Lloyd Chambers has reported that the a7R shutter vibration is worse when the camera is in portrait orientation. This makes sense if the camera is mounted directly to the tripod head, or if the mounting is very close to the camera, but my guess is that for long lenses with tripod collars that the effect is not as significant.
Here’s my reasoning. I have determined that the main vibration of the shutter is up and down when the camera is in the landscape orientation. Tripods are usually better at resisting up and down motion than side to side motion. You can demonstrate this to yourself by pressing on the top of the ball head and pressing on the side. Therefore, if the camera is mounted directly to the ball head, when it’s in landscape mode, the vibration will occur in the direction that the tripod is best able to resist. In portrait mode, the tripod is vibrated in its less effective direction.
With a big lens with a tripod collar, the camera/lens combination vibrates as a torsion pendulum even if the tripod is perfectly resistant to motion. There’s no a priori reason to think that this vibration has a preferred orientation, although it might.
Lots to test here, and the results would apply to many cameras. Modern focal plane shutters move up and down, not side to side, so that they travel shorter distances and can have higher flash synch speeds. That means that they will have the same preferential vibration plane as the a7R. Mirror slap is primarily up and down as well.
One thing to keep in mind when it seems like the sky is falling on the a7R: the contribution of camera shake of the a7R live view pre-exposure shutter closure is about a quarter of that of mirror slap on the D800E. The difference id that, in the D800E there are ways to separate the mirror-up operation and the shutter opening by enough time to let the vibrations from the mirror slap to die down, while there is no such mechanism availabe to the a7R owner.
That’s not saying there won’t be. I think it should be possible to introduce firmware workarounds.
One possibility would be to add a drive mode. Let’s call it Single Shot Delayed Shutter. It could be accessed by pressing the Fn key like all the other drive modes. When the camera was in that mode, it would operate similarly to a Nikon D800 when the dial on the top left of the camera is turned to the Mup position. With the camera in live view, the first press of the shutter release or trigger from a remote release would close the shutter. The display would go black. The second press would cause the shutter to open, make the exposure, close, and open again to enable live view.
The above scheme would have teh advantage that it would be familiar to people who have experience with modern Nikon and Canon DSLRs. A method that I would prefer borrows from H-series Hasselblads. In this approach, there would be an option in the set-up menus to assign one of the programmable buttons on the a7R to control the live view shutter position. The button would operate as a toggle. The normal mode would be the same as it is today. The shutter is normally open, and pressing the shutter release causes it to close, open, close, and open again to enable live view. The delayed mode, accessed by pressing the assigned button, would close the shutter and disable live view. Pressing the shutter release would cause the shutter to open and close, and it would stay closed. The finder would remain dark. Further presses of the shutter release would do the same thing. Another press of the assigned button would restore normal operation.
lloyd specifically noted the use of a tripod collar with the a7r. for the portrait orientation vibration tests.
you might read his articles on it; enlightening to a precise degree on this issue. for anyone involved with still life work with lenses longer than 85mm, this is a challenge.
i like the idea of a firmware fix for the function button, but as a h/w engineer i think to actualize such a solution would require sending the cameras back to Sony…not necessarily a problem.
Joseph Holmes says
Your conclusions are entirely correct from my testing over the last few weeks, which I also shared with Lloyd in some degree, prior to his series of experiments. One can indeed get a very good idea of what much of the basic pattern of shutter-induced image blur will be simply by touching and looking at the camera on a tripod in various setups. Coupled with the understanding that most of the shake is in the dimension of the shutter’s motion, it’s easy to see that when the camera is horizontal and clamped directly to the top of the tripod, the shake will be very effectively stopped, whereas a vertical shot will move much, more, and so on. Indeed when using a lens foot, the shake is far more uniform with respect to horizontal vs. vertical camera position.
Beyond your assessment of the magnitude of the shake from a single motion of the first curtain of the Sony, compared with the mirror slap of a D800E, it’s important to consider that the although the motion of the first curtain of the Sony’s shutter may shake about as much as that of a Nikon or Canon shutter’s first curtain, and the Sony’s shutter now has to move that curtain twice, for roughly double the shake, there is also the problem of the Sony being half the weight. Carefully using my own sense of touch, I estimated the A7R’s first curtain camera motion to be triple that of my 5D Mark II. The idea to use the accelerometers of an iPhone to measure the shake is good, but as you have written, the phone must be very positively connected to the camera for such measurements to be of decent quality, and thus far I’ve not devised a firm metal clamp setup to accomplish that, nor do I have the time to pursue that further.
So the Sony demands a lot more from its first curtain, with respect to avoiding shake, than those cameras with which we have been most familiar (as I gather you also found with a Leica M), and instead, because it’s a live view camera, it ironically has to move twice. I suspect that Sony already tried to minimize the A7R first curtain shake, by reducing the flash sync speed (and with it the curtain travel times) from 1/250th to 1/160th, compared with the A7, knowing that the A7R would not get relief from electronic first curtain.
At least a firmly connected damping weight in a good mass range (enough to work well but not enough to stress out the lens mount, which seems to be a mass close to 24 ounces) can come to the rescue, but not without a certain amount of hassle. The vertical grip with two batteries in it weighs 13 ounces, which will only reduce the shake somewhat. If a 4-ounce L-bracket were added, the resulting 17 ounces would be better, but I think still less than optimal. I will experiment further once my clamped-on deadweight is ready for use and report on my site, at the URL below, and likely elsewhere. Its weight will be about 23 ounces including the L-plate, the clamp and the deadweight metal block itself.
Thanks for all your great experiments and sharing!