To VR or not to VR...

[ac12]

The car in the background is in a different part of the frame, as it would be while you panned with the subject, so I would go with double exposure if asked, and given no other information. Don't really know how it would do that to you, though.

The two exposures of the subject are reasonably sharp, no motion blur except what was intended for the wheels, and there's no blur between the two positions, so although I don't know how quickly the IS system adjusts its elements, it would have to be DANG quick to achieve that!

I can see what seems like a VERY LIGHT blur between the two images

Obviously it happened.
But, it is still hard to understand how that happened.

Rather than a double exposure.
I was thinking about the sequence and timing on a 1/500 sec exposure:

Shutter opens
For the first 2/5 of the exposure, 2/2500 sec, before the IS element move = lower image
For the next 1/5 of the exposure, 1/2500 sec, the IS elements move,
Then the last 2/5 of the exposure, 2/2500 sec, the IS elements is in the final position = upper image
Then the shutter closes.

That presumes that the IS elements move to ONE position, during the exposure, rather than multiple positions.
On a 1/500 sec exposure, that is possible.


What is puzzling is, I shoot football at 1/800 at night and 1/1600 during the day, with my Nikon 70-200 and VR/IS on, and have never seen that double image effect.

I did however, have a situation where the image would jump around in the viewfinder. It was so bad, that I could not shoot. So I do not have any images of the effect of that.
I sent the lens to Nikon, and they replaced the VR module. $$$
The lens has behaved ever since.

So "maybe" the IS module or mechanism is going bad.

 

[RacePhoto]

Correct I don't know how this happened, and if I have ever seen anything like this again, I wouldn't know, because I turned off IS? Also that lens has a panning mode for IS which doesn't correct left right, but is supposed to correct up and down movement. (If I understand that right?)

"Mode 2 is intended especially for panning. It is designed to ignore motion in the direction of the panning, but to correct any detected motion at right angles to the panning direction."

And:
"When IS is switched on and the shutter button is partially depressed, the stabiliser lens group, which is locked in a central position when not active, is released. Then two gyro sensors start up and detect the speed and angle of any camera movement. The detection data is passed to a microcomputer which analyses it and prepares an instruction for the special stabiliser lens group. This instruction is transmitted to the stabiliser lens group which moves at an appropriate speed and angle to counteract the camera movement.
This complete sequence is repeated continuously so that there is an instant reaction to any change in the amount or direction of the camera shake. It takes about one second from the moment you partially depress the shutter button for the stabilisation to become really effective. The stabilisation action continues for about a second after you take your finger off the shutter release."

It might have been a focusing error, as I originally suspected, or IS, or something else. It's always been a mystery, and my answer is a guess and based on something long ago that someone said, it was the IS.

Could have been I lurched as I pressed the button and the IS compensated in real time?

I have no science or confirmation of what caused it.

I'd be able to accept it's something else and my conclusion is more superstition or a guess, than fact? What would be very convincing is someone who knows the real answer?

 

[RVT1K]

Since you can never hold your camera perfectly aligned with your subject, the image will appear to jump around as you describe.

If my lens has VR, I use it and have yet to encounter a situation where I though it was detrimental in either nature or motorsports photography.

This is a motorsports example..VR on and panning.