[comp.lang.postscript] Faster LaserWriters?

eric@batcomputer.tn.cornell.edu (Eric Fielding) (10/26/89)

We have a (still running fine) Apple LaserWriter (original model) connected
to our VAXcluster.  We are finding that it is simply too slow in rendering
complicated or large PostScript files.  For instance, one program generates
files that are about 13 MB each that take about 4-5 hours to print on the
LaserWriter.  Would it help for us to buy a new LaserWriter II?  Is the
II/NT significantly faster at processing or is the 9600 baud connection the
main bottle-neck.  Last time I was paying attention (over a year ago) the
laser printers that connect directly to an Ethernet were much more expensive
(we probably could not find more than several thousand or so).  

Oh, and a related question.  There is a Linotronic printer somewhere here 
on campus.  Would it be capable of rendering large quantities of PostScript 
efficiently if we want a publication-quality version?

				++Eric Fielding
eric@geology.tn.cornell.edu

woody@rpp386.cactus.org (Woodrow Baker) (10/27/89)

In article <9144@batcomputer.tn.cornell.edu>, eric@batcomputer.tn.cornell.edu (Eric Fielding) writes:
> We have a (still running fine) Apple LaserWriter (original model) connected
> to our VAXcluster.  We are finding that it is simply too slow in rendering
> complicated or large PostScript files.  For instance, one program generates
> files that are about 13 MB each that take about 4-5 hours to print on the
> LaserWriter.  Would it help for us to buy a new LaserWriter II?  Is the
> II/NT significantly faster at processing or is the 9600 baud connection the
> main bottle-neck.  Last time I was paying attention (over a year ago) the
> laser printers that connect directly to an Ethernet were much more expensive
> (we probably could not find more than several thousand or so).  
> 
> Oh, and a related question.  There is a Linotronic printer somewhere here 
> on campus.  Would it be capable of rendering large quantities of PostScript 
> efficiently if we want a publication-quality version?
> 
> 				++Eric Fielding
> eric@geology.tn.cornell.edu


The Laser connection in Mobile alabama sells a ps-jet+ upgrade lid.  Runs at
16 Mhz with 2 mb of ram.  Old laserwriters were about 8 mhz with 1.5 meg ram.
The extra .5 megs makes a LARGE diffrence.  Sells for 1800-2000.  I can get you
one if interested, call.  (512) 837-8317 for Woody Baker.

jeynes@adobe.COM (Ross A. Jeynes) (10/28/89)

In article <9144@batcomputer.tn.cornell.edu> eric@geology.tn.cornell.edu writes:
>files that are about 13 MB each that take about 4-5 hours to print on the
>LaserWriter.  Would it help for us to buy a new LaserWriter II?  Is the
>II/NT significantly faster at processing or is the 9600 baud connection the
>main bottle-neck?  Last time I was paying attention (over a year ago) the

If you do the math:

9600 bits    60 sec     60 min    1 byte      1 MB                    MB
--------- X -------- X -------- X -------- X --------        = 3.456 ----
1 sec        1 min      1 hr        10 bits  1,000,000 bytes          hr
                                      ^
                                      | start, stop, parity


 1 hr          13 MB
--------  X  ---------  = 3.76 hrs data transmission time for a 13 MB file
3.456 MB        1

It seems that data transmission time is a large part of the problem.  In my
experience, the speed of the communication channel greatly impacts the 
printing speed.  A faster CPU will increase the processing speed as well;
an image file that I just tested that took 132 seconds to print on the 
original Laserwriter took 45 seconds to print on an NTX.  (This was over
appletalk, where the communication time should be less of an issue.  These
numbers should not be construed as a solid benchmark, just a test that I
ran to see if there was a noticeable difference.)


>laser printers that connect directly to an Ethernet were much more expensive
>(we probably could not find more than several thousand or so).  

Geez, how many do you need?  :-)


>Oh, and a related question.  There is a Linotronic printer somewhere here 
>on campus.  Would it be capable of rendering large quantities of PostScript 
>efficiently if we want a publication-quality version?

Typesetters generally operate more slowly than the low-resolution devices
because of the differences in the imaging machinery, and because of the 
way the image is built.  A typesetter generally builds a display list on
the hard disk of the RIP, whereas the lower-resolution devices, like the 
Laserwriter, use a frame buffer in RAM and just blast the bits onto the 
page.  In the high resolution typesetter case, it would cost too much in
RAM to use a frame buffer.
 

Ross Jeynes              
Developer Support                                   jeynes@adobe.com
Adobe Systems Incorporated                 {sun|decwrl}!adobe!jeynes