#Publish! & LaserMaster
9 messages in this thread
Hi. I just got in on the tail end of the discussion about Bob Weibel's
review in Publish!, and I'd like to make a few comments.
I examined the article, and I agree that the print quality is poor. I
can think of several possible explanations, but whatever the cause was, I
can print with Baskerville now, and the output at 1000 x 1000 TurboRes is
much better than that appearing in Publish!
For anyone who is concerned about this issue, please EasyPlex me at
this signon, and I will mail actual output samples to you. If you have a
sample Ventura or PageMaker chapter you would like printed (semi-double
blind testing), ship it to LaserMaster, 7156 Shady Oak Road, Eden Prairie,
MN 55344, and I will print it for you.
TurboRes is not a buzzword. I don't think it is fair to judge the
output quality so severly on the basis of a single sample — and we welcome
the opportunity to show the critical publishing community that we have what
we say we have.
Chip Aspnes
LaserMaster Corp.
OK, Chip, you're on! I'll be Plexing my address to you forthwith, as I'd
like to see samples of the 1000 x 1000 dpi output from LaserMaster.
Would you care to conjecture as to why the output shown in Publish! is so
bad? Did they forget to flip the magic 1000×1000 switch, or do you think it
got munged at the printer (as in real printer, not laser printer)?
And I think you're correct in assuming that this is a "critical publishing
community"! <grin>
<BW>
Chip, I wrote you a long message on this topic (#49830) over in IBMAPP
because that's where you'd last been spotted. I also called Dean Schmidt
(LM sales rep) late last week to see if he could tell me anything.
I've got a 300 x 600 dpi LX 6 setup here, and in addition to my day-to-day
experience, I've got a bunch of direct print samples at 600 x 600, 800 x
800, and 1000 x 1000 dpi. I have to say that whereas I've always been
delighted with LaserMaster, I have got some real hesitation about some–not
all or even most–of what I've seen of TurboRes. The magazine page is by
far the worst example, but some of the samples I have direct from
LaserMaster show a related defect. Specifics are in my message on IBMAPP,
which I will Email to you if it has scrolled off by now.
I get the impression with the LX 6 and TurboRes that the printer is still
printing at 300 dpi vertical resolution, but that the spot size (diameter)
is being varied in real time as the laser scans across the drum. If that
is indeed what is happening, it helps explain the defect that can occur in
certain letterforms, whereby a horizontally connected group of sub-pixels
(if I may call them that) simply goes blank.
Those missing pieces of letters, I assume, occupy a vertical position
between the actual scan lines of the laser; perhaps they are unprintable
because the spots above and/or below them are white in the given character,
and the beam can't just "jump" up or down to write them when the position
in the direct path of the beam is supposed to stay "off."
(This whole discussion is severely hampered IMHO by excessive secrecy about
what exactly TurboRes does, let alone how it does it.)
I would appreciate your comments or those of anyone else from your company.
David – I have an LX6 800×800 Turbo Res on an HP IID laser and I just
haven't experienced the problems you seem to be having. So I am a bit
confused. I saw your message in the IBMAPP sig several days ago, but would
appreciate it if you would refresh my memory on the specifics of the
defects. I certainly am getting drop outs or poorly formed letters on my
setup? I just don't get any drop outs (blanks). Also, he samples I
received from LM of both the 800×800 and the 1000×100 0 are free of
dropouts? In fact I just showed some of the 1000×1000 samples to a
typesetter today and she thought it was pretty darn good, but not as good
as her Linotype CR300 <grin>.
—– bob
I have direct output samples of 600 x 600, 800 x 800, and 1000 x 1000 here.
The defects are generally slight except in that horrid looking page from
Publish! magazine, and they would normally be quite excusable. I am
trying, though, to be as critical as possible of this new approach so that
I can ultimately be convinced, one way or the other, as to whether it is a
definite step forward or not.
Do you have the 800 x 800 sample page with the photograph of the two boys?
Or the one with the letters "A" through "g" showing various sizes and
effects? If so, look at the body text (Bitstream Amerigo) and note the
gaps in the letters "n", "m", "h" and any other letter where a chiseled or
pointed stroke meets with a straight vertical stroke. The 600 x 600 dpi
sample does not have these gaps.
They've already got my check for the 800 dpi upgrade; I'm just asking for
reasonable assurance that it works better than these samples, which
requires an explanation of why these samples look the way they do. I could
conceivably be wrong about the design of Amerigo, but it is a typeface I've
used a certain amount and I think I'm reasonably familiar with it. If I'm
wrong, you or LaserMaster should feel free to tell me so. I've brought
this up with them, and haven't received any direct reply. Apparently they
are now working to clarify the situation with the Publish! article somehow.
–The message I wrote on IBMAPP was apparently deleted the other day; it
didn't scroll off the board naturally, and I'm pretty sure I did not delete
it myself. That didn't exactly increase my confidence level either.
Part of the problem, undoubtedly, is that Amerigo is a chunky face designed
for low-resolution lasers. Publish, on the other hand, uses New
Baskerville, a face with very thin serifs that start to fall off even at
1,270 from a Lino.
Since we're still applying for the patent on TurboRes, I'm limited in
what I can say, technically.
Some of the Bitstream faces have discrepancies between the Bezier data
and the arc-and-line data (neither one is wrong, just slightly different).
Our software was, occasionally, tripped up by the difference.
We are presently reviewing prints of every typeface, at every resolution,
in numerous point sizes. This will take some time, but is enabling us to
tighten up the entire process of putting those little dots on the paper.
I have shipped three sets of Actual Output to people who requested it,
and I have several more requests to fulfill on Monday. I hope that the
proof of the pudding is in the eating, and we welcome your observations and
suggestions.
Chip Aspnes LaserMaster
Chip, I remember back when I couldn't believe that Bitstream would ever
make a mistake in their kerning and width tables, and I thought LaserMaster
was BS'ing me about the fonts (for example, Bodoni and Baskerville) looking
better with PostScript width tables, but darned if you didn't turn out to
be right after all. Bitstream's kerning table for Baskerville Italic, for
instance, says that in the combination "fa", the "a" should be moved
_further_ from the "f" than normal–which is exactly backwards. As I said,
I refused to believe this until I made a list of the bad-looking pairs,
disassembled Bitstream's kerning tables and saw it for myself.
So, I think at this point I'm going to trust what you're telling me about
the troubles with certain of their fonts. It is certainly my experience
that the vast majority of the 800 and 1000 dpi printout I've seen has been
first-rate. Still, until you've dealt with the problem you mentioned, how
do I know what will work and what will not? Can I upgrade my LX 6 to 800 x
800 dpi, and then run the pre-TurboRes software and avoid the problems?
–I understand that this release of the software allows LX 6 users to use
Adobe Type 1 fonts. Have there been any similar problems with those?
Chip,
So am I correct that once you've evaluated all the faces, there will be an
updated version of TurboRes?
Also, I've seen TurboRes advertised with a product from a company called
Raster Devices. They're located up in your neck of the woods–and their ads
strike me as quite similar in style and tone. Are your companies related?