Digital BW, The Print

Peter Nelson wrote:
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, Ernst Dinkla 
> <E.Dinkla@c...> wrote:
> 
>>What Peter Nelson calls "writing 5 lines of code" is an 
>>underestimation of the problem in my opinion.
> 
> 
> In my original posting I said "if they did it right".  
> 
> As I said, I used to write inkjet drivers.  The obvious difference 
> between the 2200 and the 4000 is that the 2200 produces very obvious 
> patterns, which increase the visibility of the dots, and the 4000 
> does not.   There are essentially 2 kinds of dithers: random dithers 
> and patterned dithers.  The 2200's patterns don't look like a 
> patterned dither, they look like a badly done random dither.  So I'm 
> surmising that that's what needs to be fixed.

So far I have little proof that Epson engineers are of the lazy 
kind. If the random droplet distribution as observed is caused by 
a selected "random" algorithm for the dithering they must have 
done that for a reason.  If the random pattern is the result of 
the other points I mentioned it will be hard to get that 
corrected by whatever pattern. But one could check the different 
Gimp-print ditherings for the 2100/2200 and compare them with the 
Epson result. The Gimp-print patterns are also well described in 
the sources etc. Looks like there are more categories of patterns 
than just two. If Gimp-print shows better behaviour in one of its 
patterns you would have a strong point in the discussions with Epson.

I've written before that BO printing may become better with a 
higher nozzle quantity. Either by a bigger black head or by using 
more heads for blacks and reducing the black output per head. I 
very much think that that is a better way to go.  But it needs a 
driver like QTR to get that done.

Ernst