[Lowell Goudge]

Thanks for your comments; I understand them well.

I think you ascribe more quantitative goals to my effort than I intend. My intent is simple; it is to be able to say something like,

"Don't worry about diffraction when shooting macros until F-stop x (m+1) is greater than about 24, after that there'll be some sharpness loss."

I'm trying to get some illustrations that show what one might expect.

Dave

PS you say "and it will be lens specific..." Simple theory implies it won't be particularly sensitive to lens specifics for reasonable quality lenses. That's part of what I want to test and illustrate.

I guess where I am struggling is, I suspect that difraction may be much more of an issue as a function of focal length. Clearly a 200mm macro at F32 will have an 6.25 mm diameter aperture, but a 35mm at F32 will have a 1.1mm aperture. I would expect difraction will impact the 35mm lens sooner than the 200mm lens.

What I have not figured out is whether the depth of field as a function of magnification will be so different as to make a reasonable trade off with the shorter macro lens.

this is why I noted that difraction may be more of an issue as a function of lens, really I perhaps should have said focal lentgh.

As for something like giving a "simple rule like don't go beyond F22", might only be for a 35mm macro, but maybe it is F32 for 100mm etc... But I am sure you don't have macro lenses from 28mm through to 200 mm to take the shots you need to draw this relationship on a curve.

[View]

This is an interresting topic Had to do a google to try and follow this and found photo.net had a topic like this from 2005, Diffraction and DOF in macro?

Did a real life fratal search too and came up with Broccoli / Cauliflower Romanesco.

Not sure what the deal is but it seems to regard detail within the DOF and that it depends on the apature size. If it's in focus is it sharp or soft.

Haven't got my head round this but light bending at the apature blades will defract more over distance to the focal plane and thus a 35.2mm travel would have less defraction than one travelling 200mm - but then that assumes the angle be the same! But defraction may increase with apature diameters decreaseing from 6.25mm down to 1.1mm.

I get why you want fractals as your test subject newarts and wish you great luck in your testing. I'm now more aware of this subject as it binds together why there's a sweet spot with apatures and sharpness.

Thanks for the read and hopefully some later test findings

[Lowell Goudge]

This is an interresting topic Had to do a google to try and follow this and found photo.net had a topic like this from 2005, Diffraction and DOF in macro?

Did a real life fratal search too and came up with Broccoli / Cauliflower Romanesco.

Not sure what the deal is but it seems to regard detail within the DOF and that it depends on the apature size. If it's in focus is it sharp or soft.

Haven't got my head round this but light bending at the apature blades will defract more over distance to the focal plane and thus a 35.2mm travel would have less defraction than one travelling 200mm - but then that assumes the angle be the same! But defraction may increase with apature diameters decreaseing from 6.25mm down to 1.1mm.

I get why you want fractals as your test subject newarts and wish you great luck in your testing. I'm now more aware of this subject as it binds together why there's a sweet spot with apatures and sharpness.

Thanks for the read and hopefully some later test findings

My only comment here, and I admit I had not considered it fully, is the placement of the aperture in the lens. Depending on the lens type, construction, location of elements and groups, the aperture can be fairly close to the back of the lens or in the case of using an enlarger lens very close to the front element. As a result, each different design style, even within a single focal length could have greatly different impacts from defraction.

back to square 1 I think, go out with your gear and do a series of controlled shots, to look at both DOF and sharpness at the focal plane, with your lens and decide the best compromise.

[newarts]

I guess where I am struggling is, I suspect that difraction may be much more of an issue as a function of focal length. Clearly a 200mm macro at F32 will have an 6.25 mm diameter aperture, but a 35mm at F32 will have a 1.1mm aperture. I would expect difraction will impact the 35mm lens sooner than the 200mm lens.

What I have not figured out is whether the depth of field as a function of magnification will be so different as to make a reasonable trade off with the shorter macro lens.

this is why I noted that difraction may be more of an issue as a function of lens, really I perhaps should have said focal lentgh.
.....

Those are exactly the reasons why I'd like to do some un-ambiguous tests.

The theory is simple and well-founded. It is just Airy's Diffraction theory which is a completely accepted approximation. For a circular aperture it is:

Diffracton_spot_diameter=2.44*Distance_to_aperture*wavelength/Aperture_diameter

for macros,

Distance_to_aperture=Focal_length*(1+m)

So

Diffracton_spot_diameter=2.44*Focal_length*(1+m)*wavelength/Aperture_diameter

Using greenish light (0.55E-3mm), a Pentax sized sensor, a diffraction spot size equal to one thousandth of the display width adjusted to the Pentax sensor size, and the definition of F-Stop, the result is:

F-Stop*(m+1) less than ~ 18 to have a spot size for display less than one-thousandth of the display width.

Dave

PS there'll be disagreement about what limiting spot size to choose on the display, so I chose a simple, practical one. The effect of focal length is completely incorporated in the F-Stop term.

---

Last edited by newarts; 10-27-2010 at 01:17 PM.

[Lowell Goudge]

Those are exactly the reasons why I'd like to do some un-ambiguous tests.

The theory is simple and well-founded. It is just Airy's Diffraction theory which is a completely accepted approximation. For a circular aperture it is:

Diffracton_spot_diameter=2.44*Distance_to_aperture*wavelength/Aperture_diameter

for macros,

Distance_to_aperture=Focal_length*(1+m)

So

Diffracton_spot_diameter=2.44*Focal_length*(1+m)*wavelength/Aperture_diameter

Using greenish light (0.55E-3mm), a Pentax sized sensor, a diffraction spot size equal to one thousandth of the display width adjusted to the Pentax sensor size, and the definition of F-Stop, the result is:

F-Stop*(m+1) less than ~ 18 to have a spot size for display less than one-thousandth of the display width.

Dave

PS there'll be disagreement about what limiting spot size to choose on the display, so I chose a simple, practical one. The effect of focal length is completely incorporated in the F-Stop term.

That's fine for an elnarging lens on a bellows, but not for an internally focusing macro where the aperture is at the rear element. but any way, I think you would need to test each lens. and each focal length. On a long lens while the defraction will be there, and project to a larger defraction dot, the percentage of image is so small it will probably be seen as a loss of contrast not loss of sharpness.

[newarts]

That's fine for an elnarging lens on a bellows, but not for an internally focusing macro where the aperture is at the rear element. but any way, I think you would need to test each lens. and each focal length. On a long lens while the defraction will be there, and project to a larger defraction dot, the percentage of image is so small it will probably be seen as a loss of contrast not loss of sharpness.

My initial response is I think the rule may approximately hold for IF lenses, because even in that case the F(1+m) rule holds, and the f-stop is still F/aperture.

Your point about contrast loss for long tubes/bellows is a good one, but a separate issue.

Thanks for the thoughts - they are helpful,

Dave

[mikemike]

How about a dead cockroach?

Or maybe a seashell if you don't have the stomach to kill a bug.

[jolepp]

A vague idea: find/make a target in macroscopic size, shoot on b&w film, use (backlit) negative as target. Something as simple as designing a target on a computer and printing on paper might provide the macroscopic target (?).

[Lowell Goudge]

My initial response is I think the rule may approximately hold for IF lenses, because even in that case the F(1+m) rule holds, and the f-stop is still F/aperture.

Your point about contrast loss for long tubes/bellows is a good one, but a separate issue.

Thanks for the thoughts - they are helpful,

Dave

Dave, not so sure about contrast and long extension. Since the defraction is at large physical openings, very little of the total light from the lens, it will have the same impact as internal flair, and cause reduced contrast but the dominant image wil still have sharp edges. I thing the two will be connected.

now for IF lenses the question will be where the aperture is located. although the effective apature is still the relation you define, the physical size will be perhaps a function of location . the formula only really holds true if it is immediately behind the front element. therefore lens design does to some extent alter the amount of defraction impact on the lens. This will be for 2 reasons, first as I have always stated because the ratio of area impacted by defraction to area not impacted changes with aperture but also because the amount of defraction will change a s a function of aperture distance.

I think with everything in play, each lens design at each FL will be unique.

[Lowell Goudge]

A vague idea: find/make a target in macroscopic size, shoot on b&w film, use (backlit) negative as target. Something as simple as designing a target on a computer and printing on paper might provide the macroscopic target (?).

just so the lens you shoot it with is not suffering from defraction so the lines are sharp.

actually a pretty good idea.

[chalion]

I shoot macro at work and I first started out on junk blank circuit boards. I'm now at the point of shooting microprocessor chips & daughter boards, checking the soder points and also searching for micro cracks in the boards themselves.

It's a very boring, frustrating time consuming job.

[yeatzee]

Diffraction is a keen enemy of mine, and something I loathe very much.

Here's an example just to show its affects on images:


@ F/16, 1 image


@ Something like F/5.6, many images stacked

Both at the same magnification. Both have severe problems, but still it shows just how soft images can get

@ the op: I would love to hear your results!

[mikemike]

That being the case, I suspect you are going to be looking for a plate that has lines etched in by a laser.
When you find one, I'd be interested to know what it cost and whether or not it provided what you needed

Such calibrated scales are in the $500's-$5000's range!

Something that you might be able to find lying around for free or cheap with lots of tight lines very precisely arranged is an old motherboard or a CPU.

AMD Socket AM2 is 940 pins and is a few years old:

You can still get a brand new CPU for $25.
Newegg.com - AMD Sempron LE-1300 Sparta 2.3GHz 512KB L2 Cache Socket AM2 45W Single-Core Processor SDH1300IAA4DP - Processors - Desktops

Intel LGA 1156 processor:

These are more modern and you can't get one for less than $100 without ripping it out of a computer.

Here's a picture of all the little wires on a motherboard: