What physical/optical attributes determine the max aperture of a lens?

[emanresu]

I have always been wondering, what physical/optical attributes of a lens determine its maximum aperture?  The actual diameter of the object piece?  Looked through google, but didn't find anything.  Maybe some of the experts here can give me more information.  Thanks!

[Bob Atkins]

In the simple case, it's the diameter of the front element. This is generally true for telephoto lenses.

so, for example, a 600mm f4 lens needs a front element that's at least 600/4 = 150mm in diameter, and most will be pretty close to that.

For more complex designs (especially wideangle and fisheye lenses), maximum aperture may be determined by internal stops.

So while a 14mm f2.8 lens requires at least a 14/2.8 = 6mm diameter front element, most such lenses will have a MUCH larger front element than that.

[klindup]

Building on what Bob said, are wide angle lenses still mainly reversed telephoto lenses?

Ken Lindup

[Bob Atkins]

Yes. All wideangle lenses for (D)SLRs are retrofocus designs because they have to be located further away from the film/sensor than their focal length. If they weren't retrofocus, they wouldn't be able to focus at infinity.

With ranglefinders you can move the lens very close to the film/sensor (no mirror in the way), so you can have a very short focal length (wideangle) lens that's not retrofocus (though that has some other problems, especially with digital sensors).

[emanresu]

Thanks Bob!  Am I correct to guess that it gets quite costly to have a front element with large diameters, and that is why fast lens are usually expensive?  (well there are other reasons for a lens to be expensive, but max aperture is definitely a deciding factor)

[Bob Atkins]

Large elements are more expensive, so yes, it's one reason why fast lenses are expensive (and why fast telephotos are very expensive!).

Another reason for the cost of fast lenses is that to keep aberrations low at wide apertures requires a lot of optical correction which is achieved through the use of more elements and more complex shaped elements (aspherics). More elements with more complex shapes also leads to higher cost.

[emanresu]

I just saw DxO software from Ken Rockwell's site, and it seems to be able to correct some of these issues.  Has anyone used it?

[Bob Atkins]

I've used DxO. It works quite well, though the version I used didn't have the most intuitive user inferface.

Canon's DPP can correct for distortion, chromatic aberration and vignetting as well as perfroming noise reduction, lighting optimization and giving you control over sharpness, contrast, color, wihte balance and a few other parameters. Best of all it's free and comes with every Canon DSLR.

[KeithB]

But:

DPP will not work on JPEGS, though I am not sure why...

It will not work for every Canon lens, for example my (darn it!) 70-300 EF is not on the list and not correctable, though the 70 - 300 DO *is* on the list.

[Bob Atkins]

That's correct. DPP is really only useful with RAW files (though you can make some JPEG corrections for color etc.). DxO is more flexible in that respect.

I think that DxO, like most software, is available in a limited term trial version, so if it interests you the best thing to do is to give a a try.

There's a shareware program called PTLens which may also be worth a look. It's available as a trail version, but the full licence is only $25. It's associated with the Panorama Tools suite of open source programs.

[emanresu]

I really need to get acquainted with DPP.  So far, I have only been using it for converting RAW into JPEG, with some WB, brightness/contrast and curve manipulation.  I definitely need to play with it more to learn its full capability.  Thanks for tip, Bob and Keith!