View Full Version : Lens theory; sharpness vs. aperture

Mark Sawyer
11-Sep-2004, 16:23
Just a theoretical question...

You have two 12" lenses for your 8x10, both ground with equal precision with identical coatings, etc. One has a maximum aperture of f/4.5, the other a maximum aperture of f/9. You make identical negatives with each lens at f/64. Any difference in sharpness?

Bruce Watson
11-Sep-2004, 16:26
No. All other things being equal, the same aperture gives the same result.

Glenn Kroeger
11-Sep-2004, 17:36
Moreover, chances are at f/64 both will produce similar results even if they aren't equal at wider apertures. At f/64, both are likely to be limited by the physics of the aperture than the optics (assuming one isn't a total dog).

Dan Fromm
11-Sep-2004, 17:57
The two lenses' images will be equally fuzzy at f/64, but the f/4.5 lens' plane of least fuzziness may be nearer where you want it. If, that is, it doesn't suffer from focus shift on stopping down. This because there's less DOF at f/4.5 than at f/9, hence more accurate focusing through the lens at f/4.5.

If this is a question of, say, a 210/9 G-Claron vs. a 210/4.5 Industar-51, I suspect the G-Claron will be sharper at all apertures from f/9 down. But I say that because I don't like my I-51, not because I've done the experiment.



Richard Boulware
11-Sep-2004, 18:32
Speaking from experience of over 40 years as a professional photographer, I have learned that any optical engineer will tell you than any given lens will reach its optimim performance, two stops below it's maximum aperature.

Having tested dozens of lenses, and graphing them out using a USAF lens test chart and a 30X binocular microscope, I can confirm that, for the most part, this is true.

Lenses however, are like people. There are winners and loosers. Each lens has it's own personality. Some are lifelong friends, and some are jerks. Test your lenses, and decide for yourself. Best of luck.

John Cook
11-Sep-2004, 18:37
In the real world there exist many practical variables. In many cases, for example, a slower lens also has a smaller image circle which limits shifts and tilts. But the smaller image circle also projects less light onto the camera bellows. Since a portion of this light is always reflected back onto the film, it follows that a slower lens (at any f stop) sometimes creates less internal flare and fog than a faster lens.

Jim Galli
11-Sep-2004, 21:19
Both would be diffraction limited to around 25 line pairs.

Mark Sawyer
12-Sep-2004, 01:15
An added point from a friend offline: he offered the opinion that as larger apertures are more difficult to grind precisely, they may in general have lesser quality. Any thoughts?

12-Sep-2004, 10:51
May I take exception?

"Speaking from experience of over 40 years as a professional photographer, I have learned that any optical engineer will tell you than any given lens will reach its optimim performance, two stops below it's maximum aperature. "

I would submit that there are too many variables to make sweeping generalizations. I would use the tested results from an extraordinary site: www.naturfotograf.com

It depends on the lens. Super speed glass often often is at best wide open or a stop less. Witness 200 f2 Nikkor, 300 f2 Nikkor, various f1.2 normals etc. See: www.naturfotograf.com/lens_short.html

Alternatively peak performance from LF optics varies quite a bit... witness this page; www.naturfotograf.com/lens_LF.html

Then again, a truly rotten design like a magnifying glass may be at it's mediochre best stopped way - WAY - down to choke the life out of the worst optical abberations.

Since I am using someone else's site, experience and text to bolster my contention, I would submit this page as evidence this guy is technically capable: www.naturfotograf.com/uvfilms.html


12-Sep-2004, 21:39
The problem wih the question is in the assumption that all else would be equal it the lenses were ground with equal precision, had the same coatings, etc. etc.. In fact, if two lenses of the same focal length have different maximum apertures, it's because they are of fundamentally different designs. In other words, little else is equal. So it's impossible to make a meaningful generalization.

Except for the one that's been made here already, which I doubt specifically answers your question: all lenses are practically equally crappy at f64 because of difraction.

On the comment that all lenses perform best when stopped down two stops, I have to disagree with this generalization. My experience (which includes reviewing a stupid number of MTF curves back at a lonelier time in my life, and a lot of photography) indicates that this is true for just about every lens ON AXIS (meaning, in the center of its field of view). However, the corners are a different story. Many lenses, in fact all LF lenses that I've looked at, achieve better overall corner to corner performance (at the slight expense of their on axis performance) when stopped down one to two stops farther. This is even more the case if you're going to use movements.

This was most easily tested with my enlarging lens (a 150 apo componon hm). Center sharpness was greatest at f8 (two stops down). But corners were soft. At f11, the corners were razor sharp, and the center, while softer than at f8, was still very good. I always use it at f11. When Salt Hill had Schneider make a custom fixed aperture version of this lens for them, Schneider made it at f9 (a slightly wider compromise). Less scientific use of camera lenses in the real world found their overall performance corner to corner to be best at 3 to 4 stops down, depending on the lens. Probably the 4 stop figure comes from much of my work including buildings, which often require shifted lenses, and which put more than sky in the upper corners.