105 AFS VR Micro!

[david_lau3]

"With the enhanced VR system (VRII), photographers can capture sharp images at shutter speeds approximately 4 stops* slower [at near infinity to 3m (1/30x reproduction ratio)] than would otherwise be possible."

 

The above is extracted from dpreview news.

 

It appear that VR is only useful up to 1/30x? Does this mean that from 1/30x to 1:1 it has no VR?

[vivek iyer]

I think I will buy one of these along with a D201.

[mark_chappell]

<I> If the photographer is moving the camera slowly at a constant pace, the system won't

fix that because it has no way of detecting the constant motion. </i><P>

 

I don't understand your point in the real world of macrophotography. You're correct that a

constant velocity and direction of motion doesn't involve acceleration. However, if the

camera is moving with constant velocity, you (the macro-photographer) are going to

notice this very quickly and correct the aim, distance, or whatever. That 'correction' is an

acceleration. If you don't correct for constant-velocity motion, you will see the subject

drifting at a constant rate across the viewfinder as the camera moves in relation to the

subject. <P>

 

I can only see your assumptions as relevant in a situation in which <I>both</i> the

subject and camera are in constant, identical motion with respect to some external frame

of reference -- for example, if you are tracking an insect walking at constant velocity. At

least in my experience, that's a rarity in macro work. Typically one is attempting to hold

the camera still in relation to a stationary target, and that 'holding still' involves lots of tiny

feedback movements from the viewfinder image to the hands that are aiming the camera.

Those little movements are accelerations that could -- theoretically -- be damped by an

effective VR or IS system.

[ilkka_nissila]

The 1/30x limit just means the test result of 4 stops is applicable at long distances, I assume with short distances you may get a worse results because of what we have been discussing.

 

Mark, the problem is this: in order to correct for vibrations, the system (assuming it tries to correct for translational motion as well as rotational) has to determine the zero velocity point. This can be done by averaging over a longer time. But since this has to take place prior to the exposure, the photographer can't really know the precise time at which the VR system resets itself. So if the photographer is moving the camera laterally during this period when the lens fixes its point of reference, when the picture is taken, the lens will try to make the projected image continue that constant motion! Thus producing a blurry result since the subject is still. So the photographer has to take care not to move the camera laterally just before the shot is taken unless following a moving subject laterally.

 

Nikon may apply clever algorithms to determine where to fix the zero-point of velocity but I don't think it will work as well as traditional VR. This is supported by Nikon's announcement where it is noted that the VR system gives 4 stops from 1/30x to infinity.

 

I should look up the US patent to see what it says. Not that it's a guarantee that Nikon hasn't developed the technology further since then, they probably have.

[steffen_kluge1]

"The world's first macro lens offering reproduction ratios of up to 1:1" - what are they trying to say by that?

[steffen_kluge1]

Never mind, looks like letsgodigital.com jumbled up the features list. The original Nikon announcement has got it right.

[Mike D]

Does anyone know it this type of VR means that VR can be used when the camera is mounted on a tripod. I use a 200-400 VR and it is OK to use VR when mounted on a tripod. With 80-400 VR, it is not OK to use VR when mounted on a tripod.

[oceanphysics]

<i>The camera, not the lens, measures the perturbation and determines the amount of correction to apply, and the VR mechanism in the lens implements the correction. It's all relative - a matter of degree rather than kind!</i>

<p>

Edward, you simply have no idea what you're talking about. The camera doesn't measure jack, the lens does. And what is measures is rotation, not translation. Ever. Read the freaking reference I posted on how the system works, since you obviously don't have a clue.

<p>

The discussion about translational motion versus translational acceleration are completely irrelevant. The system has no means to detect either one.

<p>

As I predicted, Nikon's own materials now essentially admit that the VR only works well for non-macro work. Big surprise.

<p>

Once again, I'm left wondering why I bother.