Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Steven Karafyllakis" 
<stevekphoto@...> wrote:
>
> Hi Roy;
> 
> Thanks for an informative response; perhaps you're right, that the 
> question should be laid to rest, at least until next halloween when 
> the_zombies_ will_ rise_ again... but a question first and an 
> observation, and then I will shut up. 
> 
> First, how exactly is dither accomplished? how does a driver take 
> one dot, make it several dots and what information is it using to do 
> so?

The dither algorithms vary from pretty straightforward to quite complicated.
But the basic concept isn't too tough to understand.  The idea is to simulate
many grays using a smaller number of grays by averaging dark dots and light
dots.  As a simplified example: for a spot on the paper 1/360 x 1/360 inch
and printing at 2880x1440dpi there are 8x4=32 places for dots.  To start with
you can put anywhere from 0 to 32 drops of ink there.   That's 33 levels.  Then
consider multiple inks, there are lots of possibilities of a light ink and dark ink
giving many levels in between all light and all dark.  Overall the idea is to average
many dots.  The difficult part is that the dots can't look like there's a pattern.
Our eyes and brain are extremely observant when it comes to patterns so a lot
of effort goes into eliminating this. 

> 
> And the observation: you said > I'd venture a guess that most 
> all "real" pictures have less than 8 bits of "real data". Perhaps, 
> but we routinely ignore the fact that silver granules convey 
> information in more ways than just density: size, for one, which 
> variable dot size partially reproduces, when it is used. Shape of 
> the clusters, and sharpness of the edges; and perhaps even 
> transparency for partially developed granules.  Much of this is not 
> visible to the unaided eye, and perhaps not quantifiable, but it all 
> contributes. 

The nice thing about silver grain is that it gives the non-pattern look by default.
Even a very grainy film that is excessively enlarged yields a grainy but pleasing
result.   

Roy

> 
> Best regards,
> 
> Steven Karafyllakis
>