One more comment on the 12mpix vs. 6mpix setting....

I suspect this is similar (or halfway on the way) to using "RAW" mode: that's where you record exactly what the CDD output, without using the camera's on-board noise reductions, interpolation, compression etc. algorithms which the camera automatically applies to the CCD output based on hardcoded and shot-analysis defined settings. You then use desktop software to do a better final analysis and processing of the data and metadata stored with it.

For some sorts of camera work, this might be useful.

But back to the 12 vs. 6 situation:

So, if you use the 12mpix mode, you'll need an offboard processor that knows how to deal properly with the un-interpolated data.

The idea here (and in the RAW processing above) is that the offboard processor (in the desktop software) might be able to do a better job than the firmware in the camera, assuming there was a reason that the firmware on the camera doesn't support the better algorithm, such as processing time, etc.

-brendan