Phototypesetters and imagesetters - Re: VT52s, VT61s lots of DEC and DG keyboards

Toby Thain toby at
Tue Oct 13 11:28:28 CDT 2015

On 2015-10-13 12:06 PM, Paul Koning wrote:
>> On Oct 13, 2015, at 11:52 AM, tony duell <ard at> wrote:
>>> The only other terminal I worked
>>> with that could do that was a Tektronix storage scope terminal (4010
>>> or 4014, IIRC).  The Tek printer wasn't built-in, but it did take a
>>> scan of the live screen, so that was similar.  The paper was
>>> silver-grey and I remember it coming out wet too.  Everything else I
>>> worked with was either thermal or dot-matrix impact, and could only
>>> capture text as it arrived at the terminal, not a screen image.
>> The tektronix printer (or 'copier' as they called it) was photographic. The
>> paper was light-sensitive, and went past the screen of a 'flat' CRT (I think
>> it only had deflection plates for one axis, the 'deflection' in the other
>> direction coming from the paper movement). It really was an odd-looking
>> tube.
> I've never seen that printer. But the tube you describe was common
> in
the 1970s in high end phototypesetting machines, such as the Autologic
APS-5 and the Linotron 202. I think the idea was to make high resolution
CRT display feasible by having to worry only about accuracy in one
coordinate, not two. The film movement (actually paper, essentially a
roll of photo-print paper) would provide the other coordinate. Unlike
the Tektronix machine you describe, phototypesetter output was developed
in essentially the same manner as photographic print, with developer and
fixer chemicals, in a machine that would take the paper casette as input
and produce the developed/fixed/dried roll of finished material coming
out the other end. The roll of paper would then be cut to separate the
individual columns and articles, and pasted onto cardboard page size
boards to produce the finished layout.

Excellent description. I spent a few years running a Linotronic L100, 
one of the first PostScript imagesetters, capstan feed. It worked on the 
same principle but instead of a CRT it used a laser and a fast rotating 
mirror to scan the paper in 1 dimension. It could, at a pinch, expose 
film directly, but struggled to get proper density. If I recall 
correctly the bromide paper roll was about 12" wide.

A friend in Sydney, Australia, has this hardware now, but if there are 
serious collectors in that region who really want it, I think he could 
be persuaded to part with it.

> At some point, there were machines that could produce page-width
paper, but those were rather expensive and useful only once you got
software to do the page layout on the computer instead of manually in
the composing room.
> I believe that phototypesetting is obsolete now, though it hung on
surprisingly long, certainly into the 1990s -- because laser printout
was high enough resolution to read, but not good enough to use as the
master for offset printing. I think phototypesetters were good to 1000
dpi or thereabouts, unlike the 300-600 dpi of the first decade or so of
laser printers.

The Linotronic series launched with PostScript by 1987 and produced 
"camera ready artwork" from the Macintosh printing system over AppleTalk 
(PageMaker, etc, or even MacWrite with its horrendous typography) or 
anything else over serial (FrameMaker, TeX, what-have-you).

I just found an Australian Dollar price list from 1987 -
* L100 AUD $69,250
* L300 AUD $96,000

The L100 had switchable resolutions up to 1270dpi and the L300 went up 
to 2540dpi - more than sufficient for offset.

The subsequent generation of imagesetters exposed film (requiring higher 
exposure and allowing the re-photography step to be eliminated) and 
secured the film on a spinning drum, allowing industrial standard 
registration. The capstan feed setters could not offer very high 
registration accuracy. The first imagesetter I used of this type was the 
Scangraphic circa 1992 - apparently the first PostScript drum imported 
to Australia - and it was superb.


> 	paul

More information about the cctech mailing list