This comparison popped up a few days ago, I trust that their method was sufficiently scientific such that these values are representative of reality.
I own a D700, which has the same sensor as the original D3 - I’m not sure why there are data points all over the place between whole-stop points.
What I’m reading here is basically: an immediate 3-stop advantage at the D700’s base ISO, closing to 2-stops as you go up the scale.
Or in more practical terms:
- ISO 200 on the D700 is great, there’s a tiny bit of luma noise to cleanup in Lightroom
- ISO 100 on the D700 is, to my eyes, flawless. There is no noise to cleanup in Lightroom, you lose sharpness instead
- ISO 100 on the D700 allegedly comes at the cost of slightly reduced dynamic range, it’s a bit of a hack
- You can shoot ISO 1600 on the D4 and have only a tiny bit of luma noise to cleanup in Lightroom
For convenience, one of the comparison graphs (it’ll look weird due to alpha transparency):
Caption: ISO values are full stops, starting at ISO 100 on the left side. You can see that most models’ base ISO is 200.