John,
You made a good camera choice. I too think the Canon looks a bit better, although frankly in this comparison I would not expect to see much difference at all. That is because a 60X objective is not very demanding as far as sensor resolution is concerned. And this is especially so when you magnify the intermediate image on the Canon an additional 2.5X.
It is somewhat counter-intuitive, but the most "demanding" objectives are lower power, relatively high NA (for their magnification). So something like a 4/0.16 or a 10/0.30 will require a sensor with much higher pixel density than something like a 60/0.85 or a 100/1.40... nearly 8 times as many pixels.
So in your comparison, the Canon APS-C at 16Mp is far oversampling based on the optical conditions. It would really require only a 1 to 1.2Mp sensor in that camera to record just about all the resolution a 60/0.85 objective (through a 2.5X projection eyepiece) can produce. (
I know... I know... sounds unbelievable and actually a little depressing but is is the nature of imaging with full spectrum light and the diffraction caused by very small effective apertures
)
If the ToupTek is typical of 2Mp cameras of that type it has a sensor of (at best) 1/2" size (that's not the actual dimensions, it is just the funky way they designate sensor sizes). They are typically coupled to the trinocular tube with a 0.5X adapter. If that is the case with your equipment then the ToupTek would need 2 to 2.7 Mp when using the 60/0.85.
So choosing a microscope camera based on the number of pixels is not that straightforward. With low power, decently high NA objectives, (and the intermediate image properly fitted to the sensor) you will typically need a lot of pixels and your 16Mp sensor is fine and definitely worth having. The 2Mp ToupTek would not be capable of capturing the full resolved detail.
With very high power objectives the demands drop dramatically. The Canon 16Mp is now "overkill" as far as capturing resolution goes... you just don't need 16Mp. The ToupTek, at 2Mp, is now able to record output of the objective without leaving much, if anything, behind. In my mind I would greatly prefer to have a camera that is "oversampling" at high magnifications, but is still quite capable at recording the output of a good quality low magnification objectives.
We are just talking resolution "capture" ability. There are also other photographic considerations... noise levels, highlight detail (dynamic range), color quality, raw file capability (a "biggy") and more.
I spent a long time trying to figure out the relationships between sensor size, relay optic magnifications, intermediate image size, resolution capabilities of various objectives. Asked a lot of people, and read whatever I could find. To wrap my head around it I put together a spreadsheet that I still use today. You can find it here if interested:
http://krebsmicro.com/relayDSLR/relay_micro.xls
You need to enter only 3 values, the objective power, the objective NA, and any additional magnification (or reduction) provided by optics between the objective and camera sensor. With a Bayer pattern color sensor you should aim for 3.5 pixels under each resolvable detail (although in certain cases that is unobtainable).
