This is the 20th post in a series of Nikon D850 tests. The series starts here.
[I’ve completely revised this test. The first version, through error, compared the vibration from the D850’s electronic shutter to electronic first curtain shutter in live view. Unsurprisingly, it found few differences. I’m now revising the test to compare the vibrations from EFCS and the fully mechanical shutter in live view mode. I’m also changing the methodology.]
There have been a few — very few — complaints about the D850 shutter shock. The camera has lots of ways to deal with that problem if it’s objectionable, but I wanted to see how bad it could be if you didn’t employ any of those. I did mount the camera securely for the test:
- Gitzo Systematic Series 5 three-section legs
- Cognisys rail
- Arca Swiss C1 Cube
I also didn’t use a super long lens, only 85 mm. But I did pick a very sharp one, the Otus 84 mm f/1.4, and I used it at its sharpest aperture, f/2.8. I also used a moderately sharp inkjet printed target backlit with a pair of Westcott LED panels bounced off a white diffuser. ISO was 64, shutter speed 1/200 second. That’s roughly the worst shutter speed for shutter shock on the Sony a7R with this lens and roughly the same mounting arrangement.
I could have used a camera stand instead of a tripod for even more stability, but I thought that most folks would not use the D850 that way.
I focused the lens on the for each of the two runs, one using the live view and the D850’s mechanical shutter, which requires winding the shutter just before each exposure, and one using the electronic shutter, which introduces no vibration at all. I used maximum magnification, peaking, and a Hoodman loupe.
Here’s how the target appeared to the camera:
I focused once and made 32 exposures 3 seconds apart using a Nikon intervalometer, all in live view, first with the mechanical shutter, and then with EFCS.
I’ll just report the blue channel results since they are the best; I must have misfocused a bit.
The vertical axis is the MTF50 in cycles per picture height. I’ve plotted the worst, average (aka mean), and best results, and also the mean plus and minus the standard deviation (aka sigma). The points labeled V are for a vertical edge, and the ones labeled H are for a horizontal one. The difference between the V and H MTF50s is not important, and probably is mostly related to target alignment and how far off-axis the regions of interest were.
I don’t consider the differences significant. I think that EFCS is optional with moderate telephotos and shorter lenses and a good solid mounting arrangement. But why would you not use it?
I will report on other shutter modes shortly.