Since I had the 400 set up, I ran the previous tests using the D800E instead of the Leica M240.
Horizontal camera orientation, shutter tripped remotely, no shutter delay:
Horizontal camera orientation, shutter tripped remotely, 3 second shutter delay:
Vertical camera orientation, shutter tripped remotely, no shutter delay:
Vertical camera orientation, shutter tripped remotely, 3 second shutter delay:
And, just for completeness, horizontal camera orientation, live view on, shutter tripped remotely, 3 second shutter delay:
Note that we don’t see the beginning of the sinusoidal mirror-slap-induced vibration. It starts before the shutter opens. It looks like the visible trace begins about 30 degrees, or 80 milliseconds, after the mirror sets off the vibration.
Without the shutter delay, the vibration is awful in the vertical direction (camera orientation horizontal). It makes sense that most of the mirror-slap-induced vibration is vertical, since the mirror motion is upwards. The vertical 10 Hz vibration is close to 20 pixels (remember, these are D800E pixels, closer together than the M240 pixels) peak to peak, and showing no sign of damping out in the half-second exposure. The horizontal vibration is four or five pixels peak to peak, the same 10 Hz frequency, and appears to be mostly damped out by the end of the 1/2 second exposure. It’s pretty hard to see this, since the amplitude of the vibration is about the same as the diameter of the focused spot.
There is some shutter opening vibration on the shutter-delayed images with the camera in a horizontal orientation (or possibly residual mirror-slap that didn’t damp out in three seconds — I’ll have to test some more), but it’s not too big. Live view is no worse than regular mode, because of the three second delay.
So, compared to mirror slap, the Leica M240 live view errors look pretty small. I wish I could back up further so the spot size would be smaller. I’m already 30+ feet away. To get much further I need to go outside and do it at night.