Sharpness and DOF question.

[TCampbell]

The higher the f-stop, the greater the depth of field (there are other factors which affect depth of field besides f-stop). However... as you increase the depth of field by increasing the f-stop (decreasing the aperture diameter), you also increase diffraction. This is true of all f-stops, but it may not be noticeable. At some point, the camera becomes "diffraction limited".

This happens because light doesn't focus to a single point. Due to the wave nature of light, it focuses to a spot called the "Airy Disk" (discovered by astronomer George Airy in the 1800s). As the size of the aperture gets smaller and focal ratio increases, the size of the Airy disk increases. At low f-stops (large aperture sizes) you still have an Airy disk, but the diameter of the disk is small enough that it easily fits within the space of one "pixel" on the sensor (I'm simplifying because sensors usually have photo-sites and a combination of photo-sites make up 1 pixel.) At high f-stops (smaller aperture sizes) the Airy disk is bigger... so big that it may not fit within one pixel and will spill over onto others -- affecting their color. This has the effect of decreasing the contrast among adjacent pixels. It "softens" the image.

It's not as if the camera is fine at one f-stop and then disastrously awful at the next... it's a gradual process. The physical micron size of the photo-sites has a lot to do with this.

Is your camera diffraction limited at f/22 (image quality reduced)? The answer is yes... absolutely it is. No question. It's not a matter of camera, lens, or sensor quality... this is a physics problem and the math behind it is extremely well-understood and the physics assume the optics are "perfect".

But that's not the real question. The REAL QUESTION is... will you notice?

It turns out, when you look at an image that you took, you're not using ALL of your data.

My monitor has a resolution of 2560 x 1440 -- which is pretty high as monitors go. But that's only 3.8 megapixels. If I choose to print an 8 x 10 at 300dpi (a common print resolution) then that's only 7.2 megapixels. What's the resolution of your camera? What's the resolution of your monitor? How large are you planning to view or print? It turns out... you can afford to be fairly diffraction limited and not notice. If you want to print an 8x10 at 600dpi then you'd use 28.8 megapixels but you'd have to squint hard to see those dots at 600dpi... the human eye usually can't notice that.

If you plant to print at, say, 20 x 30" at 300dpi... NOW you're up to 54 megapixels... and now if you shoot at f/22 (and look at the print fairly close instead of from a distance) you can notice that the image is diffraction limited... but how often do you do that?

BTW - diffraction has nothing to do with chromatic aberration (aka color "fringing"). That's a completely different problem and has to do with the prismatic effect of glass. Low dispersion glass and and the complex use of different lens element configurations are designed to combat the problem by minimizing it.... but you can get fringing at any f-stop and it should be worse at low f-stops.

 

[shaylou]

I never understand threads like this (not attacking the OP but this type of thread in general).....you have a camera and lens. If f22 doesn't work for you simply change it and try again. If this still doesn't work keep changing and reshooting until you get the effect you want. The time wasted asking the question could have been spent experimenting with the different settings on the camera to get the desired results.

With all do respect the op was basically given bad advice. The fstop is relevant to the lens and the scene is the deciding factor of dof. Mr Petersen knows his stuff but sometimes (like all of us) gets ahead of himself. There is no hard rule for this. Hyper-focus is probably the best answer as already stated. So this is not a waist of time question, it is a time saving question imo. Get a fast clarification from us and go out and practice other things. At least this method works for me.

 

[shaylou]

Focus stacking for sharpness on landscapes?

 

[shaylou]

If you listen to him long enough you will hear him say all kinds of odd things that do not add up. He is great but like I said, he gets ahead of himself from time to time.

 

[batmura]

This video by Bryan Peterson finds him emphasizing f/22 and responding to the naysayers:

Leading Line Ep 324: You Keep Shooting with Bryan Peterson: Adorama Photography TV - YouTube

 

[lenny_eiger]

This depends a lot on what camera and lens you use. Digital cameras are far more sensitive to diffraction. Smaller formats as well, to some extent. With larger cameras and modern lenses diffraction is minimal enough to be ignored. (The two stops down from wide open is ludicrous.) I shoot at f45 all the time. I wouldn't consider f22, let alone f13. I have done the tests to prove it, up at 40 inches of print size.

I am also not alone in my opinion, this is the consensus over at the large format forum...

The important thing is to know your lenses. When you are out shooting, do quick test, change the shutter speed and fstop and back in the darkroom see what happens....

Lenny
EigerStudios