I know the Nikon D3s is about 1-2 stops better, but I wonder how many D700 owners feel the high ISO is really enough in practice. Are there many occasions that you need the high ISO as the D3s? Just wonder if the extra headroom is worth the costs. Thanks.
The advantage is probably closer to one stop than two. However, any improvement in high-ISO performance is welcome. The question for most would be, 'does the additional cost of the D3s justify the benefits?' There is also the additional weight to consider; about another 300gms at least.
For the amateur who likes to take shots of anything that interests him, whatever the lighting conditions, and who doesn't have available a truck-load of lighting equipment to create his owns conditions, and sometimes may not even have a tripod, high-ISO performance is a God-send.
Bear in mind, it's not just noise that is the issue but resolution. A D3s shot at ISO 6400 should be at least as good in every respect as a D700 shot at ISO 3200, and possibly slightly better (assuming same scene, same lighting and same lens).
I rarely use ISO 6400 with my D700, not because I rarely need it, but because image degradation becomes too obvious at that ISO setting.
However, if one examines the conditions where one is likely to need to use a very high ISO for best image quality, it will generally be in situations where a fast shutter speed is required to freeze camera or subject movement, or the lighting is just plain poor in situations where flash is not appropriate (or not allowed) and/or tripod is not available or not allowed, or when a good DoF is required which prohibits the use of a wide aperture.
With the exception of the requirement to freeze subject movement, a good lens with a good image stabilisation system will do the trick.
The D3s costs a lot more than the Nikkor 14-24/2.8, yet such a lens with a VR II image stabilisation system (boasting up to 4 stops advantage - call it a certain 2 stops and a likely 3) would be far more useful with the D700 than the switch to the very expensive D3s. Of course, the combination of the D3s with a VR version of the the Nikkor 14-24 would be something to drool over, provided there was no compromise in ultimate lens images quality, which possibly would be the case with the current state of technology.
I've recently been reading reviews of the Nikkor 16-35/F4 which claims to be the first wide angle lens with image stabilisation. I'm searching for another lens for my D700, but refuse to move backwards. What my D700/14-24 combination cannot do, my Canon gear can do.
I'm disappointed in the ultimate MTF 50% results for this lens, compared with the 14-24. This lens quality issue can affect the usefulness of image stabilisations, but one needs lots of comparison images taken under the same conditions to determine how significant this factor may be.
For example, I notice from Photozone's tests of these two lenses, that the 16-35 VR at F8 and 16mm has worse resolution outcomes, across the entire image, from corner to corner, than the 14-24 at F2.8. At 16mm the 14-24 would be even better.
Now F8 certainly provides a more extensive DoF than F2.8, and if DoF was a major issue, the 16-35 VR would be the choice within its FL range, in poor lighting conditions.
However, if I were photographing a frescoe in a church, where flash and tripod is generally not allowed, I would have to consider whether the 16-35 VR used at ISO 200 and F5.6 or F8 (at a slow shutter speed, or F2.8 at an unnecessarily high shutter speed), produced better results than the 14-24 at F2.8 and ISO 200. You get my drift?
I'll say that Nikon have done a marvelous job in raising high-ISO standards in the DSLR. I just wish they would develop some high quality, portable, VR lenses for the FX format. As Bernard says, the money is buried in the back yard. I'd really like to buy another Nikkor lens to comlement the qualities of the D700 body.