PDA

View Full Version : Resolution/contrast decreasing at larger apertures?



dh003i
27-Jul-2011, 21:42
What do people mean when they say that a lens' performance degrades poorly at larger apertures like f45 to f64 on 4x5, or f90 to f128 on 8x10? After reaching their optimal point, all lenses will lose performance when stopped down due to diffraction.

So can it really be the case that some lenses are excellent performers at f/16 - f/22 (vs. other lenses as a % of the diffraction limit), but poor at higher f-stops? That would be using the light only from the central part of the lens then.

Seems counter-intuitive unless the aperture blades are awful (I've heard diffraction explained as caused by the percentage of light-interaction with the aperture blades...but I've also heard an information-theory argument that even if "God himself" took away light beyond the central part of the lens, it would still lose resolution...less information in = less information out).

Jack Dahlgren
27-Jul-2011, 22:41
What do people mean when they say that a lens' performance degrades poorly at larger apertures like f45 to f64 on 4x5, or f90 to f128 on 8x10? After reaching their optimal point, all lenses will lose performance when stopped down due to diffraction.

So can it really be the case that some lenses are excellent performers at f/16 - f/22 (vs. other lenses as a % of the diffraction limit), but poor at higher f-stops? That would be using the light only from the central part of the lens then.

Seems counter-intuitive unless the aperture blades are awful (I've heard diffraction explained as caused by the percentage of light-interaction with the aperture blades...but I've also heard an information-theory argument that even if "God himself" took away light beyond the central part of the lens, it would still lose resolution...less information in = less information out).

First, f/64 is a small aperture, not a large one. Note that the 64 is in the denominator.

The reason that diffraction is more of a problem at smaller apertures is geometry.
Diffraction exists where light encounters the edge of the diaphragm. As you make the opening larger, the circumference changes linearly with change in diameter, while the area changes as a square of the diameter. So as apertures get smaller, the amount of diffracted light is an increasing proportion of the image. At a certain size, almost all of the light passing through is diffracted.

Emmanuel BIGLER
28-Jul-2011, 03:21
To add a few figures to Jack's post, just remember that when diffraction effects actually limit the quality of an image, for example at f/64 for a lesn designed for the 4x5" format, the smallest spot size in the image of a tiny feature in the object cannot be smaller than about 64 microns on film.

And at f/N, N microns : f/128, 128 microns and so on.

Hence the bigger the image, the less annoying diffraction effects can be : big diffraction spots but a very big image nevertheless means a comfortable number of image points.
And note that at f/64 behind a lens, a properly focused optical image is still of much better quality than a pinhole image.
A pinhole image of same size, at f/640 (e.g. a 0.4 mm pinhole diameter for an image projected at 250 mm from the pinhole) contains about 100 times less images points than an image recorded behind a lens stopped down to f/64 !

dh003i
28-Jul-2011, 04:19
Thanks for the reply. I understand the laws of diffraction (and I was calling f/64 a larger aperture number, not a larger diameter).

What I am asking about is what people mean when reviewing or comparing lenses, and they say that the lens performance deteriorates at apertures smaller than f32. Of course resolution is going to be limited by the 1500/N approximation. However, as lenses are stopped down, they tend to perform better as a percentage of the diffraction limit.

What I'm asking is what people mean when they talk about lenses comparatively, as stopped down, saying one lens deteriorates worse than others. I mean, a theoretically perfect lens that performs at the diffraction limit at every f-stop would have the largest % decreases in performance as stopped down -- but that would be good. A lens that only gets 23 lppm at f/22 and f/32 probably wouldn't deteriorate in performance at all by f/64 (where 23 lppm is the diffraction limit), right? But that is hardly commendable.

I guess my question is, are people really saying that there are say 4x5 large-format lenses that don't approach the diffraction limit very closely when stopped down to f45 or f64?

Bart B
28-Jul-2011, 04:35
Here's a link to a web site discussing diffraction. Best part is a calculator to see how much diffraction a combination of camera properties, viewing distance and f/stop effects diffraction.

http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

rdenney
28-Jul-2011, 05:09
I guess my question is, are people really saying that there are say 4x5 large-format lenses that don't approach the diffraction limit very closely when stopped down to f45 or f64?

I don't think that's what people mean, or if it is, they are wrong. The only influence degrading lens performance at small apertures (and high f-stops) is diffraction, which is caused by what looks like light bending around the edge of the diaphragm. The smaller the hole, the greater the percentage of the image affected by those seemingly bent rays of light.

"Diffraction-limited" is a term used mostly by telescope makers and users to describe optics so good that optical aberrations have less of an effect than does diffraction. Of course, telescopes are generally used at their maximum (and only) aperture. When we stop down a camera lens, we reduce the effects of aberrations (i.e., spherical and chromatic). At some point, those effects diminish to the point where diffraction is what is limiting image resolution. Poor lenses may not become diffraction limited until a tiny aperture, or they may not be diffraction-limited at any aperture. This means that the aberrations are so large that it takes a large diffraction effect to be worse. Good lenses reach the point where diffraction has the bigger effect at large apertures when diffraction is still a tiny effect. Perfect lenses are limited by diffraction wide open.

There are lenses that don't intend to be sharp, but for those that do, I don't know of any remotely modern large-format lenses that never reach a diffraction limit. Most of my large-format lenses reach that point at about f/16 or f/22. This may not seem that good compared to small-format lenses, but in return for that these lenses can do things the small-format lenses don't do as well, particularly regarding field flatness, distortion, evenness, and coverage.

We all know that as one stops down, aberrations improve and the effects of diffraction grow. The optimal aperture (from the perspective of resolution in the focus plane) is the point where those two curves cross.

Of course, being out of focus causes more fuzziness than diffraction in many cases, so we stop down as far as we must to achieve the required depth of field. We use movements to minimize how far that needs to be for a given image, of course. Diffraction is even across the image and merely limits print size, while insufficient depth of field may affect parts of the image in ways we don't want and be noticeable at any print size. So, we strike the best balance we can between those competing objectives.

Rick "who doesn't know of many camera lenses that are diffraction limited within two or three stops of their maximum aperture" Denney

Brian Ellis
28-Jul-2011, 09:25
Layman's explanation (though I may not understand your question, in which case ignore this explanation):

Lens performance at different apertures is affected by lens aberrations as well as diffraction. The more you stop down the more lens aberrations tend to be improved. So the optimum aperture is one that strikes the best compromise between loss of image quality due to diffraction when stopping down and gains in image quality due to minimizing aberrations when stopping down.

Sirius Glass
28-Jul-2011, 11:28
Layman's explanation (though I may not understand your question, in which case ignore this explanation):

Lens performance at different apertures is affected by lens aberrations as well as diffraction. The more you stop down the more lens aberrations tend to be improved. So the optimum aperture is one that strikes the best compromise between loss of image quality due to diffraction when stopping down and gains in image quality due to minimizing aberrations when stopping down.

Well stated.
The area around the cross over point is referred to as the Sweet Spot.
The Sweet Spot for 35mm is generally around f/8 to f/11.
The Sweet Spot for MF is generally around f/11 to f/16.
The Sweet Spot for 4x5 is generally around f/16 to f/22.

I used the term 'generally' to avoid bickering of this lens maker verus that lens maker.

Also see Ralph Lambrecht's book Way Beyond Monochrome, second edition, for more researched and scientific discussion of this subject. http://www.waybeyondmonochrome.com/WBM2/Welcome.html

For the Hasselblad lenses that I have, the Sweet Spot starts at full aperture and extends to f/22. That is much wider than I stated above. My experience with 4x5 albeit small has consistently shown the Sweet Spot is for f/16 to f/22 on some lenses and to f/32 on the longer focal lengths.
My selection of 4x5 lenses is on the plebeian level rather than the level of many of the well established users here. My present goal is to learn to use the hand held 4x5s that I have and get familiar with the basic capabilities. Later, depending on use and interest I will buy better lenses and look to using them in a field camera. One step at a time, I have passed up opportunities to buy well priced monorail cameras because I fear that if I spread out too fast too soon, I will never use anything.

Steve