Jan Wegener in his field test of the Canon R7 (on YouTube) commented on the camera's shutter shock with mechanical shutter, including an aside that it seemed to even occur with the electronic first-curtain shutter (EFCS). This has caused some commentary in online forums.
On a slow day, I decided to do a few tests with my R7 to see what the impact of the different shutter settings is on image quality in a relatively controlled environement. To do this I mounted the camera on a tripod, using my EF 100-400 II lens and adapter, and then shot off a series of images of a still text-based target using different shutter modes and speeds. Note the lens was attached by the foot; while not ideal for damping, it was the only option for this particular tripod. Image stabilisation was off, and I used the 10s timer to eliminate any movement from the shutter press. Images were viewed in Canon DPP4.
Please note this was just a casual test for the fun of it. I don't claim to have a proper studio set up, although I tried to eliminate as many variables as I could think of.
In the first test I shot three series of single images using mechanical shutter (MS), EFCS and then electronic shutter (ES) at a shutter speed of 1/200s and other exposure settings identical. At this speed, the MS was clearly softer than the other modes, in every image. The ES was the sharpest, with best contrast of the letters on the target. The EFCS was just a tiny bit softer, with slightly lower contrast than the ES, but I had to peer quite closely at the images at 100 percent to see the difference. This would be unlikely to be noticeable in normal use. These differences essentially disappeared when sharpening was added in DPP.
Online comments on mirrorless camera shutter shock suggest it occurs in a limited range of speeds from about 1/25s through to about 1/200s. I shot a series of images ranging from 1/25s through to 1/400 sec with MS. To my eye, at 1/25s the image quality was quite sharp. Increasing shutter speeds up to about 1/250s showed increased softness. At 1/400s the image was very sharp, suggesting shutter shock is no longer an issue at this speed.
While EFCS pretty much eliminated softness on single shots, I was interested whether, in a series of image shot at 15 frames per second EFCS, there would be increased softness due to the rapid repeat movement of the second shutter curtain. I tried this a few times at 1/200s but I didn't really get any consistent results that could lead me to any firm conclusion with my set up. However, at this shutter speed, a sequence of images shot with the MS was still softer on average than the sequences shot with EFCS.
Obviously these results come with some caveats due to the limited testing method, but they do seem to show the shutter shock effect associated with MS mode at slowish shutter speeds (in the range of about 1/100-1/200s) at least with my static camera+lens+tripod set up. How much this would affect real life photos is open to question, given other variables when shooting moving targets at 15 frames per second. Any vibration effect seems to be gone by 1/400s, and EFCS seems to solve the problem.
Simon