laser-lovers@uw-beaver (04/24/85)
From: ihnp4!utzoo!henry@uw-beaver.arpa Brian Reid very kindly sent me a bunch of samples of LaserWriter output, partly for comparison with the LaserJet stuff I've got here. The graphics are very nice, and it's obvious that Brian is having fun. (Example: a record label with the lettering in a nice circular arc.) As I've said elsewhere, Apple's attempt to break into the toy market is clearly succeeding -- first the Mac and now this. I'm afraid I'm not so impressed with the text. It confirms my previous impression of Adobe's font scaling: no, Adobe has *not* found a magic alternative to bit-tuned masters for each point size. It's better than one would expect, but visibly imperfect. Letter spacings are uneven, to the point where letters sometimes touch, and there are seemingly-pointless variations in spacing between similar characters. The fonts themselves definitely look like they could benefit from bit-tuning, although I am not enough of a font guru to have a trained eye for this. Lest I give the wrong impression, I should add that it *is* better than I would have expected from a font-scaling scheme. Adobe clearly has a lot on the ball; it is no slur on them that they haven't *completely* solved a nearly-impossible problem. They've come closer than I would have thought possible. Compared to the LaserJet... Apples and oranges. The LaserJet's letter shapes are magnificent, although HP botched the spacing to the point where any self-respecting typesetting package has to position almost everything explicitly. Given this, the LaserJet does a superior job on ordinary text in 10-point Times Roman (oops, I mean Tms Rmn [ugh]). When the fonts and/or symbology gets hairy, or graphics becomes an issue, then the LaserJet bows out and leaves the field to the LaserWriter.
laser-lovers@uw-beaver (04/25/85)
From: Brian Reid <reid@Glacier> Naturally I don't agree with Henry on this; I don't remember having sent him any samples of ordinary text to compare against. My recollection is that I sent him various samples of compressed, expanded, rotated, shrunken, and otherwise transformed text as examples of what the graphics model could do. The only printout that I sent him in which letters occasionally touched was the Organized Crime 1040 form, which uses an 8-point Helvetica on which a horizontal compression transformation of 11/13 has been done to achieve Helvetica Compressed; this is something that is simply not possible on a LaserJet, because the letterforms aren't there. If you try to duplicate that same form on a LaserJet, what you find is that the letters don't fit into the spaces provided unless you switch to a smaller type face. Henry is well-known to be a LaserJet lover; I am well-known to be a LaserWriter lover. I claim my LaserWriter will do X and Y that his LaserJet won't; he claims his LaserJet will do X and Y that my LaserWriter won't. My feeling is that the educated reader of Laser-Lovers should not believe either of us without hard evidence. I have in my office hard evidence, consisting of pieces of paper, that convince me that the LaserWriter output looks nicer and that the fonts look indistinguishable, but you don't have this evidence and Henry thinks he has different evidence. I think that what I would like to do here is to invite supporters of various laser printers to contribute print samples to some national publication whose print quality is good enough to show the differences, and then let the readers of this Laser Lovers group, and the readers of that national publication, decide for themselves. I would expect that Imagen and maybe QMS might want to get into this fray, also. I don't have any obvious candidates for the publication in question, since laser printers fall in the cracks amongst the topics on which the fast-turnaround publications are centered. I would guess that the SIGOA newsletter would be the most appropriate forum for this. I think it would also be appropriate to have things like the Imagen 12/480 and the Meregenthaler A110 (or whatever it is called) included here, even though there is no corresponding high-end HP machine. This would provide a reference for the quality of the printing process, so that people could see where the distortion was coming from. Comments? Anybody know if SIGOA would be willing? Anybody volunteer to referee this? Brian Reid Stanford
laser-lovers@uw-beaver (04/25/85)
From: Robert Morris <ram%umass-boston.csnet@csnet-relay.arpa> The quality of the composition, not just the letter forms, will affect peoples judgment of the final product. The American Mathematical Society has substantial typographic competence. Although they have a historical connection to and endorsement of TeX and Computer Modern fonts, their staff is in my opinion critical and unbiased, and would make good ``referees''. If your proposal is to reproduce things photographically for further review, then the additional complication arises of relative degradation under duplication which may be highly relevant for some and irrelevant for others. A number of art departments e.g. at RSDI and Rochester have typography and graphic arts departments which undoubtedly have the competence to form or nominate the jury.
laser-lovers@uw-beaver (04/26/85)
From: ihnp4!utzoo!henry@uw-beaver.arpa
Much of Brian's output was indeed graphics rather than text; I should
have made this clearer. Pretty damn good graphics, too.
But there were some modest chunks of text, including a lengthy quote
from Daniel Berkeley Updike. He didn't actually send me the organized-
crime 1040, or if he did it's gotten lost. (I think I have seen it
elsewhere, though.) The Updike quote was in what looked like quite an
ordinary text font (I am not enough of a font guru to identify it by
sight, and it's not labelled), and by Ghod the letters sometimes touch!
Check out lowercase r followed by lowercase i or y, and lowercase t
followed by lowercase y. Some of these pairs, interestingly enough,
touch sometimes and not other times. Some other pairs are usually so
close, due to the uneven spacing I mentioned, that they *look* like
they touch unless you dig out the magnifier. The spacing really
could stand some work; it's the biggest flaw in the output.
I should probably send Brian some LaserJet samples.
I agree wholeheartedly that a side-by-side comparison in a well-printed
journal would be superior to a handful of personal opinions, and
would also be interested in seeing commentary by professional graphic-
arts people.
Henry Spencer @ U of Toronto Zoology
{allegra,ihnp4,linus,decvax}!utzoo!henry
[[Editor's comment: I usually refrain from commenting on either
Brian's or Henry's contributions to this forum since they are almost
uniformly fun to read. However, it seems to me quite unclear what is
really being discussed in this most recent exchange of messages about
the relative merits of the LaserWriter and the LaserJet. In the
earlier message, Henry commented that the letter shapes produced by
the LaserWriter looked a little off to him, but that as he was not a
font designer, he couldn't say for sure. More definitive comment on
this by a person with training as a font designer would be very
interesting to me, and I believe others. Unless the fonts were loaded
into the LaserWriter as bits (highly unlikely), that is also a
technical issue of interest. The rest of the previous message (and
the topic of this message) concern Henry's comments that the letter
spacing in Brian's sample printout could be better. The implications
of this are uncertain to me. To know whether this is a reflection on
the LaserWriter, we have to know how Brian produced the sheet.
Perhaps this is a failure in the LaserWriter, or perhaps it is a
failure in the text formatter Brian (or Adobe) uses, or perhaps this
is a failure in the width tables used by the text formatter.
I'd also be interested in hearing what the limitations are of the
troff LaserJet software that Henry uses. I've also seen the sample
sheets from Textware's Tplus and they are indeed nice. However, I
have the strong impression that if, for example, I wanted to increase
the size of the body type by just a little bit, I'd be in trouble.
After all, there are only a limited number of point sizes available in
the LaserJet's font cartridge and it isn't possible to print more than
a quarter page at full resolution in bit map mode. --Rick]]