I am not convinced somehow that CMOS is the holy grail. Canon has
achieved low noise, but I don't think it is necessarily CMOS that is
the reason. FujiFilm has achieved amazingly low noise on a small chip
which is basically CCD as I understand it. Nikon is using a 10MP CCD
and a 12MP CMOS and their noise characteristics seem to be very close
with the advantage to the CCD which you would expect just from the
difference in pixel density. Kodak's own camera used a Belgian Fill
Factory CMOS that had terrible noise characteristics which apparently
was barely usable above base ISO. Potentially the CCD should be
superior since it has a much better fill factor, that is, a larger
proportion of the site is light sensitive compared to the CMOS which
is mostly occupied with circuitry.
Somehow I am reminded of MacPherson struts being invented for a cheap
English Ford instead of a "real" suspension and companies like BMW
turning them into world class suspension design. It is not the
design. It is what the engineer does with it. Canon has done a lot
with that cheap, noisy CMOS design, but I don't think it is
inherently superior.
Winsor
Long Beach, California, USA
On Mar 13, 2006, at 5:14 PM, Chuck Norcutt wrote:
> Kodak struck a development deal with IBM quite some time ago (maybe a
> year?) specifically for CMOS sensors. I haven't heard a singe word
> about progress on that front since the announcement. IBM has been
> building CMOS chips since before Methusala was born. And we all know
> Kodak knows how to do color as in the E-1 sensor. Maybe they'll both
> pop up and surprise us one day soon.
>
> Chuck Norcutt
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|