[UncleVanya]

Except that how the absolute aperture is defined depends on the lens design and the domain (it means something different in photography than in astronomy). In photography, it is applied behind the entrance pupil. The exact location varies by design and is critical to lens performance characteristics and effects such things as focus shift and distortion. This is a recurring concern to large format photographer attempting to adapt so-called "barrel" lenses to their cameras.

Here is a bare summary that I am sure you will enjoy

http://hyperphysics.phy-astr.gsu.edu/hbase/geoopt/stop.html

In general, a "stop" applied forward of the entrance pupil will not attenuate light through the lens unless it impinges on the angle of view (vignette) and even then, it generally does not affect the path of light (collimation).


Steve

YMMV - but in recent discussions of the 50mm f/1.4 and step ring hoods the proof was in the pudding when they went back and calculated that the lens was actually sharper due to the smaller aperture which surprise was in fact metering as though it was a smaller f/stop.

Take the 300 shoot an artificially lighted subject with known light. Then remove the 49mm filter apparatus and reshoot. Do both wide open. Submit the images for analysis and let's decide collectively.

Finally let's be clear - I may have confused the issue by concentrating on this specific example where there is a filter in front of the front element - the proposal was for a smaller lens with a much smaller profile. Anything narrower than the current design is likely to be a higher f/stop. There are a number of optical sites that discuss telephoto designs and why M4/3 for example has smaller lenses but not that much smaller when you talk telephoto and speed. The problem is pure physics and it isn't the image circle that limits the size reduction in the case of most telephoto designs - the distance from the sensor also plays a role. But ultimately long fast lenses need to be fairly large diameter.

Steve's other point that you can reduce or place a narrower ring in front of the actual aperture is due to the refractive nature of a lens - as pointed out by someone else a fresnel design can modify the effective aperture by changing the refraction (if I understand that right which I might not).

In the end - not withstanding the fact that putting a 49mm filter on the 300 f/4.5 might not cause light loss (I'll test that myself soon) the fact is that says very little about how much the overall width can be reduced.

[MJSfoto1956]

I never said the camera wouldn't do that. I said the images wouldn't be exposed right. Did they all look exactly correct? Or were those that were taken above f/4.5 but below f/6.1 incorrect?

The exposure across exposures appeared to be identical.
Further, when I took the step rings off, the exposures appeared to be the same as those taken with the step rings.

M

[UncleVanya]

The exposure across exposures was identical.
Further, when I took the step rings off, the exposures were the same as those taken with the step rings.

M

Steve seemed to have data that indicated refraction could allow for this. That's surprising but nice to know. The next question would be if the reason extends to allowing the elements farther back inside to be narrower also.

[MJSfoto1956]

I want to remind everyone that I said I was surprised -- I fully expected that a 58mm filter would be do the trick, but not a 49mm.
I'll do a more thorough test tomorrow and post the originals somewhere for examination.

Michae

[MJSfoto1956]

Ok, things are a bit more nuanced than I originally thought.

With more careful testing I have identified a more accurate description of what is going on:

@ f/4.5 a 52mm step-down ring produces almost indiscirnable vignetting on a Pentax crop-sensor body with the exposure being about 1/3 stop darker in the center and 1/2 stop darker in the corners. So there seems to be something to the concept of diffraction coming into play here.

The following list identifies what step-down rings are usable at specific apertures with no measurable difference in exposure.

• @ f/4.5 a 58mm step-down ring produces an identical exposure as no step-down ring
• @ f/5.6 a 55mm step-down ring produces an identical exposure as no step-down ring
• @ f/7.1 a 52mm step-down ring produces an identical exposure as no step-down ring
• @ f/9.0 a 49mm step-down ring produces an identical exposure as no step-down ring

As such, I was wrong in my initial impression -- sorry for any confusion I may have caused. However, I can say that should you need to use a smaller filter than 67mm on your FA* 300mm f/4.5 you can with confidence using the above as a guide.

Michael

P.S. I might add that at no time did I ever see any classic "black" vignetting regardless of the size of the step down ring or aperture. So this lens definitely behaves differently than my DA* 60-250mm and there is some magic going on here.

[Digitalis]

FF vs APS-C is independent of front element diameter

oh great, can I put a 49mm filter on my sigma 100-300mm f/4 APO EX DG now? instead of shelling out for 82mm filters? sorry, I was being sarcastc.

You, of all people on pentax forums would be the last person I would expect to say that. The entrance pupil has a very real impact upon optical performance, and the mathematical relationship it has with the effective aperture of a lens cannot be ignored. I'm not talking about just vignetting here, I would expect a 300mm f/4.5 with a 49mm filter adapter will have serious issues with coma due to the reduced filter thread occluding the entrance pupil. There might be undetectable vignetting, there might be no detectable reduction in transmission or effective aperture of the lens, but there will be other optical consequences. Lens designers don't just cram optical elements together and over state the filter diameter for their own private amusement.

[Alex645]

I thought this Nikon 300mm f/4 was interestingly small:

Nikon AF-S NIKKOR 300mm f/4E PF ED VR Lens 2223 B&H Photo Video

The fresnel element makes this relatively tiny package possible at the cost of some wiggy flare under the right conditions. I don't know that making it aps-c could save much in size though, I didn't think there was ever a major savings for telephotos (but I could be wrong).

Or similar in size, weight, and dimension, but cheaper than the Nikon would be a Pentax SMCP-DA 200mm f/2.8 with a Pentax 1.4x teleconverter. 280mm is not 300mm, but it's close.

[biz-engineer]

Lens designers don't just cram optical elements together and over state the filter diameter for their own private amusement.

It could be that larger elements than optically required would be used , for instance if it made manufacturing easier. It's only a supposition since I'm not an optical designer so I can really tell. In fact I don't know how easy it is to reduce the diameter of a lens of a certain curvature and thickness once finished.

[kh1234567890]

Lens designers don't just cram optical elements together and over state the filter diameter for their own private amusement.

The DA50-200 went from 52mm to 49mm filter dia in the WR version with no changes to the optical design and got better if anything ...

[Digitalis]

The DA50-200 went from 52mm to 49mm filter dia in the WR version with no changes to the optical design and got better if anything

I suppose it is an unintentional oversight that the vignetting is worse on the lens with the smaller filter thread.

[Rondec]

I think the thing is that even if the front element got a lot smaller and had a smaller filter size, the size of the lens as a whole probably would have to be the same size. It would be handy to have a 49mm filter size, but I just don't think it would make the lens a whole lot easier to handle or transport.

[Lowell Goudge]

oh great, can I put a 49mm filter on my sigma 100-300mm f/4 APO EX DG now? instead of shelling out for 82mm filters? sorry, I was being sarcastc.

You, of all people on pentax forums would be the last person I would expect to say that. The entrance pupil has a very real impact upon optical performance, and the mathematical relationship it has with the effective aperture of a lens cannot be ignored. I'm not talking about just vignetting here, I would expect a 300mm f/4.5 with a 49mm filter adapter will have serious issues with coma due to the reduced filter thread occluding the entrance pupil. There might be undetectable vignetting, there might be no detectable reduction in transmission or effective aperture of the lens, but there will be other optical consequences. Lens designers don't just cram optical elements together and over state the filter diameter for their own private amusement.

you are taking things out of context. The whole comment was made about a 300mm telephoto lens, and the OPs comment that he could put a 49mm filter on an F4 lens without vignetting, so why don't Pentax make a 300mmAPS-C with a 49mm filter.

But aside from that, let's all make sure we separate filter diameter from front element diameter. Front element diameter directly impacts speed.

Filters influence vignetting, and depending upon location of the front element relative to the filter threads can really impact the field of view. No argument there

[bdery]

F stop is defined as focal length divided by the maximum aperture (not f stop but actual diameter)

Careful, that's a simplification. In fact the correct way to calculate the f-number is to use the projection of the aperture stop at the front of the lens. There is rarely a reason to make the front element larger than that, but it's not a fixed physical property.

I was able to go all the way down to 49mm

You stacked and extended the distance between your opening and the front element. Putting an iris in front of the lens, close to its surface, would be a better way to actually measure this.

[UncleVanya]

Careful, that's a simplification. In fact the correct way to calculate the f-number is to use the projection of the aperture stop at the front of the lens. There is rarely a reason to make the front element larger than that, but it's not a fixed physical property.

It is common to see it defined this way. Do you have a good reference that I can use to understand the correct method? Steve posted one earlier but honestly it went around the topic a bit obliquely I think I got most of it but I would like something a bit more directly focused on this if you have anything you can share. If not no sweat.

[MikeD]

By definition the f stop of your 300mm with a 49mm filter is f6.1

F stop is defined as focal length divided by the maximum aperture (not f stop but actual diameter)

300/49 = 6.1

Seriously - you are giving F stop crop sensor lessons to MJSfoto 1956 have you visited his website, I have and if he speaks I listen. I just wish MORE of the really top photographers would post here. I never see anything from Dr. Orloff and others.