Depth of Field and "Focus by Wire" - a Thought

[GrahamS]

Tony, the NPP is a point usually found near the entrance pupil or front element of a complex lens around which the camera can be rotated without parallax becoming evident. The nodal point is a point within a complex lens designated by the designer as the nominal nodal point. This point will vary slightly as the focal length of the zoom system is changed and is regarded as the optical centre of the lens system.

This may help you:

http://spie.org/x32307.xml

[xpatUSA]

When researching Depth of Field (DOF) there is plenty of tutorial literature, here for example, on the web and a good few calculators. And also confusion galore, including the circle thereof.

The one thing that strikes me is the profusion of methods and examples on how to calculate the DOF, given apertures, distances, sensor detail and the various diameters of the "circle of confusion" (CoC), certainly varying between camera models, also between manufacturers of cameras with same size sensor and sometimes different values quoted for the same camera model.

The other thing that strikes me is the absence of methods that start with a required DOF and end up with a camera setting. I mean, if I shoot a close-up, I know what DOF is desirable. Calculating DOF and re-calculating it ad nauseam while entering apertures, sensor size, distance to subject, blah di blah, is a waste of time and quite inelegant. If I want to shoot something, I already know the DOF I need. I know the aperture I need. I know (from research) the (CoC).

For close-up photography, the calculation or graph should be for the magnification (m) that would provide that known DOF. Therefore, I have taken the standard close-up formula for DOF = 2Nc(1+m)/m^2 and rearranged it to get m from DOF, f-number and c. Wasn't easy - it involved finding the roots of a quadratic and picking the proper root.

It works. I've tried and tested it, and can now think in terms of magnification without worrying about lens pupils, distance to subject, effective aperture, the list goes on. Suppose my formula tells me I need an m of 0.1 and I'm shooting a watch of say 50mm tall from lug to lug. 50mm x 0.1 says that it's image at my sensor should be 5mm tall. That tells me to frame the watch for about 5/14 of my sensor height because my Sigma camera sensor is about 14mm tall.

No distance measurement required - the very act of framing does that for you!

Next step is to translate my formula into JavaScript and post it where I can use my smartphone to do the calc, deep in the heart of Texas!

For the mathematically inclined:

DOF/2Nc = a

am^2-m-1 = 0

m = (1+sqrt(1+4a))/2a

[jcuknz]

[QUOTE=xpatUSA;334151, The second shot was hand-held with OIS on, the first on a tripod with OIS off. The difference in DOF is quite evident, I reckon![/QUOTE]

I would suggest that if you organised the first shot with similar tone to the second you will find the improvement in DoF is less than spectacular.... there is a bit more BUT

[xpatUSA]

I would suggest that if you organised the first shot with similar tone to the second you will find the improvement in DoF is less than spectacular.... there is a bit more BUT

Yes, the shots should have been done with identical lighting. Really should compare apples with apples . .

. . but all I would do is take a shot, change the shooting distance, take another shot and "discover" that, moving back, the DOF increases almost as 1/m^2 (where m = image height / scene height, or an objects height in the image / its actual height).

For example, micro four-thirds 13mm height, CoC 0.015mm, set f/11:

Scene ht. 60mm: DOF = 8.6mm

by moving back a bit, more of the scene is imaged . . .

Scene ht. 100mm: DOF = 22.1mm

Less than spectacular but, quite evidently, 2-1/2 times as much

[jcuknz]

Maybe I have a problem but I rarely can see the difference in DoF with pictures as opposed to for instance looking along a ruler.
I couldn't see it with Grahame[Fiji] nor yours, only with my rulers. It is what it looks like or a properly set up test rather than what the calculations say for me.
Whatever I have changed my opinions thanks to Grahame and will maybe modify my approach to wee things ... pity I have bought a 4 dioptre CU lens recently becuase 280 AoV and a two dioptre doesn't give me the tight framing I had with a 430 AoV did ...I can use the same Cu Lens on both cameras.