About Diffraction

[Chad Shindel]

For anybody who can help a young lad...This question is about something I have never quite been able to grasp on my own. I understand how to properly select the optimal F-stop according to Mr. Luong, but I still don't know how to quantify the affects of diffraction. Here is an example. Let's say I photograph a scene at F22 and again at F64. I understand that the one shot at F64 is going to have a greater depth of focus that will be softer overall...even at the critical focus point, but is this going to be visible to the human eye at the original 8x10 size. If not, at what enlargement factor will the F64 shot be noticeably softer at the critical focus point than the F22 shot. Basically, what I am asking is if I shoot the same scene at F22 and F64, and I inspect just the critical focus point (not the depth of field limits), at what enlargement size will I notice a difference in sharpness when viewing the resultant prints at the normal minimal viewing distance of 25cm. By the way, I am shooting with an 8x10 and a Nikon 300mm 5.6 lens. Thank you all.

[John Brownlow]

The way to be really sure about this stuff is to learn how to do the calculations yourself. However, most DOF tables are based on a 10x magnification. You need to figure out what stop is the diffraction limit for your set up. Beyond this stop (say, f/32 for a particular lens on 4x5), stopping down actually reduces overall sharpness.

By far the best way of making judgements about this stuff is to shoot some tests.

[Ralph Barker]

There are a number of people who can give you better answers regarding diffraction limits than I, but let me complicate your question, Chad. Assuming you're talking about images made out in the field with your 8x10, there are several other practical factors that may be more important than diffraction. That would include things that move your camera (wind, and surface vibration from nearby roads, etc.) and things that move elements, like trees, within the subject (wind mostly). As such, there is benefit to considering all of these aspects to arrive at the optimal shutter speed and aperture based on the circumstances under which you're shooting each individual negative.

[Brian Vuillemenot]

With the shallow depth of field you get with most 8X10 lenses, you really need to stop down if you want the whole scene in focus. It is far better to stop down a bit too much and get everything in focus than to worry about diffraction, not stop down enough, and have parts of the scene be out of focus. Shooting at f22-45 will be the norm. Although most lf lenses have a "sweet spot" (where the lens is sharpest and diffraction is minimal) at f16-22, and stopping down past it will result in slight loss of sharpness due to diffraction, this will hardly ever be visible in real world situations at f32-64 unless you are making huge enlargements. Certainly at the original 8X10 size no difference will be visible to the unaided human eye. I have a shot of the Utah canyonlands made on 4X5 film with a 150 mm lens at f45 blown up to 20X24 and it looks extremely sharp to me (certainly as sharp as prints made from transparencies shot at f16-22). I wish I had stopped down all the way to f64, though, as a small area of the photograph is slightly out of focus!

[Kirk Fry]

It has been my exerience that there are many more important issues that lead to fuzzy results than small apertures. My major problem is not looking at all regions of the GG with my loupe. In short not focusing or getting the movements correct (same thing). I've trained myself to look at all four corners of the gg. Not locking down the movements causes the same effect. Next problem is camera movement due to wind or shutter or floor movement. Unless you are making bill board size prints and have solved the first two problems you are never going see diffraction effects. I did some tests once to see if I could see diffraction effects. I set up some USAF targets and shot them from F16 to F64 on TMX100 and developed in HC110 dil B for 7 min. I then looked at them with a 50X (combined objective and eyepiece) microscope. Sure enough you could see the slight fuzzyness. It is kind of like a very slight veil over everything. So compared to problems 1 and 2 above I would just use F64 when in doubt or you don't have to kill movement with short shutter speed or need some light on the film. Or maybe you only want one thing in focus. It is ART after all....

K

[bglick]

Chad, some good responses.... of course, testing is best, but sometimes it takes a little bit of simple math, to figure out how much film to waste :-)

Your question concerns only the "plane of sharp focus". I will address this, but you really also need to be concerned about the near and far limits at f64...but anyway, if you look at the diffraction tables, or just use the simple formual, 1500/fstop = max. lp/mm lens can deliver.

1500/22 = 68 lp/mm

1500/64 = 23 lp/mm

Now, keep in mind, this is only the "aperture diffraction" limit, but there are other forces at play here, such as gg / film alignment, shake, other forms of diffraction, MTF curve of the lens, (i.e. how it performs at that given f stop at your focus distance) type of film used, etc. So this does not become overly technical, lets assume you loose another 20% for the above factors, which brings your lens resolution to 54 and 19 lp/mm. Now, if you are using Provia 100F film at the tested contrast ratio, it will resolve up to approx. 55 lp/mm, so you run these two values through 1/R formula, which is the net effect of what the two can deliver together...

1/(1/r1 + 1/r2...)

Where r1 = lens max. lpmm delivered rez., r2 = max resolving capability of film at given contrast ratio.

If you had other variables in the optical train, you could add them also, such as filters = r3)

So, at f22 = 27 lp/mm at plane of sharp focus

at f64 = 14 lp/mm at plane of sharp focus.

Now, pick the lp/mm the human eye can resolve...a good eye can discern up to 5 lp/mm (not neccearily resolve) Average human resolution is closer to the 2 - 3 lp/mm. Using 5 lp/mm, you can enlarge the f64 shot 14/5 or about 3x before the difference will become obvious vs. 4x or 5x enlargement. And at f22 you can enlarge at 27/5 = 5.5x. This seems pretty decent, right? I would say these numbers are fairly accurate for you example, but in the photographic world, pretty accurate is +/- 50%. Just keep in mind, unless you're shooting a brick wall, or at infinity, it's the near / far subjects that can destroy the image, not at the plane of sharp focus. This assumes you can not use lens tilt to perfectly align the plane of sharp focus with the scene. Make sense?

[William Mortensen]

For 8x10 contact prints, I wouldn't worry about diffraction. I shot a 480mm Ronar at f/256 just to see what it did, and the 8x10 contact print looks quite sharp to the naked eye. Insufficient depth of field is very noticeable, and objectionable in certain veins of photography, (not so in others...)

[Capocheny]

Like Chad, I've also wondered about this very topic off and on over the last number of years. However, in practice, I've just gone with what Brian and Kirk have stated.

I shoot close-ups (indoors) using a 4x5 monorail and 240 f5.6 Nikkor and, usually, depth-of-field is quite a challenge. So, shooting at f32 or f45 isn't all that unusual! And, although I don't like to... I've even gone f64 on the rare occasion.

If I'm not enlarging to large sizes (16x20 or 20x24 being the largest I usually go)... I'd rather have the depth of field and use the smaller f-stops required to get as much of my subject matter in focus as possible.

.

wg,

Great explanation...

Any more applicable comments for folks using 4x5?

Cheers

.

[Bob Salomon]

" However, most DOF tables are based on a 10x magnification"

10x?

Linhof tables as well as the Rodenstock DOF Calculator/Scheimpflug Calculateor are for far less magnification.

Did you mean 1:10 perhaps?

[Chad Shindel]

Thank you all for your thoughtful answers. They have been helpful. I understand all the other factors in getting a sharp negative and I am not ignoring the importance of the depth of field limits, but my question's intent was to single out when sharpness loss due to diffraction becomes apparent at different aperture sizes. I used the critical focus point in my example because that would be the most telling reference point from which one could compare diffraction effects from various apertures at different enlargement sizes. I agree that doing my own tests would be beneficial, but I first wanted to understand the principle, and have a good idea of where to start since testing equals time and money. Thanks again to all of you and a special thanks to wg for his insightful response. Thanks.

Chad