Built for Macro DSLR

[niels123]

How about my situation? high-magnification macrophotography (e.g. 10-50x)? In this case, you will need much more magnification on FF compared to APS-C to have the same FOV. Here, the extra magnification will cause extra diffraction on FF compared to APSC. So APS-C has an advantage then?

Magnification doesn't cause more diffraction.

What is then the reason that images from my Mitotyo 5x NA 0.14 are very sharp, from the 10x NA 0.28 they are still quite sharp, although slightly less sharp and images from the Mitu 20x NA 0.42 clearly show diffraction? I also have the 50x NA 0.55. These photos are never tag-sharp and always soft.

 

[Zeidora]

How about my situation? high-magnification macrophotography (e.g. 10-50x)? In this case, you will need much more magnification on FF compared to APS-C to have the same FOV. Here, the extra magnification will cause extra diffraction on FF compared to APSC. So APS-C has an advantage then?

Magnification doesn't cause more diffraction.

What is then the reason that images from my Mitotyo 5x NA 0.14 are very sharp, from the 10x NA 0.28 they are still quite sharp, although slightly less sharp and images from the Mitu 20x NA 0.42 clearly show diffraction? I also have the 50x NA 0.55. These photos are never tag-sharp and always soft.

Most likely you use a coverslip objective (usually denoted with mag x /0.17 for the 0.17 mm cover slip thickness) for an object without coverslip. At around NA >0.3 using the proper objective becomes important. Then you want to use a true epi objective (denoted with x/infinity symbol). See the photomicrography forum for further details.
The next question is, how well corrected are your Mitus? achromatic, plan, neofluar, planapo? The differences become more pronounced at higher magnification.
Then, for even higher mag objectives, you often adjust for variation in coverslip thickness. I have a Zeiss planapo 40x/0.95 corr, an notice the difference. Images a free of chromatic errors when properly adjusted, this is for histological semi-thin plastic sections. Needless to say, it comes at a price (~US$5K). EDIT: $7K now
Last but not least, also consider the resolution limit of microscopy images. Most modern cameras have more pixels than any lens can resolve. Again, physics. Ideally it is 250 nm, in the real world at best 500 nm (minus NA limitations). There are formulas to help you calculate it. Don't remember them off the top of my head.

 

[scyrene]

Olympus OMD-E-M1 was mentioned earlier in the thread. What could make sense for Canon would be to mimic some of the software features that Olympus has added to both OMD and their high end P&S "Tough" series cameras. Olympus provides ability to select focus stacking in camera. Not as flexible as 3rd party stackers or as good as computer controlled rail moving the body, but very handy in a lot of situations. Select near and far focus points and specify how fine the increment (# in stack), camera does the rest. Seems to be a great way to get people interested and produces much better results than just stopping down way too far. E-M1 actually creates the resulting stacked image in camera as a jpg (limited stack size) or you can output the collection of images in the stack to process outside the camera.

Automated focus stacking is very handy - I've used the HeliconRemote software - but alas it only works with lenses that AF. The crowning jewel of Canon's macro hardware is the MP-E, which is MF only, so in-camera automated focus stacking would not be possible. You'd need a (motorised) focus rail for that.

 

[Zeidora]

Olympus OMD-E-M1 was mentioned earlier in the thread. What could make sense for Canon would be to mimic some of the software features that Olympus has added to both OMD and their high end P&S "Tough" series cameras. Olympus provides ability to select focus stacking in camera. Not as flexible as 3rd party stackers or as good as computer controlled rail moving the body, but very handy in a lot of situations. Select near and far focus points and specify how fine the increment (# in stack), camera does the rest. Seems to be a great way to get people interested and produces much better results than just stopping down way too far. E-M1 actually creates the resulting stacked image in camera as a jpg (limited stack size) or you can output the collection of images in the stack to process outside the camera.

Automated focus stacking is very handy - I've used the HeliconRemote software - but alas it only works with lenses that AF. The crowning jewel of Canon's macro hardware is the MP-E, which is MF only, so in-camera automated focus stacking would not be possible. You'd need a (motorised) focus rail for that.

Alternative for MF lenses is to move the sensor back and forth. That was done back in film days by the Contax AX body. The "AF" was very slow, but it added limited AF to Zeiss lenses. However, it also made the body very thick. It also acted as a built-in extension tube, which was cool. I very much doubt that this will come back, though.

 

[scyrene]

Olympus OMD-E-M1 was mentioned earlier in the thread. What could make sense for Canon would be to mimic some of the software features that Olympus has added to both OMD and their high end P&S "Tough" series cameras. Olympus provides ability to select focus stacking in camera. Not as flexible as 3rd party stackers or as good as computer controlled rail moving the body, but very handy in a lot of situations. Select near and far focus points and specify how fine the increment (# in stack), camera does the rest. Seems to be a great way to get people interested and produces much better results than just stopping down way too far. E-M1 actually creates the resulting stacked image in camera as a jpg (limited stack size) or you can output the collection of images in the stack to process outside the camera.

Automated focus stacking is very handy - I've used the HeliconRemote software - but alas it only works with lenses that AF. The crowning jewel of Canon's macro hardware is the MP-E, which is MF only, so in-camera automated focus stacking would not be possible. You'd need a (motorised) focus rail for that.

Alternative for MF lenses is to move the sensor back and forth. That was done back in film days by the Contax AX body. The "AF" was very slow, but it added limited AF to Zeiss lenses. However, it also made the body very thick. It also acted as a built-in extension tube, which was cool. I very much doubt that this will come back, though.

I thought about that, but they couldn't move the sensor enough, surely? Not without changing the camera body design too much, and even then... Impractical to say the least!