bclaff
•

Forum Pro
•
Posts: 11,570

Re: A way of thinking about it

Mark Scott Abeln wrote:

bclaff wrote:

What you have described is closer to depth of focus than depth of field.

However, as small magnifications (longer distances) the differences are small.

(As are the effects of pupil magnification.)

I'm still stuck in the dark ages of the symmetric thin lens approximation.

Even so you can't use that explanation.

Although this is the Beginners Questions forum let's try this.

Here's a diagram from a currently unreleased interactive widget on PhotonsToPhotos:

Best viewed "original size"

You describe the triangle on the left from the entrance pupil (P) to the in focus object on the object plane (O). So far, so good.

But Depth Of Field (DOF) depends on the size of the blur disc on the image plane (I).

So we must also consider the triangle on the right from the exit pupil (P') to the image plane (I).

I've drawn pupil magnification of 1.2 but if this were the thin model pupil magnification would be 1; P' would be the same height as P and overlap P.

We know the right triangle doesn't really come to a point but to an Airy disc; but we will ignore that (diffraction). We also ignore aberration.

Do not extend either triangle but rather consider different points on the optical axis in object space.

A point slightly closer than the in focus point at O will focus slightly to the right of the image plane (I) and this triangle will be cut by the image plane (I) to form a blur disc.

So long as that blur disc is less than or equal to the Circle Of Confusion (COC) that point on the object side will appear to be in focus.

Similarly a point further than the in focus point at O will focus slightly to the left of the image plane (I), cross and be cut by the image plane (I) to form a blur disc.

You can find the near and far points in object space by tracing the blur disc on the image plane (I) back into object space.

The key thing is that these are different triangles, not extensions of the in focus triangles.

This is true regardless of thin versus thick model.

For an in depth treatment including the math see Optics Primer - Depth of Field

Regards

-- hide signature --

Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )