To my great surprise, the specimen was far more interesting dusty than it would have been clean. The reason is that the dust revealed a fine array of hairs that I would have completely missed on a clean specimen.
Here is the beast as a single picture. Notice that it just looks like a dusty hoverfly with apparently naked eyes. You can see a few hairs near the bottom of the eye, but basically there's nothing interesting in this picture.
Here is the same stack processed into a crossed-eye stereo pair by Zerene Stacker PMax.
If you can see stereo, it will be immediately obvious that much of the dust does not lie on the surface of the eye touching the ommatidia. Instead, the dust is floating above that surface by a considerable distance.
Here's a closer view, a little bit left and below center of the full frame.
If you can't see stereo, then the above pairs are pretty boring at best. In that case, perhaps this rocking animation will make sense if you study it long enough.
Now, getting back to still pictures, here is the stacked composite alongside a single frame from the original stack.
In the single frame, it's possible to make out what is only hinted at by the stacked composites: in this area on the front of the eye, the entire surface is covered by a dense array of fine hairs, apparently one from each vertex between the ommatidia. These hairs trap most of the dust in a layer near the top of the hairs, well away from the ommatidia.
Elsewhere on the eye, the pattern is different. Over the large ommatidia on the sides and top of the eye, the hairs are shorter and more sparse, occurring at every 2nd or 3rd vertex instead of every one. At the back of the eye, I cannot see any hairs at all, even over the small ommatidia.
Now that I have this understanding, it's not too hard to generate a single image that shows some of these aspects of the distribution. Here is a single carefully selected frame, very heavily sharpened and levels adjusted to bring out the hairs.
What's most interesting to me in this exercise is that I never would have noticed most of these hairs by direct viewing of a clean eye under a stereo scope. The hairs are so thin that in direct view they get lost in the strong pattern of the ommatidia, just like they do in the first stacked result shown at the top of this thread. It was only the dust layer combined with the stereo view that gave me a clear idea there were hairs to be found, if only I looked hard enough.
All in all, a fascinating exercise. Thanks for listening!
--Rik
Technical notes: Photographed at 5X using a Canon T1i camera with Nikon CFI BE 10X NA 0.25 objective on Canon 100 mm f/2.8L IS USM macro lens. Focus stacked with Stackshot at 15 microns, 110 frames. Stereo pairs at +-4.5% using Zerene Stacker synthetic stereo.
