Lens Test Using A Reversing Ring

[georgedingwall]

Hi all,

I'm trying to find the best settings for my macro setup when using the reversing ring I recently got from ebay.

I have 3 Nikon lenses which fit the ring, 105mm/F2.8 macro, 50mm/F1.4 prime, 35mm/F2 prime.

The setup I used to test these was to fit each of them to a bellows and one set of extension tubes. I took sufficient frames to get the full depth of screw in focus. I then stacked in Helicon Focus and finished in Photoshop.

I had some trouble keeping the illumination consistent. My flash heads only have a course adjustment, so I had to move the heads in and out to try to balance the light. I was not totally successful, but I think the results are still good enough to make a comparrison.

I think part of the problem was that I shot each lens one after the other at the same F stop, when I should have shot both the F4 and F5.6 on the same lens then moved onto the next lens. This meant I had to adjust both the focus and the lighting more often that I should have.

I took a sequence of images with each of the lenses fitted at F5.6 an F4. The two images below show the results. The 105mm test is at the top of each image, followed by the 50mm then the 35mm image.

All 3 lenses seem to have performed well at F5.6, but there seems to be some deterioration in the F4 shots. The 35mm lens appears to have done worse at this stop than the other two.

In recent posts, Rik and Charlie have both commented that shooting with a wide aperture is better than shooting with the lens stopped down a little. For instance, I find my 105mm lens produces better results at F8 and F5.6 than it does at F4 or F2.8.

If I had a lens which could produce good results at F2, what would be the advantage of that, other than shorter exposures, be. You should have a shallower depth of field the wider open you are, leading to the need to take more frames for the sequence.

Now Rik and Charlie's results speak for themselves, so perhaps they might offer some technical reason why this should be the case.

The field of view at the subject for each of the lenses is as follows ;

105mm lens = 11mm
50mm lens = 4.4mm
35mm lens = 3mm

This is the F5.6 sequence.

[
The Camera used was the Nikon D200

Comments are invited.

[Charles Krebs]

George...

In recent posts, Rik and Charlie have both commented that shooting with a wide aperture is better than shooting with the lens stopped down a little

Don't recall saying that... sorry if I somehow gave that impression. If all we had to be concerned about is loss of resolution due to diffraction then it is true that the "faster" the lens (larger the physical aperture size) the higher the resolution can be.

But diffraction is not the only optical difficulty that has a deleterious effect on this type of image. There are various other lens "flaws" that are often greatly diminished by using an aperture somewhat smaller than the maximum. Now, if you have a very specialized, highly corrected photomacrography lens and you are using it at the magnification for which it was computed, you may not have to close down the aperture very much (perhaps not at all!) to reduce lens aberrations. In this case you can maximize your resolving capability by using full, or nearly full, aperture.

In practice, it is a very small number of lenses that are made and used in this way. Far more often, some testing (at a given magnification) will reveal what I had referred to as the "sweet spot" -- the point at which aberrations are best corrected while still maintaining good resolution by not stopping down too far.

I realize the difficulty of doing so when using electronic flash as your light source, but it is very revealing to run through the entire aperture range of a lens on a test subject. (A subject where you can look at some very fine detail). If you can do this you will clearly see the difference between apertures. With the 50mm, 35mm, and even the 105mm you mention, it would be extremely surprising to find that the best results were at maximum aperture. Also, at the magnifications you are showing us here, using an aperture opening that is any (physically) smaller than a "marked" f8 (such as f11 or f16) would inevitably result in severely reduced resolution from diffraction, and likely yield unacceptable results. Just where results would be best you will need to determine. (And it will vary with any significant change in magnification).

Without "stacking" software and techniques, or when using a single image for a subject, it is often necessary to arrive at a "best compromise" between resolution (detail) and depth-of-field. In this case you consciously use an aperture smaller than that which will provide best resolution in order to gain needed depth-of-field.

With "stacking" techniques, this is not really necessary. (However, while I have not done any real tests, my impression is that with certain subjects stacking results can be slightly better if you opt for a little more DOF in each shot when feasible)
Charlie

[rjlittlefield]

George,

Sorry for the late reply -- I've been traveling, out of touch with the Internet. (What a strange concept!)

In recent posts, Rik and Charlie have both commented that shooting with a wide aperture is better than shooting with the lens stopped down a little.

Hhmm... Well, I didn't intend to say that, and I'm sorry if that's the message that came across.

Trying to summarize:

• Every lens has a "sweet spot" -- an aperture that produces the best resolution at a given magnification.
• The sweet spot has to be found by testing.
• Most lenses (in my experience) have sweet spots around f/5.6 - f/8.
• The key to getting highest resolution is finding a lens whose sweet spot occurs at the widest aperture. A lens whose sweet spot is at f/2.8 will generally beat a lens whose sweet spot is at f/5.6 by a significant margin.
• A few lenses -- and only a few, which are fairly expensive and difficult to find -- have sweet spots wide open.

Please re-read my post of Tue Aug 22, 2006 HH:54 in your topic.
Also see my earlier post here and at this web page.

If I had a lens which could produce good results at F2, what would be the advantage of that, other than shorter exposures, be. You should have a shallower depth of field the wider open you are, leading to the need to take more frames for the sequence.

If the lens produces merely "good" results at f/2, but better results at some smaller aperture, then there would be no advantage to f/2. It happens that my Olympus 38mm f/2.8 bellows lens does give best resolution wide open at f/2.8, so that's where I often run it. On the other hand, it has almost as much resolution at f/4 as f/2.8, and in addition it has 40% more DOF at f/4, so sometimes I use that instead.

Does this make sense?

You might find it helpful to do some testing using incandescent lighting instead of strobes. That will make it easier to run a full series of exposures, from wide open to full stopped, along the lines of what I show here. Also, be aware that highly specular subjects (like shiny metal) are sometimes difficult to interpret. You might prefer working with some more diffuse subject like a moth or butterfly wing.

--Rik

[georgedingwall]

Hi Rik,

George,

Sorry for the late reply -- I've been traveling, out of touch with the Internet. (What a strange concept!)

--Rik

Thanks for the insights. You've given me a lot to think about.

Bye for now.