On 28 May 2008 07:56:10 GMT, Chris Malcolm
wrote:

John O'Flaherty wrote:
On Tue, 27 May 2008 19:53:23 -0700, Voodoo Thunder Pig
wrote:

RichA wrote:

from Amateur Photography magazine;

The camera has a double Gauss lens system, a design commonly used in
35mm cameras,' explains the space agency.
'Images are recorded by a charge-coupled device (CCD) similar to those
in consumer digital cameras. The instrument includes sets of red,
green and blue light-emitting diodes (LEDs) for illuminating the
target area.'

Nasa claims that the camera can focus down to 11mm and record images
at a resolution of '23 microns per pixel' at the closest focusing
distance - allowing the camera to show details 'much finer than the
width of a human hair'.



This means tri-colour imaging through three different filters onto a
monochrome CCD.

Nice try, Captain Obvious, but that's not what it means. It means the light
sources providethe three colors, and no filters are necessary.

Do you mean that the different colors of LEDs are on at different
times, providing color separation? I wonder if they might also use
filters on the LEDs, though, to increase spectral purity of the
colors.

Most unlikely. The problem with LEDs as light sources for photography
is that they're too spectrally pure.

I didn't know that. There are more details and a picture of the
robotic arm camera and its light source here -
http://www.nasa.gov/mission_pages/ph...craft/rac.html
It mentions a 1:1 magnification with resulting resolution of 23
um/pixel, which doesn't seem very fine (cf. 2-3 um for a p&s or 6-7 um
for an SLR). Maybe the light output of the LEDs is so limited that
they need larger pixels to get a clean image; or, since the camera
probably had to be designed at least a few years ago to be ready for
the mission, it may be older technology.

There's also a "surface stereo imager" camera
http://www.nasa.gov/mission_pages/ph...craft/ssi.html
that has a 1000*1000 sensor. That one does use filters, "... but SSI
exceeds the capabilities of the human eye by using optical and
infrared filters, allowing multispectral imaging at 12 wavelengths of
geological interest and atmospheric interest."
I suppose that that one depends on sunlight, though, which is
broadband to start with.

Then there's at least one more camera, the microscope camera,
http://www.nasa.gov/mission_pages/ph...raft/meca.html
which resolves 4 um/pixel, with a field of view of 1*2mm (so about
250*500 pixels?). That one uses R,G,B, and UV LEDs.

--
John