Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!spool.mu.edu!uunet!zephyr.ens.tek.com!orca.wv.tek.com!pogo!davidh8
From: davidh8@pogo.WV.TEK.COM (David Haight)
Newsgroups: comp.sys.sgi
Subject: Re: Tek pahser PX
Message-ID: <11007@pogo.WV.TEK.COM>
Date: 5 Jun 91 01:15:14 GMT
Article-I.D.: pogo.11007
References: <9106012100.AA28850@masig4.ocean.fsu.edu> <1991Jun3.223025.8536@umbc3.umbc.edu> <1991Jun4.210057.29808@odin.corp.sgi.com>
Reply-To: davidh8@pogo.WV.TEK.COM (David Haight)
Organization: Tektronix, Inc., Wilsonville,  OR.
Lines: 62

In article <1991Jun4.210057.29808@odin.corp.sgi.com> blythe@sgi.com (David R. Blythe) writes:
>>In article <9106012100.AA28850@masig4.ocean.fsu.edu> davis@MASIG4.OCEAN.FSU.EDU (Alan Davis) writes:
>>
>>  I am trying to use a Tektronix Phaser PX on a PI 4D/35 3.3.2 without much
>>luck.  I have been using both tocolps and tops to make color PostScript which
>>I am sending to the printer.  The quality of the print is extremely poor, the
>>image comes out with a very mottled appearence, like something happened
>>during the dithering, ie the whole image is speckled with white dots.  I assume
>>
>>
>
>tocolps should be adequate for the job since it does nothing more than
>copy the bits to the file for imaging by the printer.  It does, however,
>set the screendensity and screenangle for each of the magenta, cyan, yellow
>and black primaries.  These are undoubtably causing your problems.  That
>and possibly your expectations may be too high.  You might try deleting
>the manipulations of the angles and densities from the PostScript file
>since these are often better left as the manufacturers defaults unless
>you know what you are doing.  Otherwise, the printer documentation may
>give you some clue to alternate values to try.  I couldn't begin to guess
>myself :-(.  
>
>	david blythe
>	blythe@sgi.com

The problem is "blithering" or double dithering.  Try downloading this first.

%!
%% Patch to the setscreen operator to avoid unwanted behaviour and to
%% increase speed, but at the expense of disabling patterning.  Some
%% applications set their own screens, but these generally look worse than
%% the built-in ones in the Tektronix Phaser printers, so we want to use
%% the internal ones.
%%
%% The patch simply executes the spot function procedure four times, once
%% at each corner of the halftone cell, and throws the results of those
%% executions away.  Then all three arguments to setscreen are discarded.
%%
%% The proc would not have to be executed at all if it was not for those
%% applications that set variables inside of the proc and refer to them
%% outside of the proc (for determining the threshold gray level of a
%% pattern).  GEM applications and Broderbund TypeStyler do this.  Normal
%% setscreen operation could execute this proc thousands of times.
%%
%% This patch will only apply itself once.  The patch can be circumvented
%% by directly referencing the setscreen definition in systemdict, as in
%% "systemdict /setscreen get exec"

serverdict begin
0 exitserver
/setscreen load type /operatortype eq    % patch only if not patched already
{
  /setscreen {
         .99999  .99999 2 index exec pop % execute proc, discard result
         .99999 -.99999 2 index exec pop
        -.99999  .99999 2 index exec pop
        -.99999 -.99999 2 index exec pop
        pop pop pop                      % discard setscreen arguments
  } bind def
  (Setscreen patch downloaded) = flush
} if
%%EOF