DoF and ND filters

Chris Durham · August 15, 2007

I'm trying to get something straight in my head. I know that DoF is a function of both the capture surface and the amount of light. So opening the aperture as much as possible is beneficial to minimizing it; as would be using something larger than a 1/3" but that isn't an option for me. So if I open the aperture and then use ND filters to control exposure, does that defeat the purpose? Or does this have more to do with the dispersion of light in a wider aperture?

I'm confused.

Would another way of controlling this be setting the gain on my camera to -3db?

David Auner aac · August 15, 2007

Hi Chris!

Take a look at this article: http://en.wikipedia.org/wiki/Depth_of_field. It should clear up any questions you might have. If not, shoot away! You might also check out this thread: Pulling Focus

Cheers, Dave

John Sprung · August 16, 2007

So if I open the aperture and then use ND filters to control exposure, does that defeat the purpose?

No, that's exactly the right thing to do. Set the aperture to get the DoF you want, then use lighting and ND to get the exposure right.

Note that with a three chip camera, the beam splitter prism block limits your shallowest stop to about f/1.4. Also, note well that with 1/3" chips, you lose resolution to iris diffraction at stops deeper than f/4. So, that's your optimum range for that camera, quite a narrow one, just three stops wide.

-- J.S.

Daniel Sheehy · August 16, 2007

Would another way of controlling this be setting the gain on my camera to -3db?

The -ve gain would have the same effect as using ND on the lens. So, yes, pick your aperture and control the light outside of the camera, to ensure exposure for that aperture.

Michael Nash · August 16, 2007

Depth of field is controlled by three things (four really, but I'll explain): focal length, aperture and distance to subject. If you want to use a wider aperture for less d.o.f., you can use a ND filter to correct your exposure. It's done all the time. -3db gain only cuts the light 1/2 stop, which may not amount to much.

Depth of field is not a function of the size of the capture surface. That actually has nothing to do with it, at least directly. It's just that with a smaller image sensor you use shorter focal lengths to maintain the same angle of view as you would get with a larger format. So it's the focal length, not the sensor size, that's giving you more depth of field.

The "fourth" factor is the circle of confusion, which essentially is a measure of the resolution of the system you're using. In a nutshell, the higher the resolution of the format, the less tolerance there is for acceptable focus. Some marginal focus on the big screen might be acceptable on a lower-res display, like standard def TV.

Chris Durham · August 17, 2007

Note that with a three chip camera, the beam splitter prism block limits your shallowest stop to about f/1.4. Also, note well that with 1/3" chips, you lose resolution to iris diffraction at stops deeper than f/4. So, that's your optimum range for that camera, quite a narrow one, just three stops wide.

Really? I had no Idea about that. I'm doing some tests on Monday so I'll have to check that out. Of course, I probably won't notice it on a TV though...

The "fourth" factor is the circle of confusion, which essentially is a measure of the resolution of the system you're using. In a nutshell, the higher the resolution of the format, the less tolerance there is for acceptable focus. Some marginal focus on the big screen might be acceptable on a lower-res display, like standard def TV.

This makes sense, though I never thought about it really. I guess another way of thinking about it is that the smallest possible circle of confusion in digital video is the size of a pixel.

So I guess, that with these things in mind, the logical question iswhat is the effect of this in SD? In other words, is the resolution loss due to iris diffraction such that the circles of confusion are larger than an SD pixel on a 1/3" CCD? I might be using terminology a bit incorrectly, but I think you get the point. I imagine that iris diffraction wider than f/4 might impact a 1/3" HD chip worse than SD.

Thoughts?

Daniel Sheehy · August 17, 2007

The "fourth" factor is the circle of confusion, which essentially is a measure of the resolution of the system you're using.

The Circle of Confusion is not absolute. You can decide to use a CoC that is smaller than the conventionally accepted on.. and while it may give you a theoretically narrower depth of field, it has no impact on what the viewer sees.

http://www.cinematography.com/forum2004/in...2713&hl=DoF

http://www.cinematography.com/forum2004/in...8738&hl=DoF

http://www.vanwalree.com/optics/dof.html

For discussions & arguments over this issue.

Edited August 17, 2007 by Daniel Sheehy

Michael Nash · August 18, 2007

So I guess, that with these things in mind, the logical question iswhat is the effect of this in SD? In other words, is the resolution loss due to iris diffraction such that the circles of confusion are larger than an SD pixel on a 1/3" CCD? I might be using terminology a bit incorrectly, but I think you get the point. I imagine that iris diffraction wider than f/4 might impact a 1/3" HD chip worse than SD.

Thoughts?

An HD camera has more resolution than an SD camera plain and simple, and will reveal smaller/finer focus errors.