Hyperfocal Distance & Circle of Confusion

[cooltouch]

Man-O-man. When I first learned about hyperfocal distances, I learned about how to use the index marks on my lenses that shows a hyperfocal range, not a specific distance. The whole point of using a lens's hyperfocal scale was being able to set the focusing ring to a setting such that everything from a certain point to infinity, at a given f/stop, was in focus. When in doubt, add in a bit of extra windage, just to be sure.

I guess there's something to be said for old school lenses that still have aperture rings so one can eyeball a hyperfocal distance. Assuming the HF scale is still there, of course. And focusing rings with distance scales, far as that goes. Their analog displays are not unlike slide rules or the vernier readouts on calipers. As time goes by, I find myself appreciating my old manual focus gear more and more. No computational equipment required to render a decent photo.

[Patti Edens]

Wil, Take a look at Doug Kerr's article on the subject...
The Pumpkin

He speaks "engineer" too.

p

[PlanetFinder]

As nostalgic as some might be for an earlier, simpler time when camera manufacturers etched marks for hyperfocal distance on the lens barrel without regard to the resolution of the focal-plane detector, those times are gone. Notwithstanding the assertion of DoFMaster.com and Pumpkin, that the diameter of the circle of confusion (CoC) does NOT depend on pixel size or wavelength, that cannot be true, and committing the error of believing it is true can only lead to out of focus images on camera systems of adequate image quality and sampling.

This is not a small effect. The hyperfocal distance varies inversely with CoC, so the factor of two being debated here, between CoC = 0.030 mm and CoC = 0.015 , is a factor two in hyperfocal distance.

It is easy enough to discover who's right with a simple experiment, as I did this morning. Go to Online Depth of Field Calculator (or use the DoFMaster iPhone app), and calculate the case of the D3x plus lens with focal length = 70 mm, focal ratio = 2.8, and a focus distance = 91 inches. The calculator says the depth of field is 7.07 inches.

It is easy to verify that the calculator is using CoC = 0.030 mm for both the D3 and D3x. These cameras have pixel spacings different by factor sqrt(2), and I am saying they cannot have the same CoC.

Well, let's go outside and see who is right.

I took the accompanying picture of a measuring tape with the geometry and camera settings just stated two paragraphs above. The angle between the camera's line of sight and the wall was 14.6 deg.

If you compare the marks at 42.5" and 38.5" with the one at 40.5", I think you will agree that the in-focus field is smaller than 40.5" ± 2". But that's along the tape, so the depth of field (DoF) is ±1.9" along the camera's line of sight. That's just over HALF the DoF we calculated using the erroneous assumption of CoC = 0.030 mm for the D3x. The correct CoC is obviously more like Coc = 0.015 mm for the D3x.

Notice that the incorrect estimate of DoF = ± 3.5" isn't even within the magnified region of my picture! If you relied on that DoF when composing a picture, the result would be out of focus.

[tertius]

Originally Posted by Wil_Bloodworth

As I was memorizing my hyperfocal distances for various focal lengths at my most-used apertures, I came across the common accepted value of 0.03 for all of the Nikon full-frame sensors.

Originally Posted by jeffkohn

The .018 was something I chose just based on the fact that my enlargement factor for 24mp at 20x30" is slightly less than in the 12mp case. It seems to work reasonably well.

Originally Posted by Wil_Bloodworth

Good to know. Thanks. I will adjust accordingly.

- Wil

Originally Posted by PlanetFinder

Based on this, I think CoC = 0.012 mm is more reasonable estimate of the true value.

Originally Posted by PlanetFinder

Well, let's go outside and see who is right....

....The correct CoC is obviously more like Coc = 0.015 mm for the D3x.

So both were delta .003 mm (assuming 0.015 is correct). I don't care one way or the other right now. So who was right and why can't we stay inside for this test?

[Wil_Bloodworth]

Kyle,

There is really no absolute right or wrong. As Jeff so correctly pointed out, the end result (print size) dictates what you would want to choose as your CoC for determining HFD. The only obvious answer is that 0.03 is totally wrong and should not be used in any calculations.

- Wil

[donlfaulkner]

I, for one, can certainly understand why it is called "circle of confusion".