I am not sure where you are going with this. I visited the Foveon site and I really don't see evidence of them making any outlandish or misrepresentative claims.
The 10.5 million effective photodetectors is by any technical measure completely, unequivocally legitmate. That IS how many sensors are in the imaging area.
They also state the array size as 2268 x 1512 x 3, I am not sure how you would suggest stating it, but it seems quite clear to me. Foveon is not making any riduculous statements, unlike Fuji was with their Super CCD, where they interpolate even more pixels than do standard Bayer CFA imagers do.
Lets take a step back for a moment though, in a standard Bayer CFA area, if you take an area of 2 pixels vertical by 2 pixels horizontal, the sensor will have 2 green pixels, 1 blue and one red arranged as such:
RG
GB
whereas the Foveon chip will have 4 red 4 blue and 4 green sensors. So in red and blue Foveon has 4 times the sensors as other typical sensors. In red and blue Foveon could claim 14MP "equivalent capability". In green they have twice the number of sensors so perhaps they could claim 7MP "equivalent capabiltiy" perhaps you would prefer that.
According to the Nyquist theorem the Foveon chip should have substantially better resolution performance particularly if looking at pure red green or blue. The only problem is the theorem only really predicts ultimate resolution for high contrast images in black and white. An image formed from bayer pattern imager plays a little havoc with the theorm. The simple fact of the matter is 3 times the information is being captured by the Foveon chip than a standard 3.5 MP camera.
I suspect that you will find the resolution performance of the Foveon chip quite impressive. If you download any of the resolution charts that Phil shoots and magnify them 800 to 1600% you can clearly see the effects of pixel interpolation. On one side of a black line the pixels are slightly green-grey, the other side red-grey. As you get to the limits of resolution the pixels blend together and there are often strange color effects in this area as the algorithm mis-interprets the data. These effects should be absent in the Foveon chip.
Though a resolution chart might not show resolution being much better than perhaps a 6MP sensor when measuring a high contrast B+W image, the performance is likely to be much better than a 6MP sensor in lower contast conditions. Unfortunately for Foveon, there is yet to be a 10MP sensor to which it can compare itself to.
John Bower
I agree. We need a way to describe this new technology that will
give it credit where it deserves it. I feel that calling it a 10mp
sensor is not the correct way to do so. Calling the X3 a 3.5mp x 3
camera, makes it seem like there are 3 chips inside, I guess it
could be an accurate way to describe the technology. But I just
hope for a better description.
