With regard to the 400k ISO, IMO the real question should be "where does the ability of the AF to work crap out?" Super high ISO isn't of much use if you can't focus clearly.
I've seen this comment by others also and one answer could be, and this goes for me, it's not about the pitch black use of it, it's about shutter speeds. I often use ISO 5000 outdoor in not too low light to get my 1/1000s-1/4000s shutters. When shooting sports indoors and you want some dof for example, it makes sense to have a very high iso. Have I ever needed above 51k on my 1dx, no, but I like to shoot with the widest of apertures.
Well said. I'm really just curious how the rest of the ISO noise levels compare to the 1DX, which I assume we will see in the coming weeks when the tests start getting posted. It's entirely possible the D4S could pull ahead, even though its normal range still does not extend to 52k like the 1DX's. Time will tell.
I personally could make use of, and have used, ISO 102k with my 6D...but of course the noise is not good, and this was with positive exposure compensation. Not sure it would work at all with no compensation.
But yes, you could not use the D4S for say a 1/5 second exposure at ISO 400k, with a fast aperture lens. Its autofocus wouldn't work, in any case...neither would the 6D's. Although just recently I was able to get my 6D to AF with live view in very dark outdoor conditions, with my Sigma 24mm f/1.8...where the phase AF would not work at all. Obviously this is not a technique that would be useful/reliable for servo tracking, though. But then there are those that think servo AF can't even be done with the 6D in bright noontime.