppose). We've not done that. Relatively speaking, the fence is much closer to the large format setup, and we all know that there's less depth of field as the subject gets closer.


Leonard Evens
16-Jul-2004, 15:17
There is a lot of confusion about depth of field, and a lot of what you see said about it is just plain wrong.

See Bob Atkins's write-up at www.photo.net. It is pretty clear and one of the few sources which gets things right.

Let me try to address the specific question you asked. I'm sorry if it is complicated, but these things are quantitative and depend on specific formulas. Qualitative discussions are too general to be able to tell you just what is going to happen. The problem is that in some circumstances one thing will happen but in other related circumstances, the opposite will happen. Qualitative discussions aren't flexible enough to account for everything that can happen.

I will ignore diffraction, which can also be an issue. I will also suppose you don't use movements.

Suppose you do the following. You take two pictures with different size format cameras of a subject at the same distance from the camera, you arrange the focal lengths so the angle of view is the same, and you use the same f-stop. Finally suppose the final print is the same size in both cases, which implies different degrees of enlargement. Then the print from the larger format will have less depth of field. This can be made more precise. You would have to multiply the f-number you used in the smaller format camera by the ratio of the format sizes (also the ratio of focal lengths) to use with the larger format camera to obtain the same depth of field. Suppose for example the ratio is 4, and you used f/2 with the smaller format camera. Then you would have to use f/8 with the larger format camera.

But your example is a bit more complicated. First, the formats don't have the same aspect ratio, so you would have to be specific about how you were going to crop the prints. Suppose you make an 8 x 10 contact print from the LF format, and you also crop the 35 mm image to the same aspect ratio. That means you would only be using 24 x 30 mm of the smaller image. I don't know the precise size of an 8 x 10 negative, but let's suppose it is about 240 mm in the long (10 inch) direction. Then the ratio of the format sizes would be about 240/30 = 8. That means you would GAIN that much depth of field because of the difference in enlargement factors. That is, you could divide the f-number for the 35 mm camera lens by 8 in order to get the f-number for LF camera lens. If you used f/8 with the 35 mm lens, you could use f/1 with the LF lens, at least if the degree of enlargement were the only factor. BUT that ignores the difference in focal lengths of the two lenses. You have to compensate by also multiplying by the SQUARE of the ratio of the focal lengths of the two lenses. If the two lenses had the same angle of view, this would be that same ratio 8, so you would multiply by 64, which would cancel out the division by 8 and leave a multiplication by 8, as suggested above. So you would have to use f/64 with the LF lens.

But the two lenses you describe do not have the same angle of view. The ratio of the focal lengths is instead 300/50 = 6, so you would have to multiply instead by 36. The net effect would be instead that you would use f/36 with the LF lens. (The combined multiplier would be 6^2/8 = 36/8 = 4.5 instead of 8.) So you would still have less depth of field, but not to the same degree as mentioned above.

Since the two images wouldn't have the same angle of view, one can ask other questions. The answer would depend on just what you are doing.

For example, suppose you were making a portrait. You could move closer with the LF camera with the 300 mm lens so the image size on the print was the same as that obtained by the 35 mm camera with the 50 mm lens. For a fixed format, same size final print, and constant image magnification, the depth of field is approximately independent of focal length, at least for subjects not too far from the lens. So the result of doing that would be close to what you would get at the original distance with a 400 mm lens---ratio 400/50 = 8). But the f-number multiplier in that case would be 8, as discussed above, so that is just about what it would be with the 300 mm lens being closer. You would be back to using f/64 for the same depth of field.

On the other hand, if you were making a picture of a landscape, the situation might be very different.

Ken Lee
16-Jul-2004, 17:37
Thank you Leonard, for the superb explanation.

Is the following a correct understanding ?

Given prints of the same size, lenses of the same angle of view, same f-stop, and same distance to the subject, then to obtain the same depth of field, we have to multiply the f-number used in the smaller format camera by (the ratio of the format sizes) AND (the square of the ratio of the focal lengths of the two lenses).

If we are comparing a 150mm lens on a 4x5 to a 300mm lens on an 8x10, then the ratio of format sizes is 2:1. The ratio of focal lengths is also 2:1, so the square of that is 4:1. Therefore, to obtain the same depth of field, we would need to multiply the f-stop by 2 * 4, or 8.

Thus, all things being equal as stated above, an 8x10 contact print, made with a 300mm lens at f/64, will have about the same depth of field as an 8x10 print made from a 4x5 negative and a 150mm lens at f/8 - which isn't a whole lot !

... is that right ?

Leonard Evens
17-Jul-2004, 09:35

Apparently my explanation was not as clear as I hoped. Suppose A is one format (focal length) and B is the other. To convert from A to B, you MULTIPLY by the ratio of format lengths of A to B and DIVIDE by the square of the ratio of focal lengths of A to B. In your example, take A to be the 8 x 10 combination and B the 4 x 5 combination. Both ratios are 2:1 as you indicate. So multiply 64 by 2 to get 128 and divide that by 4 to get 32. So f/64 with the 8 x 10 is equivalent to f/32 with the 4 x 5.

You can keep most of this straight by remembering the following things. Use the ratio of the format lengths, and use the square of the ratio of focal lengths. These two effects work in opposite directions, so you multiply by one and divide by the other. To remember which to multiply by and which to divide by, remember that generally going from smaller to larger formats decreases depth of field, other things being equal, so f-numbers have to go up to compensate.

However, that rule is not foolproof. If the ratio of the focal lengths is low enough, the effect of the ratio of the format lengths will predominate. For example, in the extreme case where you keep the same focal length for both formats, you will actually get more depth of field with the larger format than with the smaller format. I wish it were simpler, but as I said, it is based on a quantitative formula which can give qualitatively different results in different cases.