Sensor and lens tilt and how to DIY adjust your equipment
Fred Miranda wrote:
Oh no! I just saw this:
"Update 10 Sep 2020: I’ve just learned from Ilija that the newer Sony Alpha mirrorless cameras (a9, a7R III, a7 III, a9II, a7R IV, a7S III) using 6 screws for the bayonet mount rather than 4, have a different bayonet mount design. Part of the bayonet is held down by the front body case itself, in addition to being held by the screws. It is highly likely that this change prevents these newer cameras from being shimmed at the mount. One would have to take apart nearly the entire camera to reach the sensor and adjust it there instead."
Did you try adjusting your A7R IV Samuel?
I didn't take apart my a7R IV when I had it, so I didn't realise until recently that the mount is not easily removable like in the older generation Sonys. Very annoying! I've mentioned it a while back on this thread. This was what I said:
I've also just learned that all the new Sony cameras with 6-screw bayonets from the a7R III onwards have different shaped front mounts which are not easily removed, and the design most likely prevents shimming because of the way the front panel clamps down on the mount. So those would require full disassembly to reach the sensor and shim there...
I wonder also about the need to adjust the camera, given the bayonet machining issue and that camera tilts are generally small. Maybe the best bet is to select a camera that is quite straight, never adjust the camera and simply shim all of one's lenses to that camera. When you switch cameras, re-shim any lenses that mate differently with the new one. And you don't need to re-shim every lens if you are mostly doing stopped-down landscape photography. Only the special primes for special work, for example. This was actually my plan when I got an a7R IV for a trip that never happened due to the pandemic, before I knew the mount was non-removable.
BTW: To test my A7R IV camera mount/sensor parallelism, I've ordered these items to use with my Leica 135/4 M-mount lens (cheap and sharp wide open):
I still think this is a worthwhile investigation since you are taking to adjusting your lenses now. You don't want any surprises later...
When you do this, you will be able to precisely know how much error is introduced by the camera, the bayonet machining differences between the camera body and lens mount and the lens focus plane skew.
Update: I've measured my Minolta to E-mount adapter for parallelism using a micrometer and all sides are equal
This is quite remarkable, by the way. I assume your micrometer's resolution is 1-micron? or 0.05 mils? All the adapters I've ever measured (not that it was that many) are not uniform in thickness to 1 micron. Even my highly overpriced and superbly well-made Novoflex EOS to E-mount adapter has an 18-micron difference at the highest and lowest points. If you imagine the bayonet face as the face of a clock, don't just measure at the 12, 3, 6 and 9 o'clock positions. I measure at least at every hourly position or in even finer steps, especially between, next to and on the mount screw positions, which can deform the mount plate sometimes.
