jimt@xxxxxxxxxxx writes:
<< As you might know, the high-end/professional digicamcoder have
three CCD units with independent RGB filters. Of course there will be some
image degradations whenever you start adding any elements in front of the
sensors. >>
Jim,
I don't know the actual achieved performance of the seperate 3 ccd
designs but in theory they should be capable of significantly higher quantum
efficiency and hence better signal to noise ratio and better low ligh
prformance . The argument goes as follows: A sensor with built in
"integrated" filters on each pixel throws away at least two thirds of the
light at each pixel. (ignoring spaces between sensor elements that makes it
even worse) This is light that is literally not used. (In practise the
filters are not perfect and the losses may be greater.) In a really high
performance three ccd device a special type of prismatic beam splitter is
used (dichroic prism) which is wavelength dependent although this has some
additional losses most of the light at any wavelength should be distributed
to the appropriate sensor keeping the quantum efficiency up (~3X better). In
cheaper designs of 3 CCD devices it may be they split the image with a
simpler design using white light and just use a filter over the CCD instead
of the more complex wavelength dependent beam splitter. In which case the two
systems look much the same in efficiency except for the extra area lost in
inter pixel real estate on the single chip CCD. I don't watch the video
market too much but I do know that for example Sony has used Dichroic beam
splitters on its high end 3ccd video cameras. Maybe somebody on the list has
an idea of the achieved light efficiency of these splitter systems ?
Tim Hughes
Hi100@xxxxxxx
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|