Here's my understanding of how the dither pattern at least for offset
printing is selected for "mixing" pigments. For highlights, the patterns are
chosen so that dots of different colors do not overlap. This is because when
you print one color on top of another, the one on the top isn't entirely
transparent to the one on the bottom (due to the chemistry of the pigments,
carriers, or whatever), and so you don't get a resulting color that might be
predicted by abstract theory. It is particularly important that this lack of
total transparency not come into play in the highlights, where it would
really be noticed, and fortunately because you need fewer dots in the
lightlights, this avoidance is possible when printing highlights. You can
see this for yourself by examining the highlights in a high quality printed
magazine like National Geographics. But as the density of the dither pattern
increases, this lack of total transparency in the pigments becomes less
noticeable, and it is a good thing, because as Art says, it is impossible to
avoid dot overlay when the density starts rising.
Now what I'm talking about is offset printing, where the pigments have
classically not been transparent. I haven't been in this business since
1988, and even back then we were experimenting with liquid toner pigments
from SRI (Stanford Research Institute) that WERE almost entirely
transparent, and the reason these SRI inventions were considered valuable
(and were patented) was because it was believed we would be able to choose
any dither pattern we wanted (stochastic for example) without having to
worry about getting the colors wrong in the highlights. I'm sure this
research into pigment chemistry has continued, and so today perhaps it is no
longer important to painfully construct dither patterns that do not overlay
in the highlights. I don't know. Haven't bothered to take out my 10X loupe!
Frank Paris
marshalt@spiritone.com
> -----Original Message-----
> From: owner-filmscanners@halftone.co.uk
> [mailto:owner-filmscanners@halftone.co.uk]On Behalf Of Arthur Entlich
> Sent: Thursday, March 01, 2001 12:54 AM
> To: filmscanners@halftone.co.uk
> Subject: Re: filmscanners: File sizes, file formats, etc. for printing
> 8.5 x 11and 13 x 17...
>
>
>
>
> Austin Franklin wrote:
>
> >>>> ... that dot
> >>>> from your printer can only be one of at test a dozen colors. In most
> >>>> cases ... the printer can only place on dot of
> >>>> each color in it's cartridges on any one spot. So, a CMYK
> >>>
> >> printer, can,
> >>
> >>>> for instance, place a C, CM, CY, CK, CMY, CMK, M, MY, MK,
> MYK, Y, YK or
> >>>> CMYK dot down in one of those locations.
> >>>
> >>>
> >>> I am curious how you know they (let's say the Epson printers)
> print more
> >>> than one dot at any one 'point' on the paper, ie, the CM, CY
> >>
> >> etc. 'colors'
> >>
> >>> you mention? I believed, though I could be wrong, they only print one
> >>> color, C or M or Y or K at any spot on the paper. That they
> >>
> >> did no blending
> >>
> >>> of ink, which is what I take you are saying they do.
> >>
> >> Let's call it empirical observation. With my 720 dpi printers, when I
> >> print a light green, for instance, using a loupe, I can see that the
> >> yellow and cyan dots overlay, in fact they are in almost exactly the
> >> same location.
> >
> >
> > I asked 'the power that knows for sure' and he said no, they do
> not overlay.
> > They are very close, but the definitely do not overlay.
>
> Nothing personal, but I don't believe your power. My magnifier assisted
> eyes say otherwise. Ask your "power" how the big block of near black
> CMY is created on the Adobe test print if the dots cannot overlay.
> Also, whomever the "power" is... did he work on the drivers for the
> original 720 dpi printers I use - the Epson Stylus color, pro and XL,
> which are the printers I'm using?
>
> Art
>
>
