BERTL Contex SD4400 Series Manual Download

Page: 22 / 29

Performance

September 4, 2008

All Rights Reserved. The license under which this document is made available and applicable law prohibit any reproduction or further transmission of any portion of this document. This
document may only be viewed electronically through the www.BERTL.com Web site and may not be stored in electronic or hard copy format. Any reproduction of trademarks is strictly
prohibited BERTL accepts no responsibility for any inaccuracies or omissions contained in this document

Page 22

August 2008

Contex SD4400

www.bertl.com

SCAN SPEED TESTS

We performed a series of scanner speed tests for black
and white, grayscale, and color scanning. For each test,
we measured the scanning speed between 100 to 600 dpi
resolutions in steps of 100 dpi. This test range represents
more than 95% of all wide-format scanner use and is
therefore, representative for a real user environment. Our
test drawing size is an architectural E-size (36” X 48”) for
both black and white and grayscale performance testing,
and for color performance testing we used a 36” X 24”
color drawing.

For the performance test, we used a desktop computer
running Microsoft Windows XP/SP3 with a Q6600 Intel
core 2 quad processor, 3.25GB of memory, and
connected to the scanner via the USB 2.0 interface. The
test is performed using the supplied scanner software.
While this computer is not the fastest available, it
represents a typical computer used by a typical scanner
customer.

BERTL analysts  measured the scanning time by starting
a regular stopwatch at the time we pressed the scan
button and stopping it when the scanning application
finished scanning the image into the destination file. We
call this test the "observed scanning speed" for a drawing.
For black and white drawings, we use TIFF Group 4
compression, while for grayscale and color scanning, we
used JPEG file format. The error in measuring time by
hand is approximately 1 second, which indicates that
black and white performance tests can show great
variation at as much as 25% error for less than 300 dpi
scans; at higher resolution, the relative errors converge to
less than 5%.

For each scan timed, BERTL analysts also measured the
reaction time from the time we pressed the scan button to
the time the scanner actually started scanning, and then
we measured the end time at the completion of the scan.
The reaction time is in the range of less than a ½ second
in black and white and grayscale, and approximately 2
seconds in color scanning.

We then graphed the results with both the observed time
and the adjusted time, which is the observed time minus
the reaction time. This performance is more true to the
actual true speed of the scanner. Finally, we graphed the
scanner manufacturer's claimed speed time. All
measurements are in inches per seconds (ips).

BERTL analysts caution against the absolute
interpretation of the performance result because the actual
result depends heavily on the computer in use and the
scanner interface.

The SD series has different performance specifications as
indicated in table below.

SD Series
@200 dpi

B/W

Ips

Graytone

Ips

Color

Ips

SD4410

10 10

N.A.

SD4430

5 5 1

SD4450

10 10 1.5

SD4490

10 10 3

BLACK AND WHITE PERFORMANCE

The black and white performance after adjustment is just
below the manufacturer’s claimed speed. Note, however,
that the observed speed is roughly one-third to half the
claimed performance. The scanner reaction time is low
and below 1 second, and we don’t see any differences
between the observed and adjusted scan time. We do
notice a dip in performance at 500+ dpi. The green line is
the one that should be measured against the
manufacturer’s claimed specs (blue line). The red line is
the real elapsed scanning speed from the moment we hit
the scan button and includes the scanner latency to the
time the file has been stored on the computer.

B/W Performance - SD4410, SD4450 & SD4490

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0