I just read the February 2003 issue of Outdoor Photographer.
On page 23 is a full-page ad for the new Sigma SD9 camera, the one based on
the Foveon chip. The ad claims "over 10.2 million photodetectors, to sense
red, green, and blue light at each pixel, unlike image sensors that process
only one color per pixel". From the careful distinction drawn between
"photdetectors" and "pixels", I would guess that 10.2 is the marketing pixel
count, and that there are 10.2/3= 3.4 million tricolor pixels per image, with a
1:1:1 ratio. The camera weighs 805 grams (28.4 oz, 1.76 pounds). No price is
stated in the ad. Camera also reviewed in DP Review:
<http://www.dpreview.com/reviews/sigmasd9/>, where the price is reported as US
$2,000 for the body only. From an earlier DP Review: "This new camera has a
3.43 megapixel (effective) X3 sensor which outputs 2268 x 1512 x 3 pixels. Each
pixel is 9 microns [square] which makes the sensor 20.7 x 13.8 mm (a 1.7x focal
length multiplier)." So, there are 3,429,216 tricolor pixels, and the same
number (3!
.43 million) of green pixels.
On page 72 starts a 1.5-page review for the coming Kod*k DCS Pro 14n camera,
which takes Nik*on lenses, and sports a 13.89-megapixel (marketing) sensor
yielding a 41 MByte file per image. The sensor is 4536x3024 CMOS (not CCD), is
used as effective ISO speeds ranging from 80 to 800, the camera weighs 2 pounds
(910 grams), and costs about US $4,000 (estimated) for the body only. This is
apparently intended as a studio camera. The Pro14n has 13.89/2= 6.95 million
green pixels. The ratio is 1:2:1. <www.kodak.com/go/professional>
On page 73 starts a 1.5-page review of the Can*on EOS-1Ds, which claims a 11.4
megapixel (marketing) CMOS image sensor, being 4064x2704 pixels, ISO
sensitivities of 50 or 100-1250, weighs 2.76 pounds (44.1 ounces, 1.25 kg), and
costs $9,000 (estimated) for the body only. Another studio camera. The
EOS-1Ds has 11.4/2= 5.7 million green pixels. The ratio is 1:2:1.
<www.usa.canon.com>
I can see these two studio cameras being very useful to commercial
photographers who already own Nik*on or Can*n lens collections, and these
prices will not deflect such photographers. The Sigma has one half the spatial
resolution, and Sigma cameras are not widely used by pros. One would expect
the Kod*k camera to cause downward price pressure on the Can*n. Ordinarily,
the high price of the Can*n would cause the Kod*k's price to rise as well, but
Kod*k is trying to quickly build a business in digital photography, in time to
replace the film business.
It's interesting that CMOS sensors are used, not CCDs. CMOS and CCD are
different kinds of integrated chip designs. Semiconductor memory chips are
CMOS, and CMOS optical sensors are based on memory chips. The difference is
that while CMOS sensors are easier to make in large physical chip and
pixel-count sizes, their optical performance (especially noise) traditionally
isn't nearly as good as CCDs. Apparently, the art of making CMOS sensors has
progressed to the point of practicality in photography, although CCDs still
reign in scientific applications.
Applying Moore's Law to the EOS-1Ds body, this will drop from $9K to $1K in
18*log2(9/1)= 57 months, or 4.76 years. The Pro14n body will drop to $1K in
18*log2(4/1)= 36 months, or 3 years. The SD9 will take 18 months.
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 >
|