Experimental findings on LCD response times reveals that it’s a non-issue.

Every display technology has its own set of unique strengths, weaknesses, and limitations. While each technology steadily improves over time, our impression of their initial weaknesses and limitations often persist, and can turn into demons that plague the technology forever. The best example of this was the so-called “burn-in” problem of plasma displays. This was technically overcome years ago, but the perception lingers like an 800-pound gorilla that still threatens to kill this excellent display technology.

LCDs have their own 800-pound gorilla, and it is limited “response time,” which causes motion blur. Just like plasma’s albatross, this was a significant problem for LCDs many years ago, but our research has found that it is no longer an issue. But unlike plasmas, LCD manufacturers have turned this into a brilliant marketing strategy, offering ever more sophisticated and enhanced motion processing, and ever higher 120 Hz and 240 Hz screen refresh rates to repeatedly sell a solution to a problem that is no longer a problem.

Consumers (especially the technically savvy ones) have become enthralled with the response time specifications and the various proprietary motion enhancement technologies offered by each manufacturer, which all spiral in a vicious cycle of one-upmanship. Unfortunately, none of this stands up to objective scientific testing. As we’ll demonstrate below, while the motion blur performance with moving test patterns was much worse than what’s claimed in the manufacturer’s specifications, the motion blur performance during the extensive viewing tests with a wide range of live video content viewed simultaneously on a large number of HDTVs surprised us by turning out much better than expected. In fact, motion blur proved to be a non-issue for live video in all of the mid to high-end LCDs in our tests.

A static photograph (at left) is compared to the same image moving across the screen at about 1,000 pixels per second. Photo by Lauren Soneira

LCD RESPONSE TIME AND MOTION BLUR
Motion blur is a well-known issue with LCDs. It happens when the liquid crystal, which is the active element within an LCD, is unable to change its orientation and transmission rapidly enough as the picture changes from one frame to the next. Since the standard video rate is 60 frames per second, a pixel is expected to be able to fully update its light transmission opacity within 16.7 milliseconds. If it takes any longer than that, the image will show some degree of lag, which appears as a trailing smear or blur whenever there is motion. It also affects the visibility of the leading portions of moving objects.

LCD motion blur is generally evaluated with an industry standard specification called “response time.” Unfortunately, it’s not a particularly good indicator for real picture blur because it measures the time it takes for a pixel to go from black to peak intensity white, and then back to black again. However, most picture transitions involve much smaller and more subtle shades of gray-to-gray transitions, which can take considerably longer to complete. On the other hand, blurring is much harder to detect visually when there are subtle gray-to-gray transitions because the initial and final states are so similar. But it’s even more complicated than that, because every pixel is actually made up of independent red, green, and blue sub-pixels that have their own separate intensities and frame-to-frame transition times. So visual blur within a picture that has some motion is a fairly complex and nebulous phenomenon.

Response time and motion blur depend on many factors, including the viscosity and thickness of the liquid crystal layer. Many different methods of electronic processing are used to try to speed up the pixel-to-pixel transitions. One common method is to temporarily exaggerate the drive voltage used during a transition, which is called overdriving. It’s sort of like giving the liquid crystal an extra hard kick in the pants to get it moving quickly. The problem is that it’s very hard to give just the right amount of kick for all possible transitions, and that leads to overshoot, inverse ghosting, and image flicker. Another method is edge enhancement using high frequency peaking. And then there is the one that all of the manufacturers are currently touting — updating the screen more frequently by increasing the refresh rate from 60 Hz to 120 Hz, or higher.

Because the published specifications can have a considerable impact on sales, it is often more important for a manufacturer to reduce the black to peak white to black response time value rather than improving the visually more important gray-to-gray transitions or reducing the motion artifacts that result from electronically pushing the response time. As a result, the LCD display with the fastest response time specification may not have the smallest visual blur. This was the case in our tests.

THE SHOOT-OUT
In 2009, an in-depth scientific study was conducted at the DisplayMate Technologies Demo Lab. Some of the results on display viewing angles were published in the July 2009 issue of AV Technology. The study included 11 top-tier branded HDTVs: eight LCDs, two plasmas, and one CRT Sony Professional HD Trinitron Studio Monitor, which was used as the reference standard. Our study included precise calibrations, comprehensive spectroradiometer measurements, and a large number of jury panelists that viewed test patterns, test photos, and lots of high-quality high definition video material. The shoot-out was jointly produced by DisplayMate Technologies in collaboration with Insight Media; however, all of the technical analysis was done by the author.

Two of the LCD units had 120 Hz screen refresh rates, one of the others used strobed LED backlighting, and all of the remaining units had a standard 60 Hz screen refresh rate. The goal was to determine the degree to which this varied advanced technology affected visible motion blur.

MOVING TEST PATTERNS
The first step in evaluating motion blur is to use specialized moving test patterns in order to cleanly examine and analyze the blur and related artifacts. To generate the moving test patterns and photographs we used DisplayMate Multimedia with Motion Bitmaps Edition, which includes 25 proprietary motion test patterns and 35 test photos that can be moved in different directions and speeds on screen. This digital video was fed simultaneously to all of the displays, which were compared side-byside in a “shoot-out.”

A test pattern, moving at 1,609 pixels per second, shows the 120 Hz refresh cycle of this display, as well as white “tips” on edges and fine detail, which are artifacts resulting from the electronic processing enhancements used to reduce the response time.

The differences in motion blur between the eight LCD displays was not large. The reason is that the visible blur was considerably longer than the 60 Hz video frame rate, so it didn’t matter whether the screen refresh rate was 60 Hz or 120 Hz.

To illustrate the motion blur results we took screen shots of the “best performing” LCD (in terms of picture quality and accuracy). The manufacturer’s published response time for this model is 8 ms. Since this corresponds to a double transition (from black to peak white and then back to black again) the single transition time (from black to white or from white to black) should therefore be about 4 ms. However, for the test patterns shown in figure 1, a live view by eye clearly detected blurring out to considerably more than 60 ms.

The differences in motion blur using a moving test pattern between the eight LCD displays was not significant. There were only minor differences between all of units. The reason is that the visible blur was considerably longer than the 60 Hz video frame rate, so it didn’t matter whether the screen refresh was 60 Hz or 120 Hz, or whether the LED backlights were strobed off during the frame updating. Similarly, varying the electronic processing enhancements that some models offer (which are supposed to reduce motion blur) only served to introduce objectionable contours, edges, and other artifacts onto moving objects without reducing the overall motion blur. In addition, there was no notable difference in motion blur between the top-of-the-line models (which all had 120 Hz refresh or LED strobing) and the mid-line models (which all had standard 60 Hz refresh), and which cost less than half of the high-end models.

MOVING PHOTOGRAPHS
Test patterns are perfectly valid images, no different in principle from any other image or picture content that is displayed. Absolutely every effect, artifact, and defect that you see in any test pattern also appears in every image and picture. The difference is that test patterns are often constructed to maximize the visibility of specific effects, artifacts, and defects.

Photographic images, on the other hand, typically consist of a very complex and varied admixture of blended picture elements. With so much going on in an image, motion blur is easily obscured and lost within the complex variegated imagery of a typical photograph. In particular, photographs lack the uniform backgrounds used in the test patterns, which make it easier to see the motion blur trails. Still, from what we learned with the test patterns, we expected to be able to see the effects of motion blur most easily when there are sharp bright-to-dark, or blackto- dim transitions or strong-to-weak color saturation transitions.

Moving photographs are, nonetheless, moving static images, which are quite different from live video, where the images are part of a complex and varied mixture of continually blending picture components that are themselves constantly changing in both time and position. The closest thing to moving photographs are the news and stock tickers on some television stations, and the vertical title rolls at the end of most movies.

LIVE VIDEO
While moving test patterns and photographs are very interesting and enlightening for studying motion blur and artifacts, live video is what most people actually watch on their displays. With even more screen activity going on we expect to visually detect much less motion blur in live video than with either the moving static photographs or test patterns. The first issue to consider is that there is no such thing as typical live video because of the incredible variety and diversity of content. Fortunately, we know from the moving test patterns and photographs what kinds of picture content are most likely to produce visible motion blur. And, of course, there needs to be lots of on-screen action.

Most of the live video sources that we chose were sports-based because they have lots of action and most have brightly colored uniforms and background content. We recorded them on an all-digital high definition TiVo from full bandwidth over-the-air ATSC broadcast television. It directly records the original broadcast mpeg data stream without any processing or degradation. Note that satellite and cable video sources have reduced signal bandwidth that introduces additional motion artifacts because of the extra compression needed whenever there is motion in the picture. We also didn’t use any film-based content, because it’s shot at 24 frames per second and requires considerably more interpolation and motion processing than video cameras with 60 fields per second.

LIVE VIDEO MOTION SHOOT-OUT
One important issue for live video, as opposed to the previous precision computer-generated moving photographs and test patterns, is that they are all shot from video cameras under varying conditions and may have unknown degrees of video processing. That can result in source video that is blurred with varied artifacts. We certainly didn’t want to blame an LCD for a blurred or defective picture when the cause was in the source. In order to carefully monitor the quality of the source video we used a Sony Trinitron Professional HD Broadcast Studio Monitor, which is a CRT that did not exhibit any visible motion blur or artifacts (except for barely visible tiny phosphor trails seen only in fast moving test patterns). So, whenever there was questionable content we carefully evaluated it on the CRT monitor. The shoot-out was fully operational for several months, so we had lots of time to study and compare all of the effects, and over that period of time we had several dozen people come by to see it running and evaluate the effects themselves, including industry experts, manufacturers, engineers, reviewers, journalists, and ISF instructors, all of which are trained observers.

All of the displays were fed identical simultaneous digital video from the content list above using the digital TiVo or Blu-ray player. They were all compared side-by-side in a shoot-out configuration. If any viewer thought they detected motion blur on any HDTV we would repeatedly press the 8-second TiVo backup button and watch the sequence over and over again on all of the units (including the CRT monitor) until we fully understood exactly what was happening on each HDTV. We did the same thing with the Blu-ray player and its content.

The conclusions from everyone that participated in the shoot-out were consistent across the board: there was essentially no visually detectable motion blur on any of the LCDs in all of the extensive live video content that we assembled. When people thought they saw motion blur, with only a handful of minor exceptions, the blur was either in the source video or a temporary visual illusion that disappeared when the segments in question were reviewed. Unlike the moving test patterns and moving photographs, the eye is unable to detect the blur in live video because the images are much more dynamic and complex, and undoubtedly because of the way the brain processes and extracts essential information from visual images. The results were identical for all of the LCDs, regardless of whether they had 60 or 120 Hz refresh rates, strobed LED backlighting, or advanced motion enhancement processing.

SUMMARY
Response time specifications are not a scientifically accurate or meaningful indicator of picture blur. In fact, in our tests the LCD with the shortest published response time had the greatest visible motion blur. You’ll see published values down to as little as 2 ms, but the motion blur we measured with moving test patterns on the topof- the-line displays was over 40 ms, which is more than a factor of 10 greater than the manufacturer’s specifications.

We also found that the proprietary motion enhancement processing technologies provided in most HDTVs actually just introduce ugly motion artifacts into the image rather than reducing the overall visual motion blur. The best picture quality was obtained with the motion enhancement processing minimized or turned off.

These results and conclusions will surprise many technically savvy consumers and videophiles because there has been so much talk about response time and motion blur. Like plasma “burn-in,” some of this is just old information and memories. It’s also very easy to think that you see blur when you’re looking at lots of fast action on a single TV, and a lot of it undoubtedly has its origins in the human visual system. It just doesn’t stand up to the extensive scientific side-by-side testing that we’ve described here.

Our most important and significant conclusion is that the LCD manufacturers have finally beaten the motion blur problem. So it’s time for both consumers and manufacturers to forget about this tamed 800-pound response time gorilla and focus on much more productive and fascinating display technologies, such as the upcoming generations of 3D displays. If you stick with the mid- to top-tier models from the reputable brands, you can ignore response time specifications, not worry about LCD motion blur, and don’t spend extra for 120 Hz or higher refresh rates, strobed LED backlighting, or advanced motion blur processing.

Dr. Raymond Soneira is president and founder of DisplayMate Technologies Corporation of Amherst, NH, and a research scientist with a career that spans physics, computer science, and television system design.