DCML Digest, Vol 26, Issue 24
bangfish at cableone.net
Fri Apr 22 17:10:18 CEST 2005
Here's some clarification for you:
> There were four different cels for each comic book page: one for cyan,
> one for magenta, one for yellow and one for black.
> Probably all four
> looked black and white and the dot screens used to determine the
> areas also consisted of black dots of varying size and distance.
Again, correct...but only when talking about the film negative stage.
For comics they used conventional dot screens, where the dots were
spaced evenly as measured from their centers, according to their
resolution (lpi, dpi, or whatever other terms for resolution you might
care to dig up). In fact, on screened films, solid areas are actually
dots so large that they meld together.
> Do you know how many different types of dot cels there were?
> With only one type (and three colors) one would get a total of 2^3=8
> different colors in print.
> With two different one would be able to print 3^3=27
> With four different dot screens you could print 5^3=216 different
> colors. Which seems pretty much the number of colors available back
The size of the dot determined the amount of that color - alone or in
combination with the other two - that appeared at the final printed
stage. It was very expensive, by traditional process camera methods, to
"call out" dots at other than a very limited number of sizes (expressed
as percentages of coverage), which severely limited the "visual"
pallette of colors you saw in comics. And you could not, in ordinary
circumstances, to blends or fades of any kind.
> So to color a comic book you needed a photostat of the inked (but
> uncolored) artwork, a knife and four types of dot screens.
Depends on what stage of coloring your discussing. A colorist was
someone who took photocopies - at Gladstone this was off of a standard
office copier - of the art and painted on them directly with dyes that
would approximate the printed color, and then marked these up to
indicate the color values - that is, the dot sizes of cyan, magenta,
and/or yellow - that should be used for each colored area. (There were
other methods, but this was the most common.)
The color house did need a photostat of the inked, uncolored artwork -
the base art. Over this would be applied three clear film overlays, one
each to represent cyan, magenta, and yellow. These overlays would, in
the case of Chemical Color, then be painted in various gray tones to
represent the prescribed dot sizes of these three colors. These four
elements would then be disassembled and put, one at a time, on the
process camera to create four film negatives - an unscreened one for
the base art, and three screened ones for the cyan, magenta, and
> The main task
> probably was to prevent the dots on the three color cels from
> overlapping (or at least from overlapping completely). Maybe the dots
> the screens were not spaced regularly but at individually randomly
> varying distances?
Actually, overlap was not an issue - screen angle was. And in
conventional printing it still is. You have to angle the screens -
mechanically or electronically - so that the CMY creates a proper
"rosette" on the final printed page. A good dot "rosette" is the most
visually effective way of creating all the colors of the rainbow on a
printed page using only the CMY primaries. Computers have greatly eased
the complications of this, but it's still no walk in the park to get it
> Since the end result could not be seen before the issue was printed the
> results were probably not always as expected.
In fact, it could be seen before printing, but only by the costly
method of color keys - base art with three overlays, each showing
actual cyan, magenta, or yellow - which many, if not most, comics
publishers chose to forgo.
What I describe is how things generally worked by the mid-1980s,
definitely the tail-end of the era of traditional production methods.
By then desktop publishing was starting to make itself felt, and it
wouldn't be long before it changed just about everything.
More information about the DCML