|
Although LCDs have digitally addressed pixels, internally they are actually
analog devices like CRTs, where the brightness of each pixel is
controlled by an analog signal and circuitry.
Still, LCDs behave very differently from CRTs, so they need to be tested
and evaluated from a totally differently perspective.
A good LCD needs to perform well in all of the tests outlined in the
next section. But to earn the distinction of a Great LCD it
must excel in each of the critical tests below.
Note that all of the test patterns mentioned below are proprietary
DisplayMate test patterns and that most are found only
in the professional DisplayMate Multimedia Editions.
For background information and general guidelines see
How to Test and Evaluate Displays.
1. Dark Black-Level:
LCDs have difficulty producing black and very dark grays.
The black-level for an optimally adjusted LCD should appear to be
very close to a true black except in dimly lit environments.
Verify this visually
using one of the DisplayMate Black-Level Adjustment test patterns.
Vary the display's Brightness control to see the change in black-level.
A sensitive photometer may be used to measure the residual brightness
of the black-level.
2. Perfect Pixel Tracking/Phase:
LCDs using an analog input require careful adjustment of pixel tracking/phase
in order to reduce or eliminate digital noise in the image.
Use the DisplayMate Pixel Tracking and Timing Lock test pattern.
Timing drift and jitter may require frequent readjustments during the day.
For some displays and video boards you may not be able to
entirely eliminate the digital noise.
For digital input LCDs the Pixel tracking and Timing Lock test pattern
should appear flawless, but for some combinations
of video boards and displays you may see some noise.
3. High Video Bandwidth Index
The DisplayMate Video Bandwidth Index measures the visibility of
high frequency fine detail in an image.
A value of 100 is perfect and indicates that high and low frequencies,
corresponding to broad and fine detail, are identically reproduced.
LCDs using a digital input should have values very close to 100.
For analog input LCDs values near 100 are not normally possible because
of limitations in analog electronics.
Using one of the recommended Video Boards in the
DisplayMate Best Video Hardware Guide the Index should be in the
range of 90-95.
Values lower than the above indicate second-tier performance.
Higher values may indicate over-peaking
in the high frequency compensation for fine image detail.
Values near or greater than 100 indicate substantial over-peaking
and compensation, which is a serious performance flaw.
For either analog or digital input LCDs perform this test in Green.
See item 6, below for White and other colors.
To check for over-peaking in analog input LCDs
see Ringing and Overshoot, item 16, below.
4. Smooth Gray-Scale with 256 Intensity Levels:
LCDs have difficulty producing a perfectly smooth gray-scale with 256 discrete
intensity levels that increase uniformly from black up to peak white.
Use the DisplayMate 256 Intensity Level Ramp and 256 Intensity Level Color Ramp
patterns to evaluate the smoothness of the gray-sale.
You should not see any jumps, skips, ripples, kinks, flat spots, dips, bumps,
periodicities or other irregularities.
Also be on the lookout for dithering as some panels use this method
to generate portions of the gray-scale.
5. Accurate Gamma:
The shape of the gray-scale as it increases from black up to peak white is
described in terms of a mathematical power-law function with an index called
Gamma. The internal Gamma and gray-scale of an LCD is very irregular.
Special circuitry attempts to fix it, often with only limited success.
The Gamma affects the accuracy of the gray-scale and color mixtures
so specific values are necessary when high accuracy is required.
The Gamma for CRTs has been adopted as the reference standard and all
other display technologies are expected to reproduce its 2.2 to 2.5 Gamma.
Use the DisplayMate Gamma Correction Measurement test pattern to
accurately measure the Gamma interactively on-screen.
This pattern requires that the Video Bandwidth Index (item 3) be set first.
Note that viewing angle is very critical for this test, so make sure that
you are viewing the panel exactly face on. See item 8, below.
For a more detailed analysis, use the DisplayMate Window Pattern
and a photometer to measure the entire brightness curve.
Plot the measurements on log-log graph paper.
The results should be a perfectly straight line with a slope of the
desired Gamma value.
Most LCDs will not perform well here and are therefore not suitable for
professional image color balancing, although they still produce pleasing images.
6. RGB High Frequency Balance:
The Video Bandwidth Index (item 3) for each of the primary Red, Green and Blue
channels should be identical.
If not then there will be a shift in color going from broad to
fine image detail.
An easy way to check this is to view
the Video Bandwidth Index test pattern in White. Adjust the Index
slider so that the high and low frequency blocks are as closely
matched in intensity as possible. There should be no visible
difference in color tinting between the blocks.
7. No Color Tracking Error:
The intensity of the Red, Green and Blue channels should vary identically
with signal level. If that isn't the case then a pure gray-scale will
show color tinting in one or more areas. This often occurs near black at
low intensities and also near peak intensities due to unbalanced saturation
between the RGB channels.
Use the DisplayMate Color Tracking and White-Level Saturation
test patterns. No change in color tint should be visible.
For a more precise determination use the DisplayMate Window Pattern
and a color analyzer.
8. Wide Viewing Angle:
On LCDs brightness, contrast, gamma and color mixtures vary with viewing angle
because of light polarization effects.
At large angles this leads to contrast and color reversal.
As a result LCDs have a relatively narrow viewing angle and need to be viewed
fairly close to face on. Some LCDs have much wider viewing angles than others.
The variation with viewing angle depends on intensity and
is generally strongest at the bright end of the intensity scale.
As a result the Gamma (item 5) of an LCD varies with viewing angle as do
the resultant colors in color mixtures.
There are several DisplayMate test patterns that will demonstrate these effects
both qualitatively and quantitatively.
For all of these tests it is very important to set the Contrast control
to the highest value that does not produce White Saturation, item 13, below.
In the DisplayMate Color Scales test pattern the 3rd and 4th scales from
the top are brown with RGB=255,128,0 and yellow with RGB=255,255,0.
Viewed at large angles they appear almost identical yellow because the
bright 255 intensity components dim more quickly and approach the
128 intensity components.
The dim-end of the brown scale maintains its brown identity the longest.
In the DisplayMate Color Spectrum test pattern the color banding becomes
significantly more pronounced with increasing viewing angle.
This is due to the variation in Gamma with viewing angle.
Two DisplayMate test patterns can be used to measure this viewing
angle effect quantitatively without instrumentation:
Gamma Correction Measurement and Color Explore and Match with the
Gray-Scale Overlay.
We'll discuss the Gamma Correction pattern first.
View the pattern face on and adjust the slider until the middle dither
matches the surrounding gray in intensity. Record the value.
Increase your viewing angle and note that the dither is now dimmer than
the surrounding gray.
Adjust the slider to match the intensities again.
Record the value and repeat for increasing angles.
This will give you the variation in Gamma with viewing angle.
The Color Explore pattern will give the variation in RGB intensity with angle.
Set the Color Model to HSB, set S=0 and vary the middle B slider.
Match as above and record the intensity values.
If color tinting is a problem switch to the color Green.
The slower the variation with angle the better the LCD.
9. High Quality Scaling:
There is a considerable variation in the quality of the scaling engines
found in LCDs.
When an image is produced with a pixel format that is different from the
panel's native resolution the display's internal controller rescales the
image to the panel's own pixel format.
This involves sampling, interpolation, anti-aliasing and other algorithms
that generally result in significant degradation in image quality,
particularly with fine text and graphics. Perfect or very high quality
rescaling of digital images is mathematically impossible.
Use the DisplayMate Moiré Montage, Scaled Fonts and Page of Text
test patterns to evaluate the quality of the scaling.
Fine image detail will typically appear broadened, fuzzy, pixelated
and with lower contrast.
Also look for pattern irregularities in the different
Moiré dither patterns.
Evaluate the scaling at resolutions above and below the panel's
native resolution.
10. Few Motion Artifacts:
Slow image response times and scan rate conversion result in severe motion
artifacts and image degradation for moving or rapidly changing images on LCDs.
Use the DisplayMate Multimedia with Motion Edition to test for motion artifacts.
The Motion Engine moves all of the DisplayMate test patterns in a perfectly
smooth and uniform motion over the entire full-screen in
various directions and speeds.
Look for changes in the patterns when they are moving,
particularly blurring, darkening, fluttering and disappearing elements
and variations in color.
Use the DisplayMate sample Motion Script to check for common motion artifacts
with a select set of DisplayMate test patterns.
|
Testing and Evaluating LCDs
|
To thoroughly evaluate an LCD display you should examine it using all of the
DisplayMate test patterns.
Refer to the Test Information screen for each test pattern for
details on the purpose of the pattern and what to look for in the image.
In addition to the test patterns discussed in What Makes a Great LCD above a
number of important proprietary DisplayMate patterns are also featured below:
11. Screen Uniformity:
The brightness and color of an LCD can vary over the screen.
First, it's difficult to backlight the panel uniformly over the
entire screen, so there may be hot and cold spots that are brighter or dimmer
than average.
There may also be shading irregularities that affect the corners or
portions of the screen that go from
top to bottom, left to right, or from the outside to the inside.
Variations inside the glass panel itself may produce additional
irregularities.
Use one of the DisplayMate Screen Uniformity test patterns
to check the uniformity.
Look at both bright and dim intensities using a White screen color.
Use a photometer if measurements are required.
Also, cycle through Red, Green and Blue to look for intensity variations
in the primary colors.
12. High Peak Brightness:
You can visually compare the relative brightness between two
LCD displays but it's best to use a photometer to measure this value.
Before measuring peak brightness you need to properly adjust the
Brightness and Contrast controls. The Contrast control should be
set as high as possible without producing white saturation, item 13, below.
Use the DisplayMate Window Pattern with 100% area.
DisplayMate also includes the ANSI brightness test pattern with target
positions for measuring ANSI standard brightness.
13. No White Saturation:
The bright-end of the LCD intensity scale is easily overloaded,
which leads to saturation and compression.
When this happens maximum brightness occurs before reaching the peak
of the gray-scale or the brightness simply increases more slowly
than it should near the peak.
Use the DisplayMate White-Level Saturation test pattern to check
for this effect. You should be able to make out all of the blocks
up to and including the one labeled 253.
The Contrast control should have a strong affect on this test pattern.
14. High Contrast:
Display contrast is the ratio between the peak white intensity
and the black-level intensity.
The higher the better because visibility and readability depend on contrast.
While contrast can be compared or estimated visually it's best to measure
it with a photometer.
Unfortunately the peak white-level and black-level can be specified
in many different ways.
The peak white-level depends on the settings of the
Brightness and Contrast controls.
The black-level depends on the room lighting and will be lower
in a pitch black room. It will also be lower if the entire screen
is set to black. If portions of the screen are illuminated the
intensity of the black areas of the screen will increase due
to internal reflections within the LCD.
There are two common measures of contrast: full-field sequential and ANSI.
The ANSI method uses a carefully defined procedure and a 4x3 checkerboard
pattern that partially takes into account internal reflections.
The sequential method is non-standard but
generally has the Contrast control set at maximum
regardless of saturation effects followed by a
completely black screen in a pitch black room. This is the
contrast that manufacturers generally publish. It can be a
factor of 2 or more greater than the ANSI value.
The DisplayMate Multimedia Edition includes test patterns to measure
the contrast both ways.
15. No Bad Pixels:
LCDs can have stuck pixels which are permanently on or off.
Use the DisplayMate Stuck Pixels test pattern
with Black to check for pixels that are stuck on
and then with Red, Green and Blue to check for pixels that are stuck off.
Some pixels may also have irregular intensities and may be brighter or dimmer
than average.
Use the DisplayMate Screen Uniformity test pattern and cycle through the
intensities and primary colors.
Finally, some pixels may be improperly connected to adjoining pixels,
rows or columns, so they have the brightness of one of their neighbors
instead of their own.
This is much harder to check.
Cycle through all of the full-screen Moiré patterns in Red, Green and Blue
to search for pixels that are not part of the regular Moiré pattern.
16. No Ringing and Overshoot - analog input only:
Poor signal processing or over-peaking will produce Ringing and
Overshoot, which is visible as extra light or dark edges or waves
around fine image detail.
Use the DisplayMate Ringing and Overshoot and Sharpness with Contrast
test patterns. The Reverse Video Contrast and Color Streaking patterns are
also excellent. No Ringing or Overshoot should be visible.
17. Weak Reflections:
Only very weak internal and external reflections from the screen
should be visible.
This is accomplished with high quality anti-reflection coatings and
high-absorption glass.
The internal reflections within the panel
reduce the very important local contrast.
Use the DisplayMate Dark Screen, Screen Halos, and Internal Light Scatter
test patterns to search for stray light.
View these patterns in the darkest possible environment.
18. Weak Streaking:
Streaking is visible as positive or negative shadows that trail
image detail on the right.
When there is severe streaking it will wrap around and appear to
precede the image detail.
Streaking is most apparent in high contrast situations.
Use any of the DisplayMate Streaking test patterns to look for this effect.
LCDs may show both horizontal and vertical streaking.
19. White-Point Color Temperature:
The color temperature of the screen is generally set at the factory
to 9300° for most computer monitors and televisions.
This corresponds reasonably well to the color of "Bright White" paper in
office lighting.
Many monitors include end user presets for other values,
typically 5000° and 6500°, which are
useful for graphic arts and photography respectively.
Use a color analyzer to verify the accuracy of the settings.
Note that measuring correlated color temperature alone is
not sufficient to determine the white-point. The actual Chromaticity
Coordinates must be measured and compared.
20. Brightness and Contrast Controls:
Many LCD displays do not provide Brightness and Contrast controls
that have sufficient adjustment range.
Use the DisplayMate Black-Level, White Saturation,
and a 32-Step Gray-Scale test patterns to examine the operating range
of the controls.
| 
