catchlight.. wrote:
Sometimes we get things wrong. This is worth reading Re: Effective shutter speed for "IS"... from DP Reviews Nathan Yan
Image stabilization works at all shutter speed ranges. Don't think of it as something that actually affects your shutter speed, but simply a device that balances out the shaking.
Now, for a long shutter speed, your IS system has to balance it out for longer, which is harder to do. For fast shutter speeds, your IS needs to balance it out for a shorter time, and it may be so short that there won't be blur anyway. So IS primarily helps at slower shutter speeds, although see below.
Under the same idea, when using telephoto lenses with long photo lengths, there tends to be more shake, and the IS system has to work harder to balance the greater shake. With wide lenses, there is less shake, and possibly so little that there won't be blur anyway.
So IS will always work no matter what shutter speed and focal length. However, it's main benefits are for long telephoto lenses and long shutter speeds. Some would argue that it's really not needed for wide lenses or fast shutter speeds (although I personally disagree with this).
Shutter speed "values" are all the same - they're not related to IS. You may be referring to the "effectiveness" of IS systems - they usually measure them in "stops", a 3-stop IS system claims to allow you to shoot at a shutter speed 3 stops slower than you would have without IS, and maintain the same level of blur. What this actually means depends on the individual photographer. Someone who has shaky hands can normally shoot only at 1/1000s without blur (extreme example), so 3 stops would allow him to shoot at 1/125s without blur. Another photographer with rock steady hands would be able to hold his camera steady at 1/8s, so 3 stops allows him to shoot at 1s.
Sometimes we get things wrong. This is worth readi... (
show quote)
Catchlight - I couldn't agree more - we do get things wrong sometimes. And it seems that Nathan Yan falls into that category.
His supporting data for how IS works at all shutter speeds is lacking, and I am being kind.
Clearly, IS response time and it's ability to move things in the lens has a physical limit - sensor sensitivity, inertia of lens moving elements, lag times etc. So logically speaking I doubt that IS is functional at 1/8000 of a second. By the same token I really wonder if it is functional at 1/1000 sec.
But I do not have the equipment or detailed knowledge to back this up.
But Thom Hogan and Nasim Mansurov probably do. And they are not students or academics writing opinions about stuff in 2007 like Mr. Yan.
Here is Hogan's analysis -
http://www.bythom.com/nikon-vr.htm (2010 - he observed that image acuity diminished when stabilization on Nikon cameras was left on at shutter speeds shorter than 1/500)
and Mansurov's
https://photographylife.com/proof-that-vibration-reduction-should-first-be-stabilizedBoth agree that beyond a certain shutter speed not only does keeping stabilization on not help matters, it may actually hurt image sharpness. Hogan suggests that it is an absolute shutter speed limit of 1/500-1/1000, and Mansurov suggests that it is related to the focal length, which takes into consideration image magnification, with the inverse of the focal length used as a practical limit for stabilization effectiveness.
The obvious thing to keep in mind is that as your shutter speed gets shorter, there is less of a need for stabilization, since your keeper rate will naturally improve with the faster shutter speeds - that is unless you drank a six pack of Red Bulls immediately prior to shooting hand-held.
I would say that Mr. Yan's advice is not the best, and certainly not worth reading, other than for historical value.
That being said, here is an image taken with a 600mm lens with IS turned on, hand-held at 1/25 (not a typo). The first is an extreme crop, the second is the uncropped version. I am a fan of stabilization, but it does have it's limits with respect to it's effectiveness, and I would never leave it on if I were using shutter speeds shorter than 1/500 second.