h a l f b a k e r yLeft for Bread
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Consider slide film. The colors of the
scene are (in the ideal) exactly
represented
in the developed frame of film. As a
result, it is necessary to get the exposure
time bang on, or the colors will be
screwy.
Now consider negative film. With
negative
film you have a wider latitude
in
exposure,
because you can correct for over or
under
exposure when you make the optical
print.
But negative film still has to roughly
satisfy the constraint that it must be a
negative image of the scene, because it is
made to be printed optically.
What if you
removed that constraint? Why not let
each
color layer be "represented" by whatever
color allows you to have the sharpest
possible grains? The colors in the
processed film would have no relation at
all to the colors in the scene, other than
that they would be distinguishable from
each other, and they would each
represent
one of three points within the visible
spectrum, which could also be chosen
somewhat arbitrarily, to maximize
gamut.
You would not be able to print optically,
or
even really read the processed film with
your eyes, but as long as the scanner can
tell each color layer apart, then you are
good to go. Each kind of false color film
would have a scanner profile, and the
software would convert it into the correct
colors.
Why? Well, there are affordable
film
scanners that already exceed the
resolution of available negative film by a
factor of two. So by eliminating some of
the constraints on film, maybe it is
possible to utilize some of that wasted
scanner resolution.
[link]
|
| |
Nice. I'm not sure if there is much space
for fine-tuning without locking yourself
into a custom system. Films, negative
and positive, as well as scanners,
screens and printers have been
optimised around the colour sensitivity
of human eyes. |
|
| |
If you can gain some resolution or
contrast by using a false colour film you
then have to find a scanner that is able
to distinguish those colours without
losing the advantages that you have just
gained in contrast. A solution would be
to build a spectrometer scanner so that
each pixel is scanned and stored not as
an intensity of red, green and blue, but
as an intensity of the gamut of
wavelengths present. This would give
you a 4 dimensional data file
(x,y,wavelength, intensity) which you
could then distill back to three primary
colours or use as is. |
|
| |
A bitmap of this type (if used as is)
could create a
whole raft of intesting possibilities and
associated problems. [+] Interesting
idea. |
|
| |
Unfortunately, you're still using film as an intermediary, and that technology is going the way of the Brownie box camera. So I'm afraid this is one of those day-late ideas. |
|
| |
[DrCurry] Until we have a breakthrough
in fabrication of large ICs, the quality of
digital photography will remain only
marginally as good as medium format,
and nowhere near that of large format.
This is because you can only make the
pixels so small before the (limited by
physics) ability of a lens to resolve
detail is the limiting factor in your
overall resolution. |
|
| |
The only way, ultimately, to improve the
resolution of digital photography in a
big way, is to make the sensor bigger.
The problem is that the cost of an IC is
proportional to the square of its size, so
an 8 by 10 inch, or even a 4 by 5 inch
CCD sensor is out of the question. |
|
| |
People look at the latest medium format
DSLR's, which have sensors larger than
24 x 36 mm, but smaller than 45 x 60
mm, and see that they are very
expensive. But people are conditioned
to expect that computer related gadgets
like that always start out "very
expensive" but within a couple of years
become affordable. The problem here is
that we have no way of fabricating
sensors that big in an affordable way,
so unlike CD burners, which used to
cost $10,000, but now cost $10.00, one
should not expect a digital camera with
a large sensor to be affordable any time
soon. |
|
| |
The fundamental thing here is that
digital photography has to have ICs that
are constrained in size by the
wavelengths of visible light, they cannot
be miniaturized like every other IC, and
it is that miniaturization that has driven
the cost reductions of every other
digital gadget out there. Indeed, the
cost reductions for consumer orientated
digital cameras has been driven by
making smaller and smaller sensors,
with smaller and smaller pixels, and
tests have proven that a tiny 8 Mpixel
sensor is nowhere near as good as a
"full frame" 8 Mpixel sensor, due to the
limits of the lens. |
|
| |
JP: but try telling all that to a venture capitalist. |
|
| |
I see your point, DrC. Few backers would
be behind older technology when the bulk
of the market has gone to a new medium. |
|
| |
But that doesn't stop it being an
interesting idea, even if its application
would be to a niche market that couldn't
fund its development. |
|
| |