Enlargement limits due to diffraction in camera

[Jarin Blaschke]

Hi all:

I'm a new 8x10 shooter with only a 300mm lens so far. Being aware of diffraction limits, I've been trying to keep my photographs in the f/22-32 range. However there have been more than a few subjects that have absolutely needed f/45 and f/64. I employ movements and the scheimpglug principle as I can, but the images have many layers and directions of interest beyond a single diagonal plane.

In a little time, I will have the experience to answer the question for myself, but before enough results are printed at various sizes, I'm curious to hear people's opinions regarding enlargement limits from different camera apertures. For example, at what enlargement ratios can one perceive sharpness differences between a negative shot at f/22 and those shot at f/45, and f/64, presuming otherwise ideal circumstances and technique.

I use a 300mm apo sironar-s and fp4 film developed in Wimberleys metol-pyro formula.

Thanks!
Jarin

 

[Dan Fromm]

First rule of thumb: diffraction limits central resolution to ~ 1500/N where N is the f/ number. So at f/1 the limit is 1500 lp/mm, at f/32 it is 1500/32 ~= 50 lp/mm, and so on. Understand that diffraction limited resolution is at very low contrast. Usable resolution is less than the diffraction limit. The diffraction limit is lower off-axis.

Second rule of thumb: at normal viewing distance (by convention, 10") a print with resolution no less than 8 lp/mm will look sharp.

Third rule of thumb: resolution in the print = resolution in the negative/enlargement. Contact print, enlargement = 1, resolution in print = resolution in negative. This is why an 8x10 contact print from a negative shot at f/128 will look sharp. Enlargement = 4x, resolution in print = resolution in the negative/4. Getting 8 lp/mm in the print needs 32 lp/mm in the negative, limiting f/ number is then ~ f/45. But remember that this is in the center of the negative at very low contrast. To get better contrast and better resolution off-axis, shoot at a larger aperture.

 

[Jarin Blaschke]

Thanks for this- by 4x in 8x10, do you mean a 16x20 print or a 32x40 print? I presume the former...

J

 

[Dan Fromm]

With 8x10, 4 means 32 x 40. Linear dimensions, not area, are what matters.

 

[Maris]

Rather than doing the calculations from first principles every time I use "the 3 millimetre rule".

This says that if the final photograph size from an uncropped negative is 8" X 10" then the physical diameter of the lens aperture has to be 3mm or larger to keep diffraction effects just below the limits of visibility. For a 300mm normal lens on a 8x10 camera this means f100. For a 50mm lens on a 35mm camera this means f16. For other enlargement sizes the aperture diameter requirement changes proportionately: 6mm for a 16" X 20" photograph, 1.5mm for a 4" X 5" photograph, and so on ...

Sometimes lack of depth-of-field is a bigger sin than diffraction and I'll stop down further.

 

[RalphLambrecht]

Hi all:

I'm a new 8x10 shooter with only a 300mm lens so far. Being aware of diffraction limits, I've been trying to keep my photographs in the f/22-32 range. However there have been more than a few subjects that have absolutely needed f/45 and f/64. I employ movements and the scheimpglug principle as I can, but the images have many layers and directions of interest beyond a single diagonal plane.

In a little time, I will have the experience to answer the question for myself, but before enough results are printed at various sizes, I'm curious to hear people's opinions regarding enlargement limits from different camera apertures. For example, at what enlargement ratios can one perceive sharpness differences between a negative shot at f/22 and those shot at f/45, and f/64, presuming otherwise ideal circumstances and technique.

I use a 300mm apo sironar-s and fp4 film developed in Wimberleys metol-pyro formula.

Thanks!
Jarin

Dear Jarin, I have a whole chapter on this subject but it's too are for pug.email me at rwlambrec2GMAIL.COM AND I'LL SEND YOU A FREE COPY

 

[RalphLambrecht]

Rather than doing the calculations from first principles every time I use "the 3 millimetre rule".

This says that if the final photograph size from an uncropped negative is 8" X 10" then the physical diameter of the lens aperture has to be 3mm or larger to keep diffraction effects just below the limits of visibility. For a 300mm normal lens on a 8x10 camera this means f100. For a 50mm lens on a 35mm camera this means f16. For other enlargement sizes the aperture diameter requirement changes proportionately: 6mm for a 16" X 20" photograph, 1.5mm for a 4" X 5" photograph, and so on ...

Sometimes lack of depth-of-field is a bigger sin than diffraction and I'll stop down further.

interesting! Never heard of it but sounds reasonable!

 

[Sirius Glass]

Dear Jarin, I have a whole chapter on this subject but it's too are for pug.email me at rwlambrec2GMAIL.COM AND I'LL SEND YOU A FREE COPY

Buy his book "Way Beyond Monochrome". It will be the best thing you ever bought for yourself.

http://www.waybeyondmonochrome.com/WBM/Welcome.html

I have not financial gain from the book sales.

 

[Alan Gales]

Sometimes lack of depth-of-field is a bigger sin than diffraction and I'll stop down further.

I think you hit the nail on the head.

 

[ic-racer]

I can only go to 2x enlargements of 8x10 in my darkroom. But I like to look closely at my prints. I stay away from f64 in the camera. I think f64 is just for the contact printers. Having posted that, certain subject matter will be better at f64 when the focus spread is great. In almost all cases using the method of Hansma (to minimize both the effects of defocus and diffraction) yields good results. That is a long way of saying if you need very great depth of field, f64 or smaller gives better results than f32 or f22. The equation of Hansma gives you the mathematically best compromise between diffraction and defocus. It takes the guesswork and experimentation out of the question.

 

[ic-racer]

Also, I have the focus knob on my 8x10 camera calibrated to a chart based on Hansma's paper. For example if I have to twist my knob 360 degrees for the focal spread, I calculated that is f128. And so on, so 180 degrees of twist I'd use f64, etc. So, I don't need to carry the chart around.

And just to elaborate, if the focal spread is less then needed for f32, I still use f32 because if the imperfect physics of the optics. That is to say more open than that (with my lenses) and I have a hot spot in the middle and the edges will start to loose resolution. These things are not taken account in the Hansma formula.

 

[Sirius Glass]

Back to the OP's question. It depends on the type and quality of the camera lens and the enlarger lens.

 

[Jarin Blaschke]

Does anyone know the lp/mm of a diffraction-limited lens, at f/32, f/45 and f/64?

When I am next forced to go to f/45 and f/64, it would be good to know the largest print size that I can still deliver 8 or 10 lp/mm in the finished print...

J

 

[Dan Fromm]

Does anyone know the lp/mm of a diffraction-limited lens, at f/32, f/45 and f/64?

When I am next forced to go to f/45 and f/64, it would be good to know the largest print size that I can still deliver 8 or 10 lp/mm in the finished print...

J

See post #2 above.

 

[Jarin Blaschke]

Yes! Excuse me - that's what happens when I revisit a week later, after a long work day. Thanks again.

Does anyone know a source for MTF data on the app-Sironar S lenses (or at least my 300mm?). That way I could find the sweet spot stop where the diffraction limit meets the lens' resolution limit, in case a subject with a limited focus spread warrants a truly large print down the line...

Judging from my past 8x10 prints from 6x7 cm, it seems that 4x (5x max) is a good enlargement rule for invisible grain and perfect apparent sharpness. I'm pleased that I can nearly get to f/45 without sacrificing that level of quality in a 32x40" print. And if f/64 is a must, the scope of a 20x24 still ain't bad...


J

 

[zilch0md]

Does anyone know the lp/mm of a diffraction-limited lens, at f/32, f/45 and f/64?

When I am next forced to go to f/45 and f/64, it would be good to know the largest print size that I can still deliver 8 or 10 lp/mm in the finished print...

J

You might find this useful:

Dead Link Removed

Mike