Sorry to half hijack the thread but we are on the topic of it and I didn't want to start a new thread for this as it isn't a comprehensive test.
But, I thought I'd do a quick and dirty D4 v D3s ISO 12800 test for the benefit of anyone in this thread who is interested. I've not done any other comparisons yet in the real world, and this was just a quick set up, shoot and be done.
Both camera's locked off on tripod via the lens collar of the Sigma 150 macro (so framing could be kept the same when I switched cameras over).
All settings on both cameras the same, set in manual, including white balance at 4k, even though D3s looks warmer than the D4. Also, despite the same settings the D3s looks a tad under exposed which won't help the noise.
Noise reduction OFF in both cameras and I double checked in Capture NX where I discovered the D4 file had a cheeky little amount set in Capture even though off on camera, so I set that back to zero. Sharpening set to 3 on both cameras. Essentially, these images have identical everything.
I'll supply links to larger jpegs saved at quality of 10 under each image (I uploaded full res to photo bucket but it scales them down). Happy to upload raw files if someone can point me in the direction of a good host for them.
So, nothing scientific, but just a quick play.
Settings used: ISO 12800, 1 second exposure at f2.8
I'm not a betting man, but I'd say although I'd give the nod to the D4, the vast majority would be hard pushed to see a huge difference between these two cameras in general. D4 image is scaled down to match D3s here which helps a little as well. I don't see the point in scaling down but at web size it's pretty irrelevant and as I mentioned above, Photobucket scaled both down when I uploaded them