We tested monitors with a combination of hardware and software tools that were designed to examine every aspect of screen quality, power consumption, and ease of use. To do this, we conducted more than 40 separate visual inspections on each monitor. All tests were performed at a standard resolution of 1024 by 768 pixels using a VESA-compliant vertical refresh rate of 70 Hz. VESA considers this to be the lowest recommended vertical refresh rate at this resolution to avoid screen flicker. We used #9GXE video adapters by Number Nine in Compaq Deskpro 66M computers for our test bed.
The eye is superb at seeing differences in video quality, but the brain is poor at remembering them, so we used a video splitter to send test-screen images to three sample monitors as well as to the monitor under test. The three control monitors represented the low
, middle, and high range of quality in our test sample.
The video splitter took the video signal from the computer, amplified the signal, and split it into four channels. The test monitor received one channel, and the control monitors received the remaining three channels. Because all four monitors showed the same image simultaneously, we could easily compare the test monitor's image quality to that of the best and worst monitors we evaluated.
IMAGE QUALITY
Our overall display quality score was based on a number of screen tests. To gauge image quality, we used 26 different screen images in Sonera Technologies' DisplayMate Professional to help us examine quality characteristics ranging from blooming (in which brighter images become larger and more unfocused than darker images) to local regulation (i.e., distortion caused by high-intensity images, frequently seen under Windows at the status bar). Each monitor was given a score of from 1 to 10 based on how it displayed each of the 26 screen im
ages. We summed these results and gave each monitor a final score based on that total.
To measure image sharpness across the entire display area, we wrote a custom utility using Microsoft Visual Basic. This program produces images of boxes and lines at the selected Windows resolution of 1024 by 768 pixels. The box test produces 1- and 2-pixel-thick red, blue, and green boxes at the outside edges of the screen. We examined each monitor's ability to display these boxes with a black line between the boxes.
A second series of images measures the ability of each monitor to display fine vertical and horizontal lines. Again, we tested with red and green lines separated by a black line. We rated each monitor using a set of precise scoring guides that instructed the tester to assign point values based on the screen image. We averaged the scores for the 28 measurement points that made up this test.
DisplayMate Professional also provided a systematic approach to measuring geometric distortion. Disto
rtion manifests itself as oddly shaped screen images. The image may appear as a trapezoid or a barrel shape, with the tops or sides of the display area not parallel or at right angles. We measured the length of a series of lines displayed on the monitor, and the program calculated the percentage of distortion. Prior to testing, we used the monitor's controls (if any) to visually correct any existing distortion.
The tests just described look at specific quality characteristics. Our legibility test, however, provided a real-world measurement of monitor quality. Using Microsoft Word for Windows, we constructed a document featuring six standard typefaces at point sizes from 4 to 14 points. We then determined at what point size the text samples were both readable and legible. We judged text to be readable if it could be discerned without strain from a standard reading distance of 24 inches. We considered text to be legible if it could be discerned from any distance less than 24 inches without the use of aid
s such as magnifiers.
For consistent results, only one tester conducted this test. In all our subjective tests, each tester conducted one complete set of tests on all the monitors. This system worked to eliminate scoring variances.
In our charts, higher numbers indicate better quality. The highest overall quality score we awarded was 9.64, for the NEC 6FGp, a 21-inch display. The lowest score was 7.54 for Megatron Computer Systems' MegaImage L15MG, a 15-inch monitor. We found that the difference between a score of 9.5 and 7.5 is apparent to most monitor users. A 1-point difference is harder to discern visually unless you're looking for specific problem areas. Seeing distinctions in monitors whose scores differed by less than 0.5 point requires a sharp eye and some idea of what to look for, unless you're aided by software or hardware that gauges quality.
We did not evaluate color correctness because of the difficulty in accurately measuring this and because people who need exact color repr
oduction must calibrate monitor color values to other hardware devices, such as printers and scanners. If you need accurate color, we advise you to choose a monitor that allows for color adjustments and to perform your own color tests before making a purchase (see "Color-Matching Monitors" on page 218 and "Do-It-Yourself Monitor Testing" on page 222).
To arrive at our overall quality score, we weighted the image-quality and image-sharpness tests equally; together they accounted for 40 percent of the total quality score. Convergence scores represented 20 percent of the total, followed by legibility at 30 percent and contrast ratio at 10 percent.
POWER CONSUMPTION
We tested power consumption by measuring each monitor's current draw. Using a digital multimeter connected to the monitor's power system, we took readings when the monitor displayed a full Windows screen and when the screen was blanked. For monitors with power management features, we measured power consumption in the active, standby
, and suspend modes. Only the full-screen Windows figures are published and scored.
The Roll Call lists wattage consumption for all the monitors. The summary charts in each of the display-size rankings list power scores based on a 10-point index. Higher numbers indicate that a monitor used less power than another product in that size class.
EASE OF USE
To evaluate how easy it was to set up and use each monitor, we considered the various adjustment controls, cabling, and tilt/swivel bases, as well as the documentation that came with the monitor. We ranked monitors on the placement, range, and ease of adjustment for image controls. Controls typically include adjustments for brightness, contrast, horizontal and vertical size and position, pincushion (distortion), and degaussing. Some products include controls for color matching, phase adjustment, and resetting or saving settings. A greater range of controls combined with front-panel locations earned a higher score.
For making adjustment
s, most of the products featured digital controls using push buttons instead of analog thumb wheels. We judged that digital controls offer a wider range of adjustments than analog controls. But sometimes the sheer number of controls and their poor design and layout made digital controls more difficult to use.
We also rated monitors for the range of tilt and swivel of their bases (all manufacturers included such bases with their products). In addition, we considered the ease of panning or tilting the monitor.
Products also received higher scores if their video cables were longer than average. We reviewed documentation for clarity and completeness. Finally, products received higher scores if they offered a wider range of technical-support options, such as toll-free phone service and on-line services.
FEATURES
We evaluated monitors for features such as the number of factory-preset and user-definable resolution modes, video connectors, maximum usable screen area, weight, maximum resoluti
on, and compliance with MPR II standards for emissions and EPA's Energy Star standards for power consumption.
Best Overall winners in each size category had the highest scores based on the weighted average of scores in the quality, image-sharpness, usability, features, power consumption, and distortion tests. Quality scores accounted for 45 percent of the total score, while features represented 20 percent, followed by ease of use (20 percent) and power usage (15 percent). We used the scores for quality, sharpness, misconvergence, and distortion to select High Quality winners. Low Cost winners were those monitors priced below the average for the size class and having the highest ratio between overall scores and price.
Photograph: Testing team (clockwise from left): Andrew J. Froning, Siva Kumar, Chandrika Krishnamurthy, and Alan Joch.
Photograph: Power multimeter
Andrew J. Froning, Editor/PC Digest, an NSTL publication, spent the last three years
testing monitors, systems, and computer peripherals.
Alan Joch, Senior Editor/BYTE, coordinates the combined testing between the BYTE Lab and NSTL.
Chandrika Krishnamurthy, Technical Analyst/NSTL, evaluates computer peripherals and systems.
Siva Kumar, Technical Analyst/NSTL, specializes in hardware and network operating-system testing.
Andre Whittle, Consultant/NSTL, has evaluated computer hardware for the Canadian government.
How We Tested
