Thanks all for the comments!
pierre,
I am a bit curious.
Is the shutter of your T3i completely silent like the 50D and do you proceed using raw format of pictures ?
No, the T3i makes a faint "squish" sound at the start of the exposure, but I have not seen any vibration occurring at that time. Most of the Canon bodies make a similar sound from live-view, the 50D is the only totally silent one I have seen. The 60D
does show some vibration at the time of the slight sound. While I have not actually tested one on a microscope, I
suspect the full frame 6D might also have a very slight vibration as well. (This is based solely on an in store, hand held "evaluation" and so I hesitate to even mention it. But if I were looking for a full-frame body and completely vibration free EFSC were a critical feature it needs to be checked out more carefully).
I'll confess that very frequently the out of camera jpg's are used. But in this case all three were done from raw converted to TIF files. A 50/0.50, even an excellent one, is at about the limit (past it actually!) to obtaining a reasonably sharp looking image (effective aperture being about f/50!). So the little bit of extra "sharpness" that can be squeezed out of a raw seems to make a difference. Especially on the fine lines on each scale most noticeable on the first two images. (In reality, since resolution has taken a big hit from diffraction, the slight improvement likely comes from avoiding the JPEG stage and being able to choose the best balance between "sharpening" and noise reduction).
Wim,
I am only completely confident talking about ones I have actually used. The 50D and T3i (600D) are OK in this regard. (While I have not tried them I think it is likely that the T4i/650D and T5i/700D are the same as the 600D. But they should be checked out before purchase).
Based on reports from people I trust, the 7D and 5DII (full-frame) and 5DIII (also full-frame) are OK as well.
The 60D has some slight vibration. A 70D should be introduced this summer, but it obviously remains to be seen if it will have any "issues" in this regard. As I mentioned above, I am suspicious of the full-frame 6D, but have not really tested one.
One nice feature of the mid and upper models (those above the T
xx/xxxD) is that they use separate motors for the mirror mechanism and the shutter. This means that you can operate from live-view with the mirror up at all times, with only the shutter "re-charging" between exposures (in essence, much like a mirror-less camera). The more basic models (and most SLRs) have a single motor, so after a shot both the mirror and shutter must cycle together. It doesn't cause more vibrations in the picture (unless perhaps you run off a rapid sequence of shots with no "settle time" at all), but it does make for a quieter, "gentler" overall operation when shooting an image stack.
I've often wondered about the curl in the scales myself. Without the opportunity to see a live specimen I don't know. I could put the sample used in a "relaxing chamber" and see it there is any change, but even that would not be a definitive answer. Perhaps some forum member with more knowledge could offer some insight. I would be curious about how "hydrated" scales are on a live specimen.
Chris S.
I've contemplated suspending the FO illuminators from the ceiling and would probably go that route if vibration were a problem. (This would free up more table-top "real estate" as well). My workspace already look like a sloppy mad scientists den, so that would only add to the ambiance.
Waldo,
There is no doubt that it will be easier to get fewer stacking artifacts with a 50/0.50 than with a 40/0.95, although you do give up something in terms of resolution. I'm dealing with a dry, uncovered specimen so I don't have the subject compression often afforded by a cover slip. Plus I need the working distance above the subject to allow for the type of lighting that is preferred. So the trade-off of the smaller aperture is a little more palatable (at least that is what I tell myself

). Also, if you look at the sections photographed you'll see that they were selected to have minimum open space between the uppermost part and lowest part. Even though these were shot at 50X (with an admittedly small aperture) the number of images required were 44, 43 and 34 respectively. Quite a modest number. (All it would take is one scale jutting upward to double or triple the number of imaged needed and also likely cause artifacts in the result).