First though, I thought I might try a different test subject. Minerals are fine, and of course are the ultimate final test given that they are what a actually intend to photograph, but I began thinking of some flat subject with lots of detail.
I want a flat subject to eliminate questions of focus and depth of field. I want a subject with plenty of fine detail, sufficient to push the limits of the optics I am testing.
So I thought of a semiconductor of some kind. I have lots of EPROM chips with quartz windows, but the silicon dies in them are pretty small. They are also pretty boring, usually two grid like areas of memory cells. Then I remembered I had a stash of intel 8751 microcontrollers that did not erase properly under a UV eraser. I had been keeping them for some project which could use an external EPROM, assuming the processor itself on the chip was still intact. They have been sitting around for over ten years though, and the odds of them actually getting put to use are miniscule.
The name of the game is to split the case open exposing the chip.
These ceramic cases have an upper piece with a fused quartz window,
(The quartz passes the short wavelength UV light used to erase the
internal EPROM, but has far too much distortion for our purposes).
glued to the lower case which contains the chip itself.
I propose to set it on its edge, place a chisel on the glue bond
and try to cleave it like I was a diamond cutter.
To my amazement, it splits neatly on the first hard blow.
And here is the exposed test subject. The die itself measures 0.233 inches square (5.92 mm) - we could call it a 6x6 mm target.
So, just as a warm up, here is the full image, resampled from the 5616x3744 pixel image the camera takes to a 800x533 pixel image that is reasonable to pass around on the web.
This was taken with the MPE-65 lens (at f/11) set to 5x magnification (full extension). This would make the 6mm object into a 30mm image on the chip, and the full frame chip in my camera is 36mm wide, so the way this image seems to fill the frame makes sense.
This was shot at ISO 125, Tungsten white balance, 2 second exposure with a single tungsten light from one side. Live view was active, so the mirror was locked up. The exposure was triggered by hand (bad), but with the 10 second timer so I feel confident virtually all vibration from the shutter push had calmed down.
And here is a 800x533 pixel section from the lower right of the image above (so you see the pixels 1:1 on your screen). This is 1.02 mm wide.
Lots of questions arise at this point about the MPE-65 lens. In particular, would it do better at a different aperture? We will do a series of tests one day and find out.
On to the first microscope objective. This brings up an important and unexpected side note. When I first set up to take photos with the Leitz objective, I activated live view on my camera to focus. I focus first with no zoom, then with 5x zoom, then with 10x zoom. With 10x zoom, the image was vibrating wildly. I discovered that the table I have my camera on was pressed against the wall, and mounted in that wall is a running air conditioning unit. When I moved the table away from the wall, visible vibration on the camera LCD apparently stopped, but more on this below.
This is my Leitz finite 10x objective (NA 0.23, marked for a 170mm tube).
I mount it on the belows with the minimum possible extension
in an attempt to get a magnification I can compare to the MPE-65.
This turns out to be about 55mm of extension.
I get this image:
A 800x533 crop, again from the lower right, yields this:
But this test is extremely unfair. With this small amount of extension we are asking this lens to cover a much wider angle of view than it was ever designed to do. This shows in the image due to field curvature causing loss of focus near the edges, along with obvious distortion. Consider the angle involved with a 36mm wide sensor and 55mm of extension on an objective designed for a 170mm tube. As a fairer measure, here is a crop from the center region of the above image:
Now, lets extend the bellows almost fully (which gives close to the 170mm tube length this lens expects) which will give us about 3 times the magnification obtained above. Here is the full image, note the vignetting.
Here is a 800x533 pixel crop from the lower right in the above image.
Both of these images are big surprises. They are drastically worse than the images obtained at shorter extensions and the cropped image shows artifacts that look like diffraction spikes (or image motion). I was so suspicious of this that I retook the image carefully verifying focus - it is not a focus issue.
Just on a hunch, I turned off my wall air conditioner and retook the image. Well, well, well, this is much better.
This crop shows a well defined particle of debris.
This crop shows nice detail on top of the wire bonds.
All in all I am impressed with this Leitz objective. I was told that the Leitz objectives are typically not well corrected for chromatic aberration, and I could expect to see color fringes, especially towards the edges of the image. And maybe I do -- I see some purple color in the last photo on top of the wire bonds. I need photographs from some other objectives to compare it to.
I need to carefully calculate the field size for the above crops, but it is on the order of 0.3 to 0.5 mm in width.
Tom's Mineral Photography Info / firstname.lastname@example.org