awpaeth@watcgl.UUCP (Alan W. Paeth) (12/09/85)
Here's one for all you NTSC and Video gurus.
Scenario:
Our lab (U of Waterloo Computer Graphics Lab) had been looking at a high-
quality JVC color camera for both framebuffer digitization and conventional
video use. Alas, high costs: the JVC model featuring both NTSC and RGB
output (and the latter was NOT reconstituted RGB from a comb filter, but was
taken via a backplane adaptor) was in the >$15,000 catagory. If anyone knows
of an quality, inexpensive color camera with non-composite RGB, let me know,
otherwise, read on.
Question:
Are there commercial b/w camera available at a reasonable cost? It occurred
to me that with a good set of color filters, one could make do by digitizing
color separations from a copy stand. Ideally, the camera would provide non-
composite video, and resolve about 600 lines. Linear voltage out (gamma=1.0)
would also be quite attractive, as we have linear ADCs on our digitizer. My
sad suspicion is that b/w cameras are as rare as b/w photolabs: no consumer
demands implies few products.
Some other concerns: we can digitize composite b/w video (because the Y of YIQ
is baseband, and sync pulses just generate minimal beginning of line cruft).
However, do such cameras roll off at 2-3 MHz to limit the luminance information?
I'd rather they didn't, has our ADCs are fast. If they do roll off to conform
to spec, the camera might see more "conventional" video use. Then my concern
is of full conformance to NTSC. Assuming a b/w camera doesn't add colorburst
type circuitry, I wonder if all state-of-the-art video hardware recognises
the "color-kill" concept. For instance, can off-the-shelf mixers deal with
simultaneous b/w (no burst) and color signals as input?
Are there any recommended pickup tube types? The old Sony b/w cameras I
remember had a significant bright light hysteresis, which could even appear as
edge smearing on contrasty subjects. If tubes with an intrinsic gamma of 1.0
exist, these might be of particular interest.
Replies can be mailed or posted, based on relevence to the net at large.
As should be obvious, I'm as much interested in finding the right questions as
getting answers (shades of "Jeopardy"?). The overall objective is to put
a very fast color digitizer board into action at minimal expense.
/Alan Paeth University of Waterloo Ontario CANADA