Why not a 50-135mm zoom lens?

Just picking up on John's point, a 50mm prime that was miraculously converted to a zoom, would change from f2 to f4 (2 stops). At 200mm it would lose a further two stops, and become f8.

Yet we don't see anything like that change in the max aperture range in practice. Pentax's new 50-200 has a max aperture range of 4-5.6, which is ONE stop, not four.

Clearly, there is some compensation going on or these lenses would not function like this.

G

The reason is optics. My example was grossly simplified to make the point, and your extension of the concept to 200mm is logical, but only in the same basic way.

Lenses are more complex thatn this, especially zooms, so there are many more design factors involved.

Even something like focal length becomes a blurred concept - what point do we measure it from for example in a zoom that is maybe 10cm long or more - where is the "nodal point" and does it move as you zoom...

Also, focal length often alters as we focus as well as zoom, usually because of Internal Focusing.

So although my simplification is true in that it explains the baisc principle, it becomes less true as we delve deeper, which I mentioned in the previous posts. I just love lenses!

John,

You're being very woolly. I still don't know what you are trying to say.

"Lenses are more complex than this". Complex than what?

The question (which relates to Pentax K mount zoom lenses) is whether the effective aperture changes as you zoom. I say, from limited experience, that IF it does so, it does so very little, and far less than you would expect if no effort was made in the design to compensate for the fact that a longer focal length should theoretically reduce the amount of light coming through the lens.

The question is a practical one; it doesn't matter whether the effect is achieved mechanically or optically; what matters is whether you can put a zoom lens on a camera in manual mode and expect to get consistent results at different focal lengths.

G

The actual aperture of a lens is the aparent size of the the 'hole' as you look into the front of the lens. The maximum that can be, of course is the diameter of the front element - which is usually the case for telephoto lenses. Wide angle lenses, hwever tend to have a smaller aperture than the diameter of the front element for optica reasons.

The relative aperture (or f number) of the lens is the aperture divided by its focal length.

Zoom lenses are mostly variable aperture. If you look into the front of a zoom lens the apparent size of the hole changes as you zoom (due to optical effects), but usually not as much as the focal length changes and so the relative aperture (f number) changes with focal length. AFAIK most variable aperture zooms are nomally variable at all apertures, not just maximum.

Constant aperture zooms compensate by mechanically linking the iris to the zoom action - effectively reducing the maximum aperture for shorter focal lenghths. This makes them more complex and usually more expensive but easier to use with external metering.

Of course if you adjust the aperture from the body (rather than the aperture ring) the electronics compensates and keeps the relative aperture constant as you zoom.

Yes, yes, yes, but......

My contention is that most modern zooms are in fact constant-aperture zooms. Some may have a variable maximum aperture, but at smaller apertures the effective aperture is constant. In other words, f8 gives f8, whatever the focal length

None of you are addressing that point!

G

Perhaps we are being too complex in our deliberations!

But I can address your specific question - most definitely most modern zooms are demonstably not a constant aperture as you zoom. The movement of various elements and/or groups within the lens may help to keep the effect under control to a greater or lesser degree, but the effect can be demonstrated by simple metering tests, as suggested by KimC.

The actual difference this makes to exposure can be very small, given the latitude of modern films abd/or CCDs.

I'm with John on this one. In fact with my FA 28-105 3.2-4.5 if I set it to F11 on the aperture ring and zoom on my MZ6 the LCD readout of aperture changes with focal length from 11 (at 28mmm) to 13 and then 16 (at 105mm) - so the camera 'knows' about the relative aperture variation.

Of course FAJ and DA lenses are constant aperture as they have no aperture ringand the aperture has to be set from the body. Some do have a variable maximum aperture of course.

John,

I think the gist of what you are saying is that any difference in aperture as you zoom is small. This says to me that the lens designer is trying to control the effect, but is not always 100% successful. I can't argue with that. All lens designs are a compromise, and there is no such thing as a perfect lens.

So let me re-phrase the question. In practice, do modern zoom lenses adhere to their stated apertures closely enough to be usable in manual mode?

I believe the answer is generally yes. However, my experience is limited, and it would be interesting if anybody has any real-world evidence of aperture-shifts that are sufficiently large to make it impracticable to use a particular (modern) zoom lens in manual mode?

G

Experience suggests that if you use the aperture ring then the aperture set is correct at the widest angle but 'closes down' by a stop or so when zoomed to maximum (so with the ring set to 11 on the 28-105 the actual aperture is 16 at 105). Of course if you are using open aperture TTL metering there will not be a problem since as the max aperture varies by the same amount as the set aperture through the zoom range then the metering will be correct.

Is a stop a significant issue for exposure? I would say so - so if metering externally for (e.g.) the 28-105 it might well be an idea to compensate by 1/2 a stop to 1 stop depending on the zoom setting.

George,
I will try and run some tests on the D tonight and give you a full answer. I tend to use fixed aperture lenses such as the 28-70/4F 28-20/2.8FA and 24-50/4F on the MZ-S because I like to use velvia and incident light meter readings. On the LX I tend to use the same in M or A versions. For print film it would not be such a problem. On cameras such as these the aperture is set on the lens and so a variable aperture lens will not give the correct exposure other than at it's widest setting though it is possible to "guestimate" the actual apeture based on the zoom and overall variation.

Amilner comment that the J lens are constant aperture is incorrect. The effective aperture will still vary. This is what I will investigate this evening and set up some tests with the D and the 28-90 which has a large aperture variation. The MZ-S as I said earlier and as Almilner says the MZ-6 recognizes that the effective aperture changes as you zoom. However because the aperture is set on the ring, the camera can do nothing about it, other than tell you and alter the shutter time. If I now open the shuuter till the finder tells me I have F8, regardless of the what it says on the ring, the time will revert to what it was before and the exposure will be correct and at an effective aperture of F8. The D must also "see" this. Now what I suspect will happen is this, assuming the lens is set to "A". If you use a constant tone, constant lite background and take a reading at the wide setting and the reading says F8, the camera will "instruct" the lens to close down to a true aperture of F8. When I zoom out to 90, the camera will want to use F8 and because it now controls the diapraghm, it will close the apeture down to a wider setting which might be F4.5 at 28 but will give an effective aperture at 90 of F8. It will do this by closing the aperture to the same position it would go to if you set 4.5 on the A ring

Now if this is the case, it will only work if both the camera and the lens have the KAF/KAF2 mount with the extra data pin. A variable apeture A lens will still be a variable aperture lens on a D and vice versa. I will let you know how I get on.

Kim

Thanks, Kim.

I await your findings.

G

Hi George,
I have had a play for a couple of hours and come to some basic conclusions. It has also thrown up some more interesting questions which I hope to investigate later. (I am away all next week on the narrowboat) When I have done that I will try to post a more complete record of the findings and my understanding of them. However, the short answer is as follows:

A "fixed" aperture lens such as the 28-70/4 FA AL is designed such that if you set F8 on the aperture, you will get a "true" F8 throughout the range and therefore the lens can effectively be used on any body in any mode. All the "variable" aperture lensses such as the 28-90/3.5-5.6 FA will only give the aperture set on the A ring at the wide setting. If you zoom out the effective aperture will be smaller and this will cause exposure problems on a non KAF/KAF2 mount camera. On a KAF mount camera, it will depend on how the lens is used. On such as the MZ-S on which the aperture is set on the lens, you must use the aperture readout in the viewfinder rather than what the A ring says to get the correct exposure.

On a camera such as the D and I suspect the Z1P, (although I can't try it at the moment) where the aperture can be set on the body with the lens on the A setting or lenses such as the J models, the camera can compensate for the aperture change. If you set F8 on the camera, it will give you F8. BUT the actual aperture on the lens does not equate to the same opening that you would get if you set F8 on the lens. Although this will give you the correct exposure, it could have quite an effect on lens quality. The lens I was using, the 28-90 has a large change. At 90mm if you select F8 on the D, the diapraghm will move to the same position it would be in at about F4.5 at 28mm. In other words at the longer lengths you may have to stop the lens down far more than at the wider lengths to get the same improvement in quality.

So to answer your question do modern zoom lenses adhere to their stated apertures closely enough to be usable in manual mode? The answer is no and there is nothing the lens designer can do about the laws of physics. However, the electronics of some camera bodies in some combinations can take account of the effect and hide it by compensating to give you the correct exposure with perhaps the slight loss of quality that you would expect if you used the lens at a wider aperture.

Finally a word of caution, such bodies can only do this with lenses that give the correct info ie KAF/KAF2 mount lenses. Variable aperture A series zooms such as the 35-70/3.5-4.5 will only be correctly exposed at their wide setting. At all longer lengths, there will be a degree of under exposure with such bodies as the D and MZ-S.

Hope this helps, it has certainly given me a better understanding of why some camera/lens combinations work better than others. Although I was instinctively using the right combinations, I now know why they are right.

Kim

To get back to the original subject:

woo-hoo!! I found one of the 45-125 lenses at KEH in Atlanta, Geogia. It is in "excellen"condition (which at KEH means pretty much mint, I buy "bargain" equipement all the time and it's fully functional and totally clean. Just some brassing, nothing to effect the photograph).

This is after fruitless searching on Amazon, KEH, ebay in seveal countries, and other odd auction sites. Anyone looking for one go to KEH.com right now!!

In fact they had two, I figured since Bojidar Dimitrov's site lists availability as "roughly once a year" I figured it would be wise to grab it. Now I just need to find the odd double-peice hood!

The next trick is to modify it for matrix metering with the D. I'll just put a body on a scanner and scan the mount, so I have a template.

On the subject of varifocal lenses (which is the technical correct name for a variable-apeture zoom) you can only trust the set apeture if the camera sets the apeture, because the apeture blades do not change with focal length.

That's why I stick to constant apeture zooms, they are more complex to make and tend to be the manufacturer's best offerings for zooms.

There was a time I wouldn't go near a zoom. I've relented in my old age.

Sorry to disagree, but a vari-focal lens is one which has a variable focal length. This is subtely different from a zoom. The zoom is supposed to maintain it's point of focus as you change the focal length (essential in a movie, useful for still photography) but a vari-focal has to be refocused after each alteration of focal length.

One of the early Vivitar lenses (was it a 38-80mm?) was widely acknowledged even by Vivitar as a Vari-focal design.