Quote:
Originally Posted by gregbradley
Sorry Rick.
Here it is:
Correct, but in practice there are limitations. It is dependent on the
charge capacity (full well) of both the pixels in the output shift register
and the output node into the digitizer. When binning, you clock multiple
pixels into a single pixel in the output shift register. If its well depth
is not N * imaging pixel well depth (N is vertical binning factor), then the
pixel can be saturated (clipped), even though the individual pixels binned
into it are not. Same holds for the output node - when multiple horizontal
pixels are clocked into it, clipping can occur. Every chip is different in
this regard - most do NOT have 2x well depth in these registers, so the
bright stuff (>50% full well) will generally be saturated.
Kurt
|
Thanks, Greg. I don't know if that's specific to the KAF8300, but it's what Tandum mentioned earlier. If the CCD readout registers aren't designed to handle the larger values generated by binned data then binning won't have the desired effect. It's not too hard to test with a lightbox, so I'll give it a go with my camera some time soon...