PDA

View Full Version : Depth of Focus - Wide Angle

Wilbur Wong
21-Oct-2005, 09:32
I think I have a better understanding now from the earlier thread.

Specifically in the use of wide or superwide lenses, the "circle of confusion" works in conjunction with the "cone of acceptable focus". My understanding is that the smaller the f stop, the narrower the "cone" will be, and therefore will increase the tolerance of how far from the plane of focus, the film may be permitted to deviate (given the same tolerances for eventual print sizing etc.).

I think I understand that this angle of the cone is independent of focal length and is related strictly to focal ratio (f stop). Is that correct?

If the above is true, isn't there an additional factor which comes into play with wide angle lenses? Billiard balls elongate into elipses in the corners of wide angle shots, I think in inverse of the tangent of the angle off axis from the center of the picture. Doesn't the cone of focus also elongate similarly? It seems to me that if it does, than the tolerance for depth of focus becomes even less at the edges of a wide angle photograph as the film will fall off of that cone with less distance from the plane of sharp focus.

Of course this is where I have increasing difficulty viewing the image, both visually on the ground glass, and with a loupe. I admit to doing as JJ mentioned in a previous post that many photographers simply stop down as far as possible, with some frequency.

What do you think?

Bill_1856
21-Oct-2005, 09:45
What do I think? I think that you're overanalyzing.

John Cook
21-Oct-2005, 09:54
This is the sort of technical thing my art school avoided teaching. Not sufficiently emotional. I recall, during a first-term State-mandated math class, someone saying that the answer to the problem on the blackboard. “Feels like six”.

But my impression is that the image projected out from the back of a single-element uncorrected lens focuses in an arc. The apex of which is at the optical center of that lens. In other words, the best image would be that projected onto the inside of an old Beetle hubcap, painted flat white.

Even a well-corrected multi-element objective has difficulty with the extreme angle of approach of the projected image in the negative corners from a closely-positioned wide angle lens. Shoot a water pistol at the wall at a great angle and see the oval water pattern. In the same way, the image tends to “smear” creating those oval billiard balls.

Not sure how I would relate this phenomenon to circle of confusion.

As a practical working photographer with a long shot list and toe-tapping studio owner boss, it has always been sufficient for me to realize that depth of field was a function of the image size on the film.

Both long/telephoto and large format lenses project a larger image with shallower depth of field than wide angle and 35mm lenses. Only way around that, besides a smaller f/stop, is to back up the camera, decreasing the image size.

ronald moravec
21-Oct-2005, 10:06
Stretching in the corners is due to the light cone hitting the film plane at an angle. Shine a flashlight on a table (torch if you are UK) perpindicular and at an angle to demonstrate.

You can observe the corners with plain ground glass by moving you eye off axis.

Debth of field is the same for any given picture segment, ie up close with a wide, further back with a tele, as long as the lens sees the same picture.

Brian Ellis
21-Oct-2005, 11:17
Three factors affect depth of field: distance from subject, focal length of lens, and aperture. The greater the disance of the camera from the subject, the longer the focal length of the lens, and the wider the aperture the less the depth of field and vice versa, assuming that everything else affecting depth of field remains the same (i.e. that when you change one of these three factors you don't change either of the other two, that the magnification factor remains the same, etc.). .

Only one plane can be in focus (i.e. represented by points rather than circles) at any given time. Everything on either side of that plane is represented by circles and is out of focus. However, the circles may appear to be "in focus" if they are close enough to the plane of focus, if they are small enough, if the print is small enough, if our distance from the print is great enough, etc. We control the plane of focus by using tilts and swings and we control the size of the circles on the film by the three factors mentioned above, i.e. distance, focal length, and aperture.

I'm sure all this isn't totally accurate from a purely scientific point of view (e.g. use of the term "points") and anyone more knowledgeable about the science behind it could pick holes in some of it. However, I think it's sufficiently accurate for practical photographic purposes and I think it's about all a photographer needs to know in order to understand depth of field and use it in his or her photography.

You used the term "depth of focus." FWIW, I don't worry about depth of focus because there's nothing I can do about it once the lens is on the camera (or if there is I haven't heard about it and I don't think I want to). I worry a lot about depth of field because that's something used to control the look of the photograph. Do I want everything from front to back to appear sharp? Do I want only the subject to appear sharp with everything in front and in back of it to appear blurry? If so, how blurry?

I skimmed but didn't study your previous thread. If I had I perhaps would know what you mean by "cone of acceptable focus" and its relationship to the circles of confusion but I don't know what you're referring to. I don't offhand remember ever seeing the term "cone of acceptable focus" before and I don't know how it would relate to the circles of confusion.

Dan Jolicoeur
21-Oct-2005, 11:30
"what do I think"

Go take some pictures, try different things. That is the only way you are going to really learn and establish a style of your own. Then look at them yourself with a loupe.

Richard Schlesinger
21-Oct-2005, 12:56
Stroebel (7th edition) has an excellent explication of depth of focus on pages 159-160. In essence:

1. "Depth of focus increases as a lens is stopped down."

2. "Depth of focus increases as object distance decreases."

3. "Depth of focus is not affected by focal length."

4. "Depth of focus increases as the film size increases."

Dan Fromm
21-Oct-2005, 13:06
Wilbur, let's go at this from a different angle.

What would you do differently if you were sure you knew what depth of field and depth of focus meant? What are you trying to control, and why?

Cheers,

Leonard Evens
21-Oct-2005, 15:00
If you are still talking about depth of focus, as opposed to depth of field, then what you say is correct. Depth of focus depends just on the choice of circle of confusion, the f-number (i.e., the relative aperture), and the magnification. The way the magnification enters is as a factor 1 + M and for distant objects the magnification M is so close to zero that you can ignore it. So for distant objects, only the relative aperture and the choice of circle of confusion matter. For close-ups, the magnification can play an important role. For example, at 1:1 the total depth of focus would be doubled. Since magnification is related to the ratio of subject distance to focal length, you could say that for close-ups focal length does play an indirect role.

Your geometric picture is more or less right, but you should draw some pictures, label them, and play with similar triangles if you want to get it precisely right.

The issue of "distortion" with wide angle lenses at the edges is another matter. There is no distortion of circles which are in planes parallel to the film plane. Imagine taking a picture of a flat wall with the back parallel to the wall and the lens axis perpendicular to the wall. If you have a series of equal sized circles on the wall, they would be reproduced as equal sized circles in the image. For three dimensional objects it is a bit more complicated. Often a sphere near the edges will appear to be ellipsoidal, but this is not really a distortion. It results from the camera position relative to the sphere. If the sphere is far enough to the side, by the laws of perspective it will appear to be ellipsoidal. But when you look with your eye, you can't see that far off to the side without turning your head, and then you are looking at the sphere head on, so it doesn't appear distorted. When you take a picture, you need a wide angle lens to include such a sphere in the image. If you made a print and then placed your eye relative to the print in the same position that the camera was relative to the scene, you would also not be able to see the sphere. So you move back to see the whole scene by viewing the print at a distance appropriate to the print rather than the center of perspective of the scene. That introduces an illusion of distortion, although the sphere is shown exactly as the laws of perspective demand.

There is another kind of "distortion" which results from the use of wide angle lenses and viewing prints made from them at a normal viewing distance determined by the print. Namely everything even moderately far away looks too small. You lose the sense of scale you saw in the original scene. For me, this is a much more serious problem than ellipsoidal spheres and related "distortions" of three dimensional subjects. Skilled photographers know how to compose the scene to minimize such effects and provide a convincing rendition of the scene. I work at this myself, but I must admit I seldom accomplish it.

Wilbur Wong
21-Oct-2005, 15:44
Thank you Leonard for addressing the "depth of focus" issue, I am sorry my billard ball illustration led some posters away from my real question, but it was only to illuminate my thought regarding depth of focus at the edges of a wide angle shot.

My real question here is when the cone of focus is NOT normal to the film plane when it intersects the film far from perpendicular to the film, is the depth of FOCUS actually less than it would be near a normal intersection to the film plane?

Leonard Evens
22-Oct-2005, 08:28
Wilbur,

Each point in subject space will produce a cone with base on the aperture, assumed circular, and vertex somewhere in image space. The vertex is where the subject point produces an image point. (Of course all this is an idealization only approximated by a real lens system, but it is pretty close.) If the film plane is parallel to the lens plane, i.e., no tilts or swings, and it is slightly off from the image point, then the film plane intersects the cone in a circle. If I understand your question, you are convinced of this fact when the image point is on the lens axis, but you are asking if it is still true if the image point is off the lens axis. It is. That follows from the basic fact, that any cone, not necessarily a right circular cone, is intersected by parallel planes in similar curves. If one of a pair of parallel planes intersects a cone in a circle, then the other does also. (This doesn't, by the way, depend on which side of the image point the film plane is. Instead of a single cone, you have to consider the double cone with vertex at the image point. In this application that would consist of all complete lines passing through the vertex image point and the circumference of the aperature. )

When the lens plane is not parallel to the film plane, the situation is quite a bit more complicated. The problem is that the section of the cone in the film plane is no longer a circle, but instead an ellipse. Moreover, the exact shape and orientation of that ellipse depends on just where in the film plane you are. I believe it is still true, to a high degree of approximation, at least in normal circumstances, that you can choose two planes equidistant from the exact image plane such that within those two planes, the focus is adequate. In other words, in most circumstances, the depth of focus about the exact subject plane should not vary signficantly by position in the film plane. But I have to admit I haven't thought this out completely but am relying on what I have manged to figure out for the related question of depth of field in case of tilts or swings.

Michael S. Briggs
22-Oct-2005, 13:24
Wilbur, what I don't understand is why you are concerned about depth of focus. Unless someone has mispositioned a focusing screen, I don't believe that it is an issue in practice for 4x5 photography. The depth of focus at the usual taking aperture is so large compared to what the equipment tolerances are supposed to be, that there shouldn't be any problem. I gave some calculated numbers in an answer to one of your previous questions.

If the real issue is depth of field, it would be better to directly discuss depth of field then its conjugate, depth of focus.

If your problem is seeing the the image from a wide-angle lens in the corners of a screen, with a loupe? What I do is use a magnifier, so that I can tip the magnifier to be perpendicular to the rays. Perhaps the Silvestri's Tilt Loupe would help, I've never used one. Others use screens with Fresnel lenses, but I don't like the lines.

Wilbur Wong
22-Oct-2005, 21:23
Thank you Leonard and Michael for staying with me. I think that often I am on this forum to keep LF photography in my psyche when I am unable to get out and actually shoot. Being a self employed general contractor, I can't get out more than a few times a year.

Leonard, as I have thought about it more, I kinda of realize that the off axis cone should be narrower in angle which helps equalize the increasing size of the intersected cone due to it's tangental aspect. So in average it might be a moot point.

Michael, I modified my 4x schneider a few years ago to allow me to view the corners of the ground glass similar to the silvestri's ability. But to be honest, I seldom like to use a loupe, instead relying on my 3.5 plus readers. (60 year old eyes). And I am annoyed at at picture I took a few years ago of some pounding surf, during a pacific storm which has one corner out of focus. Of course it had brilliantly lit waves, gorgeous clouds (infinity) with the opposite catty corner having the closest DARK rocks which are out of focus.

Well, I do think that I have learned from the previous experience, as well as your contributions.

Thanks,

BTW, tomorrow morning I am leaving for a WEEK of photography meeting Per Volquartz's group, and I intend to enjoy it.