Changing tube lens/objective distance changes magnification

[seta666]

Hello,
These days I am doing some test to find out best tubelens for APS-C from the lenses I have.

I have been using the Raynox 5.9 dioptes for a couple of days and I quite like it.

I wanted to see if I could find the sweet spot for distance between raynox and objective but initial tests show me something I did not expect. Tube lens is always in the same place, only thing that changes is separation between both lenses.

Changing the distance between the tube lens and the objective changes magnification. Is this supposed to happen?

In this picture you can see the different FOV; tube lens is focused past infinity to get designed magnification

Green Raynox 13mm away from mitutoyo at 10X (tube lens, Iris, mitutoyo)
Red Raynox 40mm away from mitutoyo (tube lens, 27mm M42 tube, Iris, Mitutoyo)
Black Raynox 67mm away from mitutoyo (tube lens, 54mm M42 extension tube, Iris, Mitutoyo)





I do know yet what works best because single shot were not enough to determine it; I will run some stacks tomorrow. Meanwhile this single shots test shows that the further awy the lenses the less the magnification


EDIT

Today I have tested the Raynox +5.9 as tube lens focused to infinity; magnification with mitutoyo 5/0.0.14 is 4.1X

I have run three stacks with lenses separated 13mm, 40mm and 67mm; subject was sand papper, like Rik normally does.

I can see some improvement in the corner when the lens is separated 40mm and 67mm; difference is not that big anyway, you have to do some serious pixel peeping to see it



Full size
http://farm9.staticflickr.com/8442/7996 ... 4dfd_o.jpg

Regards
Javier

[rjlittlefield]

The magnification will change with separation if and only if the rear lens is not focused at infinity.

The change goes one way if the rear lens is focused closer than infinity, but the other way if it's focused farther.

--Rik

[seta666]

Interesting!! I do not remember reading this before; so I will better test it infinity focused

Regards
Javier

[rjlittlefield]

As far as I can recall, the effect has been mentioned only in passing, never discussed in detail. I spent some hours experimentally confirming the theory one day, but never wrote it up.

Without getting into calculation details , it's easy enough to "explain" why the effect happens.

Imagine that the front lens forms a virtual image that the rear lens must focus on. If the distance between rear lens and sensor stays the same, then the virtual image must stay in the same place too, or it will go out of focus.

Now think of the subject and front lens as a pair.

If the rear lens is focused at infinity, then the distance from subject to front lens must be such that the virtual image is at infinity. In this case the subject plus front lens can be placed anywhere and the image will stay focused with no further adjustments. Magnification will not change with separation.

But suppose the rear lens is focused at some distance other than infinity. In this case when you move the subject and front lens as a unit, the virtual image moves with them and goes out of focus. To make it focused again, you must slightly change the distance between subject and front lens, so "it makes sense" that the magnification would change also.

The reason I've put quotation marks on "explain" and "it makes sense" is because the explanation I've offered turns out to have some hand-waving in it. From the explanation as written, it seems like there should be a simple relationship between magnification and distance between subject and front lens. But there isn't. In all cases, when the separation between lenses gets larger, the distance from subject to front lens gets smaller. But the change in that distance is miniscule in comparison to the change in magnification, and in some cases the relationship is positive (more distance, more magnification) while in other cases it's negative (more distance, less magnification).

I have not yet figured out an explanation that is both intuitive and computationally correct.

Instead, I generally restrict my advice to something like "When the rear lens is focused at infinity, the magnification is just the ratio of the focal lengths. When the rear lens is not focused at infinity, the magnification will change, but it's difficult to predict exactly how much."

--Rik

[seta666]

Thank you Rik for the extra feedback; yesterday I did not understand what was happening but I did not stop to think I was using the set up in a way it is not designed for.

I normally get good results when pushing up infinity lenses, but sometimes it does not work as I expect. There is lots of thing I still have to understand about infinity set ups

Ideally one would have several tube lenses at different focal lengths, but to be able to get different framing one would need quite a lot of them. This is why I just use them as if the whole thing was a finite set up.

With the 5X and 10X lenses works very well, may be with higher NA lenses the drop in performance is more evident

Regards
Javier

[rjlittlefield]

With the 5X and 10X lenses works very well, may be with higher NA lenses the drop in performance is more evident

Yes, exactly. See HERE and note the loss of contrast for fine detail as the tube lens is focused away from infinity.

That said, a lot of the degradation can be made up by strong sharpening, even by something as simple as a single USM with well chosen radius and strength. See for example HERE.

--Rik

[seta666]

Thank you for those links, very interesting piece of reading. I guess a little loss of contrast is not a big deal; is it the only inconvinience of working out of infinity ? Worst thing would be other kind of aberrations like CAs, astigmatism, etc.


I happen to have found another interesting tube lens, my old summar 120/4.5 performs very well corner to corner at 6X (infinity) with nikon CFI 10/0.25; at least on the NEX-5N

Now I have a 120mm summar, the 170mm Raynox (+5.9 diopters), morfanon 173mm and SOMCO 240mm; it would be nice to fill the gaps with a 150mm and a 200mm. I also could try my Canon USM 100/2.8 macro (non-L) but I read that it might vignete at infinity

Regards
Javier

[rjlittlefield]

is it the only inconvinience of working out of infinity ? Worst thing would be other kind of aberrations like CAs, astigmatism, etc.

I don't really know. In general lenses used away from their design point are vulnerable to many strange effects. Using an infinity objective, focused at infinity, combined with a tube lens that also works well at infinity, is a sort of "sweet spot". Of course whatever CA is present in each lens alone will add up in the combo, so if you're lucky they cancel and if you're not it gets twice as bad. But if you drag the lenses far from infinity then I don't know what aberrations they may pick up beyond spherical.

--Rik

[ChrisR]

it would be nice to fill the gaps with a 150mm and a 200mm

Now, if you could borrow Harold's 150mm Componon and his 180mm Wedding Anniversary Tamron...?

(I've used the former with a Nikon 10x CFI, which seemed to work fine.)

[seta666]

Now, if you could borrow Harold's 150mm Componon

I sold him my componon-S 150mm few weeks ago ;-)
Anyway, I am already trying few differente lenses and they all seem to work pretty well, at least on APS-C sensor

[Harold Gough]

Now, if you could borrow Harold's 150mm Componon and his 180mm Wedding Anniversary Tamron...?

I wish to make it clear that, after some four decades of a relationship with Tamron SP Adaptall-2 lenses, it is a purely working relationship. It was Tamron's 35th anniversary!

Harold