[stevebrot]

So, may I assume then that the K 200/"2.5" and the Takumar [Bayonet] 135/"2.5" specs were "designed" in the Pentax Marketing Department?

I have a couple of f/2.5 lenses. Whether they are really 1/3 stop faster than their f/2.8 cousins is a matter of speculation. The same might be said of f/1.8 vs. f/2.0 or f/3.5 vs. f/4 (also 1/3 stop difference).


Steve

---

Last edited by stevebrot; 11-10-2015 at 06:01 PM.

[stevebrot]

The key for smaller long telephoto lenses seems to be in the use of a Fresnel element. Canon and Nikon use them, giving them different names. But they're both Fresnel concepts. I don't think Ricoh would make enough lenses to justify the cost of developing these elements, as they seem to be quite pricey to make.

Fascinating!

Here is the full explanation at Nikon:

Nikon | Imaging Products | NIKKOR Lens Glossary

There is a penalty in the form of strange flare under some circumstances, but as with most good things there is usually a piper to be paid somewhere.


Steve

[Digitalis]

From my experience with diffractive optics, is that the images taken with lenses using DO often require a bit more sharpening than standard lenses do, there are also Bokeh artifacts caused by the DO elements itself - unless a diffractive corrective element is used - which has to be the geometric opposite of the primary DO lens, this make the optical construction considerably more complex, and prone to loss of optical transmission.

as with most good things there is usually a piper to be paid somewhere.

what happens is the piper was terrible to begin with?

[bdery]

My A* 300/4 is pretty compact although the front element is pretty big with a 77mm filter size. It's a FF lens so it would think it could be made smaller as a DA.
The real question for Ricoh is would people want to buy it?

It's only my guess, but I believe short-term development will focus on FF lenses. Pentax has the most complete APS-C on the market, but with the upcoming FF THAT lineup should be built.

The key for smaller long telephoto lenses seems to be in the use of a Fresnel element.

Fersnel optics will make lenses shorter, but not smaller (in diameter). And as others have said, it's not an easy thing to do properly.

[mattb123]

It's only my guess, but I believe short-term development will focus on FF lenses. Pentax has the most complete APS-C on the market, but with the upcoming FF THAT lineup should be built.



Fersnel optics will make lenses shorter, but not smaller (in diameter). And as others have said, it's not an easy thing to do properly.

I'm sure you are correct! The OP's dream is just that.

[clawhammer]

A couple points I want to make here.

Zerost (Zerst? Zeroist? 0*ist?): Guys, optics are confusing. This post involve way more rabbit holes than I thought it would. Still not sure I've got it all right.

First, almost all stated f/ numbers are approximations in one way or another. For example, this f/4.5 lens is really f/4.6, and this f/2.8 lens is really f/2.9-3.0.

Second, while we like to refer to f/stop being defined by a physical property (the aperture opening), it's a little more nuanced than that. It is defined by the focal length divided by the entrance pupil (source; I know it's wikipedia, but the source is an optics textbook). This pupil, as bdery said, is a projected opening, not a physical opening. If there were no lenses bending the light between the thing you were taking a picture of and the aperture assembly (like a pinhole camera), they would be the same (source).

This MIT lecture gives a slightly different definition for f/# calculation; this OSU presentation shows the math on how they are equivalent. Also note that the field stop, which defines the field of view of the lens, is normally placed within the optics near the rear of the lens (but not always). My guess is that telephoto and retrophoto designs make what sounds simple—the location of the various stops and pupils—and makes it totally counterintuitive. Because complex engineering always does that.

I'd love to hear from a real optical engineer to see if I've got this all correct or if I'm way off course.

And finally, some lens design humor.

Further reading: photozone's series on lens design

Edit: The calculations for the nerd-level of the author of this post is left as an exercise for the reader.

[bdery]

I'd love to hear from a real optical engineer to see if I've got this all correct or if I'm way off course.

Well, optical designer is my job title, and I've got a Master's and Ph. D, in optics, so I guess I quality

You summarized things pretty neatly. Kudos.

[Lowell Goudge]

So, is it possible then, for the "projection of the aperture stop at the front of the lens" to be larger than the actual physical clear diameter of the front element?

I am specifically thinking of certain lenses that seem to be "overly generous" in their rated f/stop. For examples:

The Takumar [Bayonet] 135/2.5 (unlike the K 135/2.5, which has a filter thread of 58mm) has a filter thread of only 52mm, so that its actual front element diameter can't be more than, say, 50mm for an absolute maximum (and maybe only 48mm or so). But, if I divide 135mm by 50mm, I would get an aperture value of f/2.7, which is certainly numerically larger than "f/2.5". Or, another way to look at it would be to divide 135 by 2.5, which would suggest an aperture of 54mm (difficult to accomplish, I have thought, with a 52mm filter thread).

The K 200/2.5 has a filter thread of 77mm, so that its actual front element diameter can't be more than, say, 75mm as an absolute maximum (and perhaps only 73mm or 74mm). But, if I divide 200mm by 75mm, I would get an aperture value of f/2.67, which is certainly larger numerically than "f/2.5". Or, another way to look at it would be to divide 200 by 2.5, which suggests an aperture of 80mm (difficult to accomplish, I have thought, with having a 77mm filter thread). [I have always suspected that the nearly-optically-identical A* 200/2.8, is more "honest" in its rated aperture.]



Or...? Hmmm? ("Inquiring minds want to know.")

I think there is some tolerance on the claimed apertures, the examples you care are good ones. Look at the shorter lenses, many F1.2 lenses are not actually 1.2 but a little slower