2 stops rule

[msvadi]

okay, I'm new to SLRs and I do spend to much time studing the MTF data from www.photodo.com . So, I wonder where the 2 stops rule is coming from. For the vast majority of lenses the MTF peaks at f/8 even for 1.8 and 2.8 lenses.

[hmhm]

The "two stops rule" isn't a rule, of course. Nor is it really even a rule of thumb, as it is almost never exactly true.

It is true that _most_ lenses are a little softer wide-open, and that softness (due to aberrations) clears up a bit as you stop down. Beyond that, lens aberrations are specific to a particular lens design, and some lenses (typically long, fast, telephoto primes) are very sharp wide-open.

I've found that some "reasonable practices" in photography have somehow been passed around by word of mouth long enough that they've unjustifiably acquired the title of "rule". I can imagine a "practice" whereby some photographers who don't walk around with MTF charts might try to avoid using a lens wide-open, and try to stop down a couple stops when they don't strictly need the largest aperture, as that's a reasonable compromise between sharpness and speed.
-harry

[ron chappel]

Can i ask how you can view the charts at photodo?I have no luck-just can't get them to show at all.It's been that way for ages,i think that last time i consistantly got them to show was about a year ago.
i have IE6

[msvadi]

I can see charts for some of the lenses, but I cannot see most of them. Anyway, they give the numeric data too, weighted MTFs.

[DaveG]

okay, I'm new to SLRs and I do spend to much time studing the MTF data from www.photodo.com . So, I wonder where the 2 stops rule is coming from. For the vast majority of lenses the MTF peaks at f/8 even for 1.8 and 2.8 lenses.

I always use it as a rule of thumb. And I don't say two stops I say middle apertures. Wide open lenses are soft on the edges and especially fast ones. Stopped down too much introduces diffraction. So middle apertures (whatever they are) are the best technical solution.

Now in the real world almost any aperture is OK. I've never actually seen a photograph and said, "Gosh look at the diffraction." I mean I can't see it at f22, or f128 on my 8x10's 300mm lens. Although I can see lack of depth of field with my 16-35 f2.8 AT 2.8 I don't see a lack of sharpness that's lens induced (operator induced maybe :D )

So understand that middle apertures are better, and then forget about MTF charts and actually go and press some shutter buttons.

[robertwgross]

I've never actually seen a photograph and said, "Gosh look at the diffraction."

I've seen lots of problems with digital photos, but I'm not sure that I know what diffraction would look like. Some of my landscapes are with aperture of around f/27. What should I look for?

---Bob Gross---

[DaveG]

I've never actually seen a photograph and said, "Gosh look at the diffraction."

I've seen lots of problems with digital photos, but I'm not sure that I know what diffraction would look like. Some of my landscapes are with aperture of around f/27. What should I look for?

---Bob Gross---

Diffraction is an optical defect caused when the aperture is too small, and is caused when the light bends coming around the edge of the very small diaphram opening. The aperture where diffraction becomes a problem is dependent on the size of the "negative" which why large format lenses routinely stop down to f64, while f22 is as far as you can go with 35 mm, although my 55 Nikkor macro stopped down to f32. Heck my 300 mm Nikkor for the 8x10 STARTS at f9, so f128 isn't all that weird. In any case the manufacturers limit the defraction problem by just not making lenses that stop down far enough for it to be a noticible.

The effect is just that the image is soft. But the softness is relative and will be almost impossible to see unless you had a series of shots at f22, f16 and f11; and of course where the greater depth of field of f22 wouldn't be important. Then, in comparison, you mght see something. But as I say I don't look at images shot at f22 and wish that they had been shot at f11. But if I don't NEED the depth of field of f22 then I'll pick something wider, on the basis of why borrow trouble.

[robertwgross]

I understand basic optical diffraction from physics class, but I've never seen an image at f/27 or so that seemed to suffer from overall softness that could be attributed to the aperture.

I like to shoot landscapes where everything is in focus from about four feet out to infinity, and about the only way to get there is by getting around f/22, f/27, etc.

At the next time I arrive in my magic landscape spot (at about 7500 feet on the rim of Yosemite Valley), I'll have to try different apertures to see if I can find any difference.

---Bob Gross---

[DaveG]

I understand basic optical diffraction from physics class, but I've never seen an image at f/27 or so that seemed to suffer from overall softness that could be attributed to the aperture.

I like to shoot landscapes where everything is in focus from about four feet out to infinity, and about the only way to get there is by getting around f/22, f/27, etc.

At the next time I arrive in my magic landscape spot (at about 7500 feet on the rim of Yosemite Valley), I'll have to try different apertures to see if I can find any difference.

---Bob Gross---

What format are you shooting?

[robertwgross]

What format are you shooting?

Normally I shoot my D60 until I see something good showing up, and then I pull out a Canon 35mm film camera and shoot Velvia 50.

Since the 1.6 factor is in the digital body, I can move zoom lenses over to the film body, adjust a bit, and get almost the same shot on film.

---Bob Gross---

[hmhm]

I always use it as a rule of thumb. And I don't say two stops I say middle apertures.

An "f11 for best sharpness" rule of thumb (or f/8) is pretty reasonable, as it's almost always "true enough", or at least close enough to true that the difference shouldn't matter much. This assumes, of course, that you're not dropping into slow shutter speeds as a result, i.e. you've got a sturdy tripod that's not swaying in the wind, mirror lockup, etc.

I've never actually seen a photograph and said, "Gosh look at the diffraction."


The effects of diffraction are fairly subtle, it's not a "hit you over the head" effect. I did a simple test here:
http://forums.dpreview.com/forums/read.asp?forum=1019&message=4915567
-harry

[Jesper]

I've never actually seen a photograph and said, "Gosh look at the diffraction."

I've seen lots of problems with digital photos, but I'm not sure that I know what diffraction would look like. Some of my landscapes are with aperture of around f/27. What should I look for?

---Bob Gross---
I don't know exactly how it works, but I don't think diffraction is a certain kind of distortion that you can't recognise immediately in an image. It makes images less sharp, but doesn't have a characteristic look of its own.

I did a lens test sometime ago - put a piece of paper with printed text on a wall, put camera on a tripod, use mirror lock-up and the self timer and make test shots with a lens at different apertures. I noticed that my 17-40L was sharpest at f/5.6 and f/8. At larger and smaller apertures, the image was less sharp than at f/5.6 and f/8. Probably the unsharpness at larger apertures was caused by diffraction.

[msvadi]

Can i ask how you can view the charts at photodo?I have no luck-just can't get them to show at all.It's been that way for ages,i think that last time i consistantly got them to show was about a year ago.
i have IE6

I just checked it again, I can see the charts for 70-200 4L, for example, but not much else.