Hi,
My guess is that the Alpha 900 was limited by it's 12 bit processing pipe line. The impression I have is that there is a balance between readout times and noise. Very clearly, the Nikon D3X was much better than anything else regarding DR, and probably also on resolution, due to thinner OLP filter, and possible better lenses.
Regarding noise, there seem to be little to do about shot noise. It's proportional to the square root the photons captured. So shot noise is essentially limited by full well capacity, the amount of electrons each pixel can hold. This quantity seems to go down with pixel size. The noise floor will also be limited by readout noise, coming from electronics. Great progress have been made in reducing readout noise with the Sony sensors. Nikon and Pentax seem to make better use of the sensor than Sony themselves.
A bigger sensor will collect more photons, and will therefore have an advantage.
Best regards
Erik
Yes, you will get lower shadow noise (also known as dynamic range) with both the Sony and Nikon, even less with the Nikon thanks to the way Nikon has been able to process signal information.
The interesting thing is that Sony semi-conductor and Sony Imaging are two different companies. Nikon has been a much larger customer of Sony semi-conducator than Sony imaging is. Fun isn't it? :-)
The gap in terms of shadow noise was already very large between the D3x and A900 (even if it was never acknowledged by some folks). To give you an idea, that gap was larger than the gap seen between MFDB and the best DSLRs. It seems Nikon did it again with the D7000. Pentax also did a very good job using that same sensor by the way.
I would think that Sony Imaging finally got it right with the A77 sensor, but we'll have to wait for actual DR measurements to find out.
Cheers,
Bernard
