DOF calculation - marked f number or effective f number ?

AndrewC · Post by **AndrewC** » Thu Nov 13, 2008 1:27 pm

If you are working in the 1-4x magnification regime and calculating DOF using the widely distributed DOF = 2 N C (M + 1 / M x M) formula, do you use the marked f number of effective f number in the calculation ?

Andrew

rjlittlefield · Post by **rjlittlefield** » Thu Nov 13, 2008 1:39 pm

Marked.

If your lens is asymmetric, then there's an additional correction that depends on pupillary magnification factor. See http://en.wikipedia.org/wiki/Depth_of_field.

--Rik

AndrewC · Post by **AndrewC** » Thu Nov 13, 2008 1:46 pm

So the effective f number is just to get a handle on when you might start to see diffraction blurring ?

Andrew

rjlittlefield · Post by **rjlittlefield** » Thu Nov 13, 2008 2:46 pm

It's good for that. Effective f-number is also what determines the exposure time, should you ever be reduced to computing that instead of just measuring it.

Speaking of which...

In my experience (and I'm a math guy), the DOF formulas are good for general guidance and that's about it. By the time you factor in pupillary magnification and diffraction issues, it may be faster and it's certainly more reliable to just test your actual setup to see what works best for your application.

If you're stacking, for example, then you'll probably be looking for the smallest aperture that gives you adequate resolution. But almost by definition, diffraction will be significant at that point. The standard DOF formulas don't take diffraction into account, and as a result, they give a number that's too big. Choosing focus step based on the calculation will not give as good a result as good as one obtained by experiment.

--Rik

AndrewC · Post by **AndrewC** » Thu Nov 13, 2008 3:18 pm

In my experience (and I'm a math guy), the DOF formulas are good for general guidance and that's about it.

Ah, well I'm an engineer / lapsed plasma physicist. In our world 2+2=5 (just to be safe), well to be really safe let's make it 6. Actually, sod it, to be really sure let's just round it up to 10, oh and do a test run, or two, anyway

To be nice to my neighbours I don't do machine work at night (I need to mill a recess into a small drive housing so I can couple a miniature linear bearing to it - pictures will follow, someday soon) so sometimes I play with numbers

Like you say though, unless you have a really good proven model, sometimes they just point you in the general direction.

Andrew

rjlittlefield · Post by **rjlittlefield** » Thu Nov 13, 2008 4:01 pm

2+2=5 (just to be safe)...round it up to 10...and do a test run anyway

Ah, excellent -- a fellow of like mind!

--Rik

Charles Krebs · Post by **Charles Krebs** » Thu Nov 13, 2008 5:06 pm

... does this mean if my grand-kids think 2+2 =5 they are either going to be mathematicians, engineers, or plasma physicists?...

rjlittlefield · Post by **rjlittlefield** » Thu Nov 13, 2008 6:02 pm

Or accountants?

I've heard that one of the hallmarks of good accountants is that before they calculate the answer, they ask what you'd like it to be!

--Rik

DaveW · Post by **DaveW** » Fri Nov 14, 2008 5:15 am

Or con-men, maybe statisticians, even government ministers, or are those all the same?

DaveW

Harold Gough · Post by **Harold Gough** » Fri Nov 14, 2008 6:07 am

DaveW wrote:Or con-men, maybe statisticians, even government ministers, or are those all the same?

DaveW

I hear that they all had the same father!

Harold

Harold Gough · Post by **Harold Gough** » Fri Nov 14, 2008 6:17 am

The other thing to remember, if using flash illumination (and not metering), is that, in macro setups, the light source is usually very close to the subject and intensity at the subject varies proportionally to reciprocal of the distance, not to the reciprocal of its square. That is, halve the distance and you get double, not four times, the intensity.

Harold