f stop and depth of field

Rob Featherstone · April 29, 2009

Can any one explain exactly why a smaller lens opening results in greater depth of field?

I imagine it has to do with less scattered light hitting the target.

Thanks!

Paul Bruening · April 29, 2009

Here's a good start:

http://en.wikipedia.org/wiki/F_stop

http://en.wikipedia.org/wiki/Depth_of_field

Hal Smith · April 29, 2009

The Wikipedia articles tell you all about how to calculate DOF, etc. but they don't explain WHY. And frankly, even though I've taught Physics at the prep school and college level and know some optical theory, I don't know why when f-stop is decreased to smaller values DOF decreases and that it's a universal concept independent of lens design.

I've got a copy of Applied Photographic Optics (all $140 worth of it). I'm going to have a look and see if it covers the "why".

Patrick Neary · April 30, 2009

Hi-

In very simple terms you are constricting the light rays (from your subject) as you close the aperture; think of a pinhole camera with virtually unlimited dof.

Tom Jensen · April 30, 2009

Hi-

In very simple terms you are constricting the light rays (from your subject) as you close the aperture; think of a pinhole camera with virtually unlimited dof.

True.You have spherical aberration and diffraction. When you stop down, you have less spherical aberration and more. At some point you will get a sweet spot where the lens is at maximum performance. Not only does the dof increase but the depth of focus which is the area behind the lens that lands on the film plane. It works in converse of focal length. Long lenses have small dof and a large depth of focus. Wide lenses have a lot of dof and little depth of focus. This is why the flange focal depth of the camera is so important. If you are out of the country and you think your flange is off, stick to the long lenses stopped down

Michael Collier · April 30, 2009

Read Kris Mankewitz's 'Cinematography'. That was the book where I had the ahha moment where I finally understood why. I could try and explain it fully, but I really need the diagrams in the book, and I am not about to scan those and rip them off (not with one of the co-authors on the board anyway).

Basically, without getting too much into the details, out of focus areas are a projection of the shape of the iris. If you see iris' bokeh that looks perfectly circular, the lens is likely wide open. If not, you can probably count the blades in the iris. If you imagined a point of light out of focus, keep in mind the projection idea, and as you iris up or down, that point of light gets bigger or smaller in relation to the size of the hole (bigger as you iris up, smaller as you go down). Now if each of those points of light were a point of detail, as you iris up you get more overlap and it seems more out of focus. Iris down and you see the overlap decrease, and focus seems sharper (even though hyperfocal and distance to point hasn't changed.) Make those circles smaller than the circle of confusion of your given format, and (presumably) that is now in focus, since the circle is smaller than the resolution of the format, or more correctly, smaller than the 'acceptable' sharpness limit of the format.

Buy the book and study the diagrams to see why the iris projects itself onto the emulsion. Take some time with it. I think I studied and thought about it for 2 days before I got it, but at that time I was 15, so you might get it a little quicker.

Tom Jensen · April 30, 2009

Read Kris Mankewitz's 'Cinematography'. That was the book where I had the ahha moment where I finally understood why. I could try and explain it fully, but I really need the diagrams in the book, and I am not about to scan those and rip them off (not with one of the co-authors on the board anyway).

Basically, without getting too much into the details, out of focus areas are a projection of the shape of the iris. If you see iris' bokeh that looks perfectly circular, the lens is likely wide open. If not, you can probably count the blades in the iris. If you imagined a point of light out of focus, keep in mind the projection idea, and as you iris up or down, that point of light gets bigger or smaller in relation to the size of the hole (bigger as you iris up, smaller as you go down). Now if each of those points of light were a point of detail, as you iris up you get more overlap and it seems more out of focus. Iris down and you see the overlap decrease, and focus seems sharper (even though hyperfocal and distance to point hasn't changed.) Make those circles smaller than the circle of confusion of your given format, and (presumably) that is now in focus, since the circle is smaller than the resolution of the format, or more correctly, smaller than the 'acceptable' sharpness limit of the format.

Buy the book and study the diagrams to see why the iris projects itself onto the emulsion. Take some time with it. I think I studied and thought about it for 2 days before I got it, but at that time I was 15, so you might get it a little quicker.

What? When you are using a lens wide open, you are seeing light gathered from the entire lens including the edges. When you stop down you are eliminating the light on the edges and the light is coming more from the center. I don't think you are focusing on the iris.

Edited April 30, 2009 by Tom Jensen

John Sprung · April 30, 2009

The drawings in this article give you a handle on it:

http://en.wikipedia.org/wiki/Circle_of_confusion

You can see from them how a point at some distance other than the plane of critical focus gets rendered as a circle, and that the diameter of that circle depends on the diameter of the simple lens in the illustration. Stopping down a complex lens basically makes it act as if it had a smaller diameter.

-- J.S.

Dan Diaconu M · April 30, 2009

Although answered already, for the "why" part is because any smaller section of the lens is closer to the design than the whole lens (hence the high price of fast lenses/primes and the less expensive digital zooms starting at 4/5.6).

But I have another question: from a practical point; how does one measure the confusion? Some people like to add a personal touch to their images http://fc01.deviantart.com/fs27/i/2008/107...by_pixel_ah.jpg

and those are far from circles. Daytime images (without lights) still show a fair amount of bokeh (confusion) in their natural shape (not necessarily circles)

Is there an apparatus (caliper or something) to measure the confusion?(charts aside) How do you do it?

The confusion grows with the projected screen size and distance (not only lens, aperture and originating format).

Help, I am diffused.

Dominic Case · April 30, 2009

I haven't seem a simple answer to the question yet.

Forget aberration, where the iris is and all the rest. Forget high cost lenses. Forget sweet spots.

Light from a point souce at a certain distance spreads out. The lens collects it and refocusses it to a point on the film plane IF the object is correctly in focus. So you have a cone of light spreading from the point to the lens, and another cone returning to a point on the film plane.

For any point source closer or further away, the point of focus is further or closer to the lens (which can only bend light to a certain amount. So when the cone hits the film plane, it isn't a point but a small circle of light.

If the lens has a smaller aperture (higher f/stop value) then the light is formed as a narrower cone. So the circle of light on the film plane is smaller. The point source object is closer to appearing in focus.

If you accept small circles up to a certain size as being "in focus" then a narrower cone of light means that a greater range of distances are "in focus".

That's all.

In a diagram it would be even more straightforward.

Tom Jensen · April 30, 2009

I haven't seem a simple answer to the question yet.

Forget aberration, where the iris is and all the rest. Forget high cost lenses. Forget sweet spots.

Light from a point souce at a certain distance spreads out. The lens collects it and refocusses it to a point on the film plane IF the object is correctly in focus. So you have a cone of light spreading from the point to the lens, and another cone returning to a point on the film plane.

For any point source closer or further away, the point of focus is further or closer to the lens (which can only bend light to a certain amount. So when the cone hits the film plane, it isn't a point but a small circle of light.

If the lens has a smaller aperture (higher f/stop value) then the light is formed as a narrower cone. So the circle of light on the film plane is smaller. The point source object is closer to appearing in focus.

If you accept small circles up to a certain size as being "in focus" then a narrower cone of light means that a greater range of distances are "in focus".

That's all.

In a diagram it would be even more straightforward.