Focal length, distance, and depth-of-field: clarification?

Drew, while I am all for hands-on experience (Try it!, as Fred Picker would exclaim), and I practice with such matters as often as able (I just read about a focusing technique on this forum and was out practicing in hours); and while I would be extremely unlikely to use a 135 at 3' to make a portrait in attempting to get more DOF than my 210 would allow at 5', I was inquiring here about operant principle explaining the difference between what has been stated by some accomplished authors and what the calculator showed. Really, that's all.

If we understand a principle, we can apply it, e.g., the three variables determining DOF. My question here was a derivative, which it seems to me I managed to explain reasonably clearly in my OP.

Quote:

---

Originally Posted by Ulophot

Another thread prompted this query, whic for me is just a matter of curiosity. The possibility of substituting a shorter focal length lens and moving closer was suggested as a way to increase DOF... "

---

I think the advice was to use a shorter focal length at the same distance (keeping same perspective), then crop back down to the original image to increase DoF.

Choice of lens and subject distance should be dictated by what image you are trying to make. With that said, to first order:

- when you change the subject distance to get the same horizontal field of view, the distance is proportional to the focal length.

- when you keep the f-stop constant, the diameter of the entrance aperture is proportional to the focal length. (That is, for a 210mm lens at f/16, the aperture is 210/16 = 13 mm diameter. for a 135 at f/16, the aperture is 8.4 mm diameter.)

The blur of an object ahead/behind of the plane of focus is related to the subject distance and the entrance aperture. This is most easily understood for a point behind the plane of focus. The image forming light from this point is a cone starting at the point and entering the aperture of the the lens. When this cone intersects the plane of focus, the intersection is a small circle, so the point is imaged as a small circle on the film rather than a point. The size of the blur circle is proportional to the aperture diameter divided by subject distance.

So to a good approximation, if you change the focal length and subject distance together, and hold the aperture constant, the DOF stays the same, but the relation between objects in the picture changes.

reddesrt, thank you for this explanation, which is helpful. Your "to a good approximation" is pretty much what Simmons wrote.

I guess I'll let the discrepancy I posted in the two examples rest; probably the explanation would get too complicated, although, based on your explanation, I'm not sure why it would. I'll review from a practical standpoint -- read, play around with it in practice -- when I get the chance, just for curiosity's sake.

Quote:

---

Originally Posted by Ulophot

reddesrt, thank you for this explanation, which is helpful. Your "to a good approximation" is pretty much what Simmons wrote.

I guess I'll let the discrepancy I posted in the two examples rest; probably the explanation would get too complicated, although, based on your explanation, I'm not sure why it would. I'll review from a practical standpoint -- read, play around with it in practice -- when I get the chance, just for curiosity's sake.

---

There's no real discrepancy in the two examples you posted in the first post. There is a mistake though. Go back to the online calculator and try the numbers for the 135mm again. You have:

Quote:

---

135

HFOV 9'
Distance 10'
TDO 5'11.8"

---

It should be a distance of 9.6 feet for a horiz field of view of 9 feet, bringing the TDOF down to 5'5.6".

Anyway, the depth of field from the two lenses then comes out within a few percent of each other at constant field of view and f-stop. If you want to worry about it in greater detail (second order effects), you have to consider the effects of extending the lens out to focus and so on - see the notes on that DoF calculator page. It doesn't matter in practice because DoF is a gradual effect, not a cliff that you fall off of, so if the DoF is a few percent different between two cases, you would not notice in practice.

Thank you for pointing out the error.!I have corrected it, tfough I had chosen the 2nd method; I believe you chose the first. My correction for Example 1 (edited there now) is

210 mm

Distance 15.6'
HFOV 9' (stated as 8'12'', probably due to rounding)
TDOF 5'8.9"

The discrepancy, now down to ~2" is certainly small, and yes, DOF is not a cliff.

Thanks again.

If you search back through threads I started on Photrio, you’ll find one where I posted exact calculations - no approximation - for specifically these relationships.

Here it is: https://www.photrio.com/forum/thread...ling-f.169727/

-Jason

Philip,

I must apologize for sending you down this rabbit hole to begin with. It was likely my less-than-clear post on another thread that started this whole questioning.

To clarify, I did suggest using a shorter focal-length lens to get more depth of field. Ideally, you keep the same camera position and crop back to your originally-planned borders. However, if you end up cropping extensively, you run into problems with grain, etc.

Now, if you use a shorter focal-length lens and move closer so that an important object in the scene is the same size on the ground glass as it was at a greater distance with a longer focal-length lens and keep the same general framing, you get no advantage with depth of field; it will be about the same as before, since the magnification is the same.

However, if you split the difference, so to say, i.e., move in a little and use a shorter focal length lens, you'll gain a bit more depth-of-field and avoid having to crop so much to get your original general framing.

Keep in mind that any time you move the camera position, you're going to change the relationship of objects in scene, i.e., the perspective. Therefore, when you move that bit closer, you'll be recomposing your photograph, which is also a compromise and maybe not acceptable.

Hope that clarifies things,

Doremus

Thanks, Jason, I have a look.

Philip U:

Thank you for that. In retrospect, I did not carefully read your OP.