Product Reviews

As with most product claims, you have to study the small print. The good news: Small
print is legible with all these printers.

The last time I reviewed high-end network laser printers [GCN, June 28, 1997, Page 37],
the test machines were difficult to network. I applaud Hewlett-Packard Co., IBM Corp. and
Xerox Corp. for making printer administration much easier.

Xerox deserves a medal for the most improvement, and HP garners a Reviewer’s
Choice for its next-generation Mopier 320. Without doubt, getting the Mopier together and
onto the network was a breeze.

The GCN Lab examined five printers:

Without exception, they would all perform well in any networked environment. But their
promise was high speed. In the race to deliver a variety of print jobs, most fell short of
their promised ratings.

In this report, the GCN Lab introduces click-to-clunk speed tests. Like real-world
users, we count the time from clicking the Print button until the final page clunks into
the output bin. The most challenging test tool turned out to be a 44-page Adobe Acrobat
file downloaded from Adobe Systems Inc.’s Web site. This complex 6.6M document
stymied the printers with its mixture of graphics, fonts and layouts.

Printing under PostScript or emulated PostScript, the Xerox DocuPrint N40, which
boasted the fastest speed rating, was the slowest. It took longer than eight minutes on
the Adobe test file. IBM’s two models took more than seven minutes each. Both HP
machines finished in less than six minutes.

The Xerox and IBM printers both use true Adobe PostScript Level 3, while HP emulates
PostScript Level 2— HP’s own version of PostScript, which might give it a speed
advantage. But even a simple 100-page text document took longer than the rated times for
all but the IBM InfoPrint 32. The 40-ppm Xerox took three minutes when it should have
taken 30 seconds less.

Finding out how the printer makers come up with their ratings is a challenge. IBM
documentation promises each model will print up to the stated speed. HP’s Web site
just lists 32 ppm and the words “text and photo.”

In GCN Lab testing, most of the printers did come close to their ratings on the
simplest jobs; on one test, the InfoPrint 32 exceeded its 32-ppm rating. But overall, the
vendor ratings were overly optimistic.

The lab rated each printer on a variety of one-sided, letter-sized test documents. The
results:

Despite the fact that no printer bought these days is likely to reach its stated
rating, the five network printers may still be the speediest on the market. But the Xerox,
which should have been the fastest overall, printed slower than all three 32-ppm units.
Moreover, the IBM InfoPrint 32 and Xerox have the same external hardware under different
labels. Their internal software and drivers differ, however, which may account for the
Xerox’s speed.

Speaking of twins, the two printers from Hewlett-Packard are the same, too, at the most
basic configuration. The LaserJet 8100N proved a little easier to set up than the Mopier,
which has additional media capacity and functions.

What most caught my attention was simple network installation. It took me a couple of
hours to assemble the Mopier, but getting it on the network was a snap. A wizardlike
process with clear explanations brought the Mopier to life in less than five minutes.

In my last network laser printer comparison, I took printer makers to task for their
arcane network installation methods. Even the original Mopier had antiquated
character-based commands, difficult TCP/IP addressing and old drivers that sometimes just
did not work. Not this time around.

Both HP printers grabbed IP numbers with ease, finding the Dynamic Host Configuration
Protocol service provided by Microsoft Windows NT Server. DHCP assigns an IP address to
each node.

HP’s installation even checked driver versions on the Web to see if updates were
available. Unfortunately, the Windows 9x client driver could not determine the Mopier
320’s configuration, so a manual update would be necessary for all network clients.
HP’s printer management tool, JetDirect, worked well but did not notice non-HP
printers.

HP offers limited management via the Web. While visiting the Mopier’s Web page, I
assigned the printer an administrative password. The JetAdmin application on my server
asked for the password. I gave the correct one, but a dialog box said it was wrong and
management would be restricted.

Canceling a print job on the IBM InfoPrint 32 and Xerox DocuPrint N40 required pressing
several buttons. I could never successfully stop a job on the InfoPrint 20. The control
panel read-out kept saying it was flushing, but the job kept printing. With both HP
printers, one press of a button deleted the job.

Moreover, the two-line HP LCD read-out gave more information. If I sent 25 copies of a
multiple-page job, the display showed how many copies out of 25 had been printed. And HP
backlit the display for legibility.

The Mopier is easy to like but difficult to afford. It starts around $14,000 and can go
beyond $18,000 fully loaded. But it can serve as a high-end copier replacement, so it
might be worth the extra cost (see story, Page 28).

Its print quality was good, showing smooth gradations and little banding. PostScript
proved slightly better at text. Words became legible at 2 points under PostScript and 3
points under PCL.

Diagonal hairline rules gave all the printers problems under PCL. The Xerox DocuPrint
N40 did a good job with text using any language.

Although the Xerox and 32-ppm IBM were almost identical, the Xerox did not arrive
put-together, as did the IBM. Attaching the Xerox duplexing unit turned out to be quite a
chore. The IBM’s duplexer arrived installed.

The Xerox CentreWare management software worked well and could at least see HP printers
on the network but not IBM or other models. CentreWare’s features were outstanding,
especially on the Web side, although consumable status could be a bit more detailed about
percentage used.

The original version of IBM’s Network Printer Manager shipped with the printers
crashed, but a new version posted to IBM’s Web site worked fine and discovered all
printers on the network. The Printer Manager lacks the polish of HP and Xerox tools,
however. IBM offers a Web Java version, but not as part of the printer, as does Xerox and
HP.

Both IBM printers claimed to use DHCP to obtain an IP address but could not do so. I
had to key in the numbers. The InfoPrint 32 was more complex to assemble because of its
finisher, duplexer and bridge, but the InfoPrint 20 took longer to put together because it
arrived unassembled.

The InfoPrint 20’s output looked good overall but was heavier on toner use than
the others. Banding appeared, especially on the PostScript sample.

The InfoPrint 32’s PCL photos lost gradations almost completely, showed a high
degree of banding and gave images a Sunday-comics quality.

In the end, rated speed may not matter that much. Administrators and users will be
delighted by the capabilities of all five printers, but they would especially enjoy the
Mopier 320. 

The GCN Lab examined five network printers rated at 20 pages per minute or faster. The
lab set up and installed the printers according to manufacturer specifications.

Each printer turned out more than 500 test pages over two weeks. The primary test
document, composed in Adobe Illustrator 8.0, had bitmap photographic images at several dot
pitches, vector chart illustrations, text sized from 1 point to 12 points, line art,
blends and black gradations.

The lab sampled each printer’s output in both Hewlett-Packard Printer Control
Language and Adobe PostScript formats as available, always printing from the same client
PC at the default load of the respective drivers.

We recorded the speed results on a click-to-clunk basis. Timing began at the click of
the Print button and stopped when the final page of the job slid out.

Lab staff timed the following:

The time it took to communicate between client PC and printer varied little from
printer to printer. The maximum was seven seconds for the Adobe Acrobat file. The averaged
speed ratings in this review do not include the time it took for print jobs to arrive at
the printer from the client PC and print server.  

"The following are two versions of the original printer test output file used with
this review. The original was a 65M Encapsulated PostScript file,too large to post here.
Both of these PDF files require Adobe Acrobat 3.0 available from http://www.adobe.com. PRINTHI.PDF
is about 2M and contains printable bitmap images. PRINTLOW.PDF
is a quick download at 400K, but it contains a much lower bitmap resolution, causing
photos to appear dark."

GCN Lab assistant Donovan Campbell contributed to this review.

HP has answered the critics with the Digital Copy 320 addition. Rather than tag on a
scanner, HP enhanced the whole package with an automatic feeder and excellent options.

The Mopier 320 with Digital Copy 320 makes a robust copier that could challenge almost
anything on the market.

The touch display panel shows all options understandably—a challenge these days
for any copier. The Digital Copy 320 does a good job of reducing, enlarging, making
multiple or duplex copies, stapling and collating. It can copy both sides of a two-sided
page at the same time, so there’s no feeding paper through twice. And the N-up
feature reduces two original pages to one.

The DC 320 doesn’t act as a scanner, which might complicate the machinery too
much. But HP should consider adding the features of its 9100C Digital Sender [GCN, Oct. 26, 1998, Page 24].

If you’re thinking of buying a Mopier 320, get the DC 320 addition and throw out
the office copier. Maybe someday HP will tell us whether to pronounce the first syllable
“mop” or “mope.”