I looked up the datasheet of the Kodak KAF-5100CE CCD imager chip that Olympus
will use:
Its imaging area is 17.8 by 13.4 mm, or almost exactly one quarter of a 35-mm
frame. There are 2614*1966= 5,139,124 pixels. However, there is a catch:
These pixels are either red, green or blue, so we actually have only 1,713,041
three-color resolution cells here.
The pixels are square, 6.8 microns (0.00027 inchs) on a side.
Let's take a lens with resolution of 60 lines per millimeter. Each such "line"
is in fact a pair of lines, one black, one white. There are 24*36*(2*60)^2=
12,441,600 resolution cells in a 24x36 mm frame. Each such cell has three
colors.
So, a 35-mm film image contains (12441600)/(1713041)= 7.26 times as much
information as does the 5.1 megapixel image. Using Moore's Law, it will take
Log2(7.26)*18= 51.5 months, or 4.3 years for Kodak's CCD images to equal 35-mm
film.
The dynamic range is quoted at about 70 dB, or about 12 stops. The dynamic
range of film is more like 7 or 8 stops. The linearity and large dynamic range
of silicon image sensors is why they have displaced film in astronomy.
Kodak also makes the KAF-16801CE imager, whose sensitive area is square, 36.7
mm on a side, with square pixels 9 micron on a side. There are 4080^2=
16,646,400 pixels, each of one of the three colors, for a total of 5,548,800
three-color cells. This chip likely costs thousands of dollars today, but in
ten years they will be giving them away.
Joe Gwinn
< 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 >
|