Lens Test With Four 50 mm Lenses

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

georgedingwall wrote:The more I get into this subject the more complicated things seem to get.
It's that way with every subject I can recall getting into. After a while, if I'm lucky, my understanding converges to "here's a simple model that's pretty good" and "here are some gory details not captured by the simple model". For optics, I'm still going through that process. Trust me, you're doing fine!
georgedingwall wrote:If you were put on the spot, what would you say is the best lens I could get for my sort of set up. I get about 8 to 1 magnification on the D200's 23.6mm sensor as a maximum if I use all of my accessories. Is there an ideal lens that would work best for this magnification, assuming money was not a limiting factor.
No, I don't think there is an "ideal lens".

Go back to my mini-report posted here. I tested 4 lenses that might reasonably be used at 10:1 (more precisely, to image a 2.3mm subject width). Those lenses are a 10X microscope objective, a 16mm Luminar, and the 20mm and 38mm Olympus. They span quite a wide range of delivered resolutions (as illustrated in the mini-report). But based on other considerations, I can easily imagine using any one of the four for a particular subject.

From what I've seen of your tests, my guess is that no camera or enlarger lens is going to do much better than what you currently have. Based on my tests, I think there's a good chance that switching to a microscope objective would get you a lot more resolution. That comes at the cost of smaller DOF = deeper stacks, some less pleasant handling of OOF features (especially bright spots), and no feasible mechanism to adjust aperture. Lots of downsides, but the objective could significantly broaden your range of capabilities, where another camera or enlarger lens will not.

--Rik

georgedingwall
Posts: 207
Joined: Tue Aug 01, 2006 11:15 am
Location: Invergordon, Scotland
Contact:

Post by georgedingwall »

Hi Rik,
rjlittlefield wrote: Based on my tests, I think there's a good chance that switching to a microscope objective would get you a lot more resolution.
--Rik
I've managed to fit a 10X microscope objective to my setup today. I found a lens cap that was exactly the size of my M42/Nikon T adaptor and after drilling a 3/4 inch hole in it, the objective fits perfectly.

I was a bit surprised at how close you need to get to the subject to get it in focus, but I suppose I shouldn't have been, given how close it is to a slide when mounted in the scope.

I took this image of my old faithful test screw with the objective. I used only a 1.4 X tele-extender and 8mm of extension tube for this image. The tube was necessary to minimise the vignetting present without it. I think I may need to add more extension to completely get rid of it.

This is not a very good quality lens, so I'm not sure how the result compares with a good lens. However, the quality in the central area seems to be very good, but falls away quite fast as you move out to the edges.

I also see some of the problems you mentioned with OOF parts of the image.

All in all, I'm quite pleased with this first try. Can you suggest a good quality 10 X objective that would give better results that I got with the one I've used here.

The stack is made from 52 frames with an adjustment of 0.04mm between frames. This image is uncropped after stacking.
Last edited by georgedingwall on Mon Jul 26, 2010 6:33 am, edited 1 time in total.
George Dingwall

Invergordon, Scotland

http://www.georgedingwall.co.uk/

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

George,

Excellent first test -- I'm happy to see this! :D

The best objectives to use for this job are designed for older microscopes in which there are no lenses between the objective and the eyepiece, just as (except the teleconverter) you have no lenses between the objective and the camera sensor. Those objectives can be identified by some notation like "160/-", referring to a 160-mm tube length and no cover glass.

It's also best to use an extension that places the objective close to its design point, with the shoulder of its mounting threads about 150mm away from the image plane / sensor position. Any closer and you're likely to get vignetting and/or serious loss of quality away from center. (Objectives are only designed to project about a 20mm field at 150mm distance.) Much more extension, and you'll probably get empty magnification at best.

I have worked recently with only three different objectives, so I don't have much feel for the span of quality. Those three all work OK. They are
  • aus JENA 10X NA 0.25 (marked only "10/0.25", "160/-", presumably an achromat),
  • Edmund Optics "International Standard" 20X NA 0.40 achromat, and
  • aus Jena 3.2X NA 0.10 semiplan.
Working distance (WD) is a big consideration and something to be cautious about. It can vary quite a bit even between lenses of similar spec, and some of the more highly corrected objectives can have very short WD's. For example, of lenses that I can see right now at Edmund Optics, the WD of inexpensive 10X objectives varies from 10mm (part number NT38-342, a semi-plan), down to only 1.5mm (part number NT43-907, a plan). When I bought my 20X, I think I got to choose between WD=3.3mm for an achromat and WD = 0.33mm (!) for a cheap semi-plan. There are also special long working distance objectives, but they tend to be pricey (like 10 times more expensive).

I have no experience buying objectives through eBay, but that's surely an attractive way to go. At this moment there's a Nikon LWD Plan Achromat 10X bidding at 24.95 USD (item 280036183472). LWD here stands for "long working distance", but for this particular lens that's only 10.5mm. Not long at all in comparison with other 10X LWD's, but it makes you wonder just how short their "normal working distance" would be!

BTW, for evaluating lenses at these magnifications, I suggest switching subjects. Something like a moth wing is likely to have more detail at a wide range of scales (pardon the pun!) than a manufactured object like that screw.

--Rik

georgedingwall
Posts: 207
Joined: Tue Aug 01, 2006 11:15 am
Location: Invergordon, Scotland
Contact:

Post by georgedingwall »

Hi Rik,

Thanks for the info. It gives me a lot to think about.

The lens I have has this etched on it.

Plan 10/0.25 160/0.17

As for the screw, I only used because it seemed to have a lot of surface texture in the flaking plating, and also it doesn't droop or dry out under the lights of my setup.

I'll try your suggestion of the wing when I can get hold of one.

As an aside, I tried this lens on the Caddis Fly I recently shared in the gallery, and it did not come out as good as the lens I used for that image. I seemed to loose some of the definition in the fine hairs, bvut it may also be me trying to get used to the very shallow D of F that I get with the microscope lens. I'll keep trying.

Bye for now.
George Dingwall

Invergordon, Scotland

http://www.georgedingwall.co.uk/

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

George,

That objective is designed to look through cover glass that's 0.17 mm thick ("160/0.17"). Using it without cover glass might soften the image a bit, but I have no experience. (Cover glass is very important at high magnifications, but at only 10X, I don't know.)

Try it without the TC and at 150mm extension.

You definitely need a small focus step when working with these large aperture lenses. Nikon MicroscopyU quotes the DOF of 10X NA 0.25 as being only 8.5 microns = 0.0085 mm = 0.000335 inch. That's consistent with my experience. My latest 10X/0.25 moth wing was stacked at 3 steps per 0.001 inch. When possible, I stack at 4 or 5 steps per 0.001 inch to give myself more leeway for an occasional sloppy move. My table has tick marks at 0.001 inch, with 100 marks per rotation, so steps are getting pretty approximate in this range. It should be easier with your finer screw.

BTW, with such small DOF, there's no need to let the stacking software "correct" for magnification between images, and it's liable to hurt more than it helps. The moth stack was done with all of HF's autoadjustment parameters set to zero, which is typically the first thing I try these days. If the focus steps were smooth, it works fine. Once in a while there's a bit of lateral shift between frames, in which case I enable vertical and horizontal position adjustment, but not rotation or magnification.

--Rik

georgedingwall
Posts: 207
Joined: Tue Aug 01, 2006 11:15 am
Location: Invergordon, Scotland
Contact:

Post by georgedingwall »

Hi Rik,
rjlittlefield wrote:George,

Try it without the TC and at 150mm extension.

You definitely need a small focus step when working with these large aperture lenses. Nikon MicroscopyU quotes the DOF of 10X NA 0.25 as being only 8.5 microns = 0.0085 mm = 0.000335 inch. That's consistent with my experience. My latest 10X/0.25 moth wing was stacked at 3 steps per 0.001 inch. When possible, I stack at 4 or 5 steps per 0.001 inch to give myself more leeway for an occasional sloppy move. My table has tick marks at 0.001 inch, with 100 marks per rotation, so steps are getting pretty approximate in this range. It should be easier with your finer screw.

--Rik
I did try it without all of the attachments, and although the central circle was OK, there was heavy vignetting. I'll try it with the extension and without the TC as you suggest.

My sliding table has an adjustment of 0.00002 Mr. Which gives 50 x 0.02 mm divisions per turn of the adjuster. I've just done a stack of 66 frames with an adjustment of 0.02mm bewtween exposures. I'm not all that happy with it. I tried your advice of zeroing the auto-adjust in the Prefs, and didn't notice much difference.

The main problem is with some fuzzy areas in the final image which are not there in the idividual frames. I'm thinking that Helicon is struggling a bit dealing with some of the out of focus areas and is overlaying some of this fuzzyness onto the frames where the same area is sharp.

Here is the stacked image. This is straight from HF, I have not done any cleaing up of artifacts yet. You can see several fuzzy areas in the hairs above the eyes. If you look at the fly's right eye, you can see a heavy ghost of the in focus hair that crosses it. Also in the area between the two antennae.

There also seems to be some chromatic abberation which is obvious on the hair near the fly's left eye.

I'm not entirely sure if it's the lens, my technique or HF struggling with the characteristics of this lens. Probably a combination of all three.

I can definitely see that this lens, or something similar, has the potential to produce great images. I'll keep at and see if I can't improve things a bit.
Last edited by georgedingwall on Mon Jul 26, 2010 6:34 am, edited 1 time in total.
George Dingwall

Invergordon, Scotland

http://www.georgedingwall.co.uk/

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

George,

Some subjects are more difficult than others, and you've chosen one of the hardest.

It's not unusual to have visible chromatic aberration. Your image shows more than I usually see, though. That could be due to the coverslip correction, or to running the lens at an extension different from what it's designed for, or perhaps some funny interaction with the TC, or perhaps the objective is designed for use with a strongly correcting eyepiece.

Whatever the cause, CA can be greatly reduced by software such as PTLens. For example:

Image

The other problems are harder to deal with and I think some of them are intrinsic to the use of large-aperture lenses.

The ghost of the infocus hair over the fly's right eye (left side of this image) looks like what commonly occurs whenever there are strongly contrasting features in very different depth planes. There is a fundamental problem that when the lens is focused on the eye, it's still seeing the hair as a very blurred but visibly bright overlay on the dark eye. Check your individual images and I'll bet you see the same thing there.

If so, then there's not much that software can do about it. It's theoretically possible for software to figure out the relative contributions of in-focus background and OOF foreground, and use that to remove the effect of the blurred foreground. But to do it in practice is very difficult and I don't think anybody has demonstrated success yet. To the best of my knowledge, there's nothing in HF that even attempts it.

The mystery to me is why you don't have the same problem with the hair on the other eye. But I notice that the geometry of that hair seems a bit strange, like it's thicker over the eye than over the face. By any chance, is that hair broken off and lying against the eye?

By the way, this problem is also more visible with bright foreground and dark background, because adding a little bright to dark makes a lot bigger percentage change than adding a little dark to bright. It sounds weird, but given free choice of subjects, you'll often get better results by picking ones with dark hairs.

The other problem is getting detail replaced by fuzzy haze, such as behind the hairs near top of image. I suspect the loss of detail is not fundamental (though the brightening is, as discussed above). You may be able to make some improvements by tuning the HF parameters, but I don't know how much. Again, this is a common problem with dense mats of hair, see for example the long hairs at the end of my click beetle trigger pictures.

--Rik

georgedingwall
Posts: 207
Joined: Tue Aug 01, 2006 11:15 am
Location: Invergordon, Scotland
Contact:

Post by georgedingwall »

Hi Rik,

Thanks for those pointers.
The mystery to me is why you don't have the same problem with the hair on the other eye
The hair on the left eye is broken and lying directly on the eye surface. I suppose the best answer to this is to be a little more careful when setting up the scene and to avoid those circumstances which will cause artifacts when stacking. I find your observation s on the Light over Dark background fascinating. I will watch for this in future.
It's not unusual to have visible chromatic aberration.
I can get rid of most of the chromatic abberation using Photoshop CS2. It has a lens distortion filter which has a control for chromatic abberation.
The other problem is getting detail replaced by fuzzy haze
I'll have to do a bit of further experimenting with the HF settings to see if I can find a better combination.

Many thanks for your input on this topic.

Bye for now.
George Dingwall

Invergordon, Scotland

http://www.georgedingwall.co.uk/

Post Reply Previous topicNext topic