Wanted: DoF table for Canon 20mm f/3.5 Macrophoto Lens

[3477jgu]

I'm thinking that the table will help determine step distances for focus stacking.

Jim

[enricosavazzi]

I'm thinking that the table will help determine step distances for focus stacking.

Jim

The only lens-specific parameter necessary to compute an accurate DoF table is the pupil ratio. If you already have this lens, you may be able to measure this parameter with some accuracy, and I believe enough information about this is available in these forums (e.g. see http://www.photomacrography.net/forum/v ... php?t=8895 ).

[3477jgu]

Thank you for the link. Building a table will be a challenge.

Jim

[rjlittlefield]

What camera do you intend to use, and at what magnifications?

--Rik

[3477jgu]

I'm using a Canon 5D Mark II along with a StackShot rail and a 3200˚K illuminator. My current interests are in mineral specimens and pond life. Switching from a Canon MP-E 65mm to the Canon 20mm Macrophoto I hope to see better image resolution from 4:1 to 8:1.

Jim

[rjlittlefield]

Switching from a Canon MP-E 65mm to the Canon 20mm Macrophoto I hope to see better image resolution from 4:1 to 8:1.

I understand the goal, but I suspect you won't reach it that way. Resolution in that range is mostly determined by the aperture of the lens. The MP-E used wide open at nominal f/2.8 is probably sharper than the 20mm Macrophoto at f/3.5. At 8X, even the MP-E has effective aperture of 2.8*(8+1) = f/25, while the 20mm f/3.5 is around effective f/32. Either way, you can see that diffraction is going to be cutting into your sharpness pretty badly.

If you want to make big improvements over the MP-E 65, you'll be needing to use microscope objectives. See our FAQ: How can I hook a microscope objective to my camera? for more discussion about the general technique.

At the upper end of your range (near 8:1) a good option would be a Nikon CFI Plan Achromat 10x NA 0.25 WD 10.5mm (part number MRL00102), pushed down slightly using a converging lens ("tube lens") shorter than 200 mm. See HERE for test report. That will still cover full frame with high quality. See Lenses for use at 4-5X on an APS-sized sensor for comparison between an MP-E and a pushed-down 10X objective. (But note that the 10X objective will not cover full frame if you push it down much below 8X. The corners will go fuzzy.)

The Nikon 10X is prone to producing purple casts in out of focus regions, due to longitudinal color aberration. If that sort of thing bothers you, then the alternative is Mitutoyo Plan Apo 10X NA 0.28 like HERE. The tradeoff there is that the Mitutoyo is not quite as sharp and does not cover as wide a field.

At the lower end (near 4:1), you might consider using a Mitutoyo Plan Apo 5X NA 0.14. It's not notably sharper than the MP-E 65 (effective f/18 at 5X, versus f/17 for the MP-E. But the Mitutoyo has the advantage that it's almost free of chromatic aberration, where the MP-E 65 is not. There are other objectives with even wider apertures, but they're generally hard to find and tend not to cover full frame well.

Anyway, back to the original question about DOF...

Can you handle Excel spreadsheets, or would you prefer to have your DOF table as just a bunch of numbers?

--Rik

[3477jgu]

As I read through the information that you have sent I understand most of the math involved in using the formulas to calculate the table. The measurement of the entrance and exit pupils seems rather subjective, but I suspect that there are self-canceling errors at work. I’m going to photograph the Canon 20mm and take some measurements in pixels to obtain the pupil ratio.

After thinking everything through, the question is: at a given magnification what is the largest step distance I can use without degrading image quality? Now I’m starting to think that I could arrive at this number empirically with a stage micrometer at an angle to the lens axis.

I have a Nikon CF E Plan Achromat LWD 10X, MSK00102. The images are not as sharp as I would expect. The objective is mounted on an RMS to M40 cone which mates to an EOS lens mount on a bellows. The objective is being used without a cover glass which may contribute to the softness.

Regarding the format for a DOF chart, I use a Mac which has a spreadsheet app that will accept Excel formatted data.

I really appreciate your expertise,

Jim

[rjlittlefield]

The measurement of the entrance and exit pupils seems rather subjective, but I suspect that there are self-canceling errors at work. I’m going to photograph the Canon 20mm and take some measurements in pixels to obtain the pupil ratio.

I don't understand the part about "subjective" and "self-canceling errors". The procedure described at http://www.photomacrography.net/forum/v ... 1516#61516 is pretty cut-and-dried to take the measurement.

After thinking everything through, the question is: at a given magnification what is the largest step distance I can use without degrading image quality? Now I’m starting to think that I could arrive at this number empirically with a stage micrometer at an angle to the lens axis.

My usual recommendation is to determine step size empirically by finding the largest step that does not produce visible focus banding. It's pretty hard to figure that out from a single shot of a tilted stage micrometer, because the pattern will just go from blurred to sharp to blurred, one time. It works better to shoot & process a stack, then look for multiple cycles of sharp/blurred/sharp/blurred. I also recommend doing this with a subject that is similar or more demanding than the subjects you actually care about. I don't know how well a stage micrometer qualifies there.

If your stacking rig will do very small steps, then one good method is to shoot a single stack with very small steps, then process it repeatedly with frame skip factor = 1, 2, 3, etc. The biggest step at which you don't see focus banding is just right. See the discussion and example HERE.

I have a Nikon CF E Plan Achromat LWD 10X, MSK00102. The images are not as sharp as I would expect. The objective is mounted on an RMS to M40 cone which mates to an EOS lens mount on a bellows. The objective is being used without a cover glass which may contribute to the softness.

I have no experience with that objective. But three thoughts come to mind:

1) Presence/absence of cover glass usually makes no visible difference with a 10X objective. The NA is too small for the cover glass to introduce enough aberration.

2) All objectives with the same NA generally have the same resolution at image center. Contrast and thus apparent sharpness can vary, especially if there's any stray light. But contrast can also be adjusted in post-processing. The big difference between objectives is how well they do away from center.

3. 10X NA 0.25 is effective f/20, so there's always a significant amount of diffraction blurring. Fresh out of the camera the images never look sharp. The best 10X shots you see posted by other people have always been sharpened pretty heavily.

If you'll post some 100% crops perhaps we can tell whether what you're seeing is typical.

--Rik

[3477jgu]

I set up and photographed the entrance and exit pupils on the Canon 20mm Macro using the “shallow DOF method” that you suggested and pixels as a unit of measurement .The results show a very small ratio of PR=1.01575. At this point I’m not sure if this number is correct in that it is not much greater than 1.

Regarding the Nikon 10X objective. I will find an image or two and post them for comment. Knowing that the cover glass is not a factor helps.

Arriving at a step distance empirically seems to be the best option. I have used large enough step distances to see rows (bands) of unfocused subject. My StackShot has a minimum step distance of 0.5µ, which should allow me to make a simple chart with aperture, magnification, field of view and step distance columns.

Jim

[rjlittlefield]

See also the spreadsheet introduced at http://photomacrography.net/forum/viewt ... 606#126606 and discussed more on page 2 of the same thread.

--Rik