Re: What is APS?

[P3D Gregory J. Wageman <gjw@xxxxxxxxxx>]

Michael Kersenbrock speculates:

>In summary, given processing-power-on-a-chip say, ten years from now,
>we should be able to generate a digital camera at a somewhat reasonable 
>price (*if* sold in volume) that will do much of the processing, including 
>window manipulation (and such) probably in real time, and in the camera.  And
>in stereo.

I wouldn't be quite so sure of that.  The problem today is not lack of
"processing power".  There are single-chip microcontrollers that are
more powerful than the early PCs, which can easily be embedded in
a camera.  This won't do anything for your basic image quality, which
requires resolution, not merely processing power (remember, "Garbage
in, garbage out" still applies).

No, the high cost of photographic-resolution digital imaging is due to
the laws of physics.

In order to make a digital imager with the resolution of film emulsion,
you need to get the cell size down to near the size of film grain, which
is the size of a large molecule.

The number of photons that will strike a cell of this size during a normal
photographic exposure is quite small.  If the imager is operated at
room temperature, self-noise within the chip (due to the equivalent of
"Brownian Motion" in a semiconductor) tends to completely overwhelm
the miniscule "signal" from those few incident photons.  This is why
the digital imagers used on e.g. telescopes tend to be a) refrigerated,
often with cryogens, and b) operated at extremely long exposure times,
to minimize heat-induced noise and maximize the number of incident
photons.

Refrigerating the imager in a hand-held camera is possible, but requires
either a lot of power (to operate a heat pump, for example), which would
limit portability by tying you to a wall socket, or drastically shorten
battery life, or require a source of cryogenic fluids.  Long exposures
(on the order of minutes) aren't practical, nor are extremely bright
lighting situations, except in some obviously limited applications.

The current crop of CCD-based imagers (Charge Coupled Device) has pretty
much reached its practical limits.  There will probably need to be a
breakthrough in materials technology before digital imaging will achieve
the quality of film with the same or better east-of-use at competitive
cost.

Of course, having said that, I am reminded of a wise man's words:

"When a distinguished and well-respected scientist tells you that something
is possible, he almost certainly right.  But when the same distinguished
and well-respected scientist tells you that something is impossible, he is
almost certainly wrong." :-)

	-Greg (who is neither a scientist, nor particularly well-respected)


------------------------------