Motion Blur Reduction Backlights
Simon Baker, with input from Mark Rejhon
21 March 2013 (updated 17 Sept 2013)

Twitter: Stay up to date: @TFTCentral |    Forum: Discuss this article  | RSS Feed: Article Alerts

Quick Browse:

 

 


Introduction

Manufacturers have been trying for years to improve the gaming experience for users of LCD displays. When they first became widely available and affordable as a desktop monitor solution there were certainly limitations in many areas with the technology. One area which has always been a focus for buyers has been the response time of the display, how quickly the individual pixels can change from one colour to another. For a long time this had a major impact on how the screen can handle fast moving scenes in movies and games and has for a long time been seen as one of the most important specs buyers should look at. Slower panels can suffer from issues like ghosting (trailing images) and high levels of motion blur and smearing. Over the years manufacturers have managed to drive down pixel response times significantly, helping to eliminate a large part of the problem. The advent of overdrive (Response Time Compensation) several years ago signified a step change in pixel response times, driving down grey to grey (G2G) changes and helping to ensure all panel technologies including IPS, VA and TN Film could offer decent response times where it was used well.

More recently manufacturers have improved gaming again with the arrival of native support for 120Hz (and now above) refresh rates. This has significantly helped improve smoothness and the perceived motion in practice for users, as well as allowing support for higher frame rates and 3D content.

One area which has not been as easy to address over the years is the sample-and-hold nature of LCD, which affects the way the human eye perceives the image from an LCD monitor. This is an area which has seen some investment over the years and in this article we will have a look at a new technique which is reported to greatly reduce motion blur from LCD's.
 

 
The Sample-and-Hold Problem

LCD displays are sample-and-hold displays and are continuously illuminated, unlike a CRT. The following high-speed camera video of LCD versus CRT being refreshed in real-time gives you a view of how the two differ:

 

Phosphor illumination is measured in microseconds, upon electron beam excitation.  However, phosphor decay takes much longer than that, and can affect ghosting. CRTs can have shorter persistence phosphor which can lead to dimmer images but reduces ghosting, or CRTs can have longer persistence phosphor which can lead to brighter images but more ghosting. On this specific CRT, the phosphor decays within 1-2 milliseconds (to the point where it is mostly dark), so the remainder of the CRT is dark for most of a refresh cycle. The CRT phosphor is many times brighter than the LCD backlight, but for a very short period. The scanning is cycled every refresh very rapidly, 60 times a second on most CRT displays (60 Hz), so it appears as a continuous image to the human eye.

Since LCD’s are a hold-type display the images are “held" on the screen for the entire refresh.  When your eyes are tracking a moving object on a display, your eyes are continuously moving.  This means that on a sample-and-hold display, your eyes are continuously moving across static refreshes, which creates the perceived motion blur. This applies even if response times were instantaneous (0ms) and even at 120Hz refresh rate.  With modern low response times at a pixel level (1 – 2ms), the actual speed of colour changes is now very fast. Higher 120Hz refresh rates also help drastically with the perceived smoothness of motion and aid the reduction of motion blur. However the issue of motion blur is still there as a result of eye-tracking based motion.


 
Motion Blur Reduction Techniques

    

Various techniques have been tested in the past by manufacturers in an effort to eliminate this perceived motion blur. Instead of using extra Hz, motion blur can also be reduced by inserting black period between refreshes, using various technologies such as Black Frame Insertion (BFI), scanning backlights, and/or full-screen stroboscopic backlights. Some televisions create extra Hz by using frame interpolation (e.g. 120 Hz HDTV's). This is the process of artificially generating intermediate frames, that are displayed between actual frames. However, this is not used for computer monitors due to the input lag and artefacts that frame interpolation creates.

Early attempts in reducing motion blur within computer monitors have included some ideas such as BenQ’s Black Frame Insertion (BFI) which was first discussed back in 2006. The idea was to insert intermediate black frames in between the normal frames to reduce motion blur. When models sporting BFI were introduced (e.g. BenQ FP241WZ) it seemed the technology had evolved and was instead a scanning backlight system instead, similar to that of a CRT. This refreshed the image at a certain interval from top to bottom to try and help eliminate the motion blur.

Other manufacturers also had a go at similar technologies, including efforts like Samsung’s Motion Picture Acceleration (MPA). We tested this in fact in the Samsung SM245T where the CCFL tubes were turned off in sequence one at a time. While the results are pretty subjective, we didn’t feel it helped much in practice really with moving images. On static images a visible flicker was introduced and contrast was impacted somewhat. We felt the same when we tested NEC’s version in their Motion Picture (MP) mode.

These early attempts at reducing motion blur were good on paper, but didn’t really deliver a workable solution which helped users in practice. Most commercially available displays with scanning backlights use a longer illumination duration than CRT phosphor due to insufficient brightness. In addition, many of these technologies do not produce a sufficiently long enough black period between refreshes, since CRT's are dark for a longer period than illuminated. Backlight leakage within scanning backlights can interfere with motion blur reduction - this is the leakage of light between on-segments and off-segments of backlight.  
 

  
The Theory - Ideal Scanning Backlights

To achieve decent motion-blur elimination with a scanning backlight system like this three main pre-requisites are necessary as outlined below. If all 3 were achieved then in theory you could have an LCD display which behaved far more like a CRT, and was able to deliver CRT-like motion quality.

1) Very Fast Pixel Response Times

Until recently pixel response times were not sufficiently low to eliminate pixel persistence factors from the motion blur equation. However with modern 120Hz LCD's manufacturers have needed to improve response times further to support their fast refresh rates and active shutter glasses operation.  LCD display manufacturers have been working hard to clean up as much pixel response trace as possible before the next refresh.  As a result, minor ghosting on such new LCD panels no longer leaks noticeably into the next refresh cycle. Thus, 3D-compatible panels are excellent candidates for scanning backlight designs. In theory the scanning backlight could be implemented during a ghost-free moment during the LCD refresh, once per refresh. As a result, pixel response would no longer be the absolute barrier for motion blur reduction.
 

2) Very Short Backlight Strobes With No Leakage Into Black Period

One very short strobe is needed for each point in the display per refresh, similar to the whole phosphor illumination-and-decay cycle in a CRT display. Different CRT displays have different phosphor decay cycle times. Common CRT computer monitors typically have approximately 1 to 2ms of phosphor decay. During a typical CRT refresh, a small fraction of a CRT, such as one-tenth (1/10) of the screen, is visibly illuminated at one time during a 1/60th second refresh. For a scanning backlight operating at the same scanning rate, it would need to be similar. The shorter the illumination is, the better the motion quality is.

Backlight diffusion also must be well-controlled, to prevent too much leakage of light from on-segments of backlight, leaking to the off-segments of backlight. A different solution to reduce or eliminate backlight diffusion from being a limiting factor to motion blur elimination, is to use a more rapid scanning backlight pass per refresh, or to use full-panel backlight strobes where the whole backlight is turned off at once, as opposed to a scanning backlight which turns it off in portions from top to bottom. Full backlight strobes are also more practical on newer LCD panels that do individual refreshes more rapidly. The important motion-blur elimination factor is that each point in the display is flashed as briefly as feasibly possible, only one strobe per refresh.
 

3) A very Bright Backlight

If you were to have a very fast scanning backlight with a high portion of off-time to on-time, the very short on-time will normally result in a darker image. The backlight needs to be very bright (at least 10 times brighter than normal) to compensate for the very short on-time.  Phosphors on a CRT illuminate extremely brightly for a very short time period - at least 10 times brighter than a typical LCD backlight. The extreme amount of extra brightness can be very expensive to engineer into a backlight system. Fortunately, the prices of LED's have fallen dramatically. LED's are well suited for scanning backlights due to their fast switching speed and brightness.

 

  
Strobe Backlights and LightBoost

We’ve already discussed the theory of scanning backlight systems and what would be required to produce a high quality image, greatly reducing motion blur issues. We also touched on the fact that a full strobe backlight, where the whole backlight is flashed on and off very quickly, could also be even more efficient and could work very well in this application.

The ‘LightBoost’ technology, found in new monitors supporting NVIDIA 3D Vision 2, is a form of a strobe backlight. This was introduced as part of NVIDIA’s 3D Vision 2 system to help boost the brightness of 3D content and improve performance in 3D gaming. Note that this helps improve brightness in 3D uses compared with older non-LightBoost capable displays. However, enabling LightBoost as a feature makes the screen darker than with the feature turned off which we will talk about later on in the article.  

An interesting secondary benefit has been studied in recent times with LightBoost. Studies have been conducted by Mark Rejhon at http://www.blurbusters.com/ based on using this system to reduce motion blur as well in 2D applications and games. In many respects, LightBoost manages to simultaneously meet all three critical criteria of an ideal scanning backlight design. It seems that the LightBoost system can be used as a modern-day strobe backlight and from many user testimonials, it seems to work very well. This is accomplished by turning off the backlight completely for the whole screen between refreshes, while waiting for pixel transitions. The backlight is strobed only on fully-refreshed frames, bypassing pixel persistence as the motion blur limiting factor. There is no backlight diffusion, unlike within a scanning backlight. In this case, strobe backlights can be superior to scanning backlights, and according to the work at BlurBusters can “allow LCD displays that have the motion clarity of a CRT.”

This is an unofficial technique for LightBoost systems but has been found to help reduce trailing, reduce motion blur and improve the overall gaming experience on LCD’s further. If it proves to be successful and widely used it is possible that manufacturers will find a way to implement this kind of system natively. We would like to see them revisit advertising this feature, as it may prove to be more popular than the original purpose for this feature (3D Vision).  In addition, vendors need to consider providing an easy method of enabling the stroboscopic backlight feature, via a hardware button or easy software toggle.

 

Tests have been conducted by Mark over at BlurBusters and a very high speed camera has been used to capture the results in slow motion. The following is a video of the Asus VG278H (note: not the newer VG278HE model) recorded at 1000 frames per second (fps).

 

You can see that the backlight strobing occurs directly between LCD refreshes.

Updated 19 June 2013

Using their own response time testing tool, similar to PixPerAn, BlurBusters were able to take some measurements of the level of motion blur visible on a monitor, running at different refresh rates and with LightBoost enabled and disabled. These UFO objects in the test image were moving horizontally at 960 pixels per second, at a frame rate matching refresh rate, on an ASUS VG278H LCD. These pictures were captured using a pursuit camera using a 1/30sec exposure (exposing multiple refreshes into same photo). Pursuit cameras are used by display manufacturers for testing (e.g. MotionMaster, and other MPRT pursuit cameras). This is simply a camera that follows on-screen motion. These expensive cameras are extremely accurate at measuring motion blur and other artifacts, since they simulate the eye tracking motion of moving eyes.


60 Hz Refresh rate:
Each refresh is displayed continuously for a full 1/60 second (16.7ms)


120 Hz Refresh rate:
Each refresh is displayed continuously for a full 1/120 second (8.3ms)
This creates 50% less motion blur.  This includes regular and overclocked 120Hz.


120 Hz with LightBoost
The backlight is strobed briefly, once per refresh, eliminating sample-and-hold.
With LightBoost, 120fps @ 120Hz has 85% to 92% less motion blur than 60Hz.
 

At 120fps@120Hz, a 1/30second camera exposure captures 4 refreshes. All 4 refreshes are stacked on each other, because the pursuit camera is moving in sync with the 120fps@120Hz moving object at a 1/30second camera exposure.

The brief backlight flash prevents tracking-based motion blur. There is extremely little leftover ghosting caused by pixel transitions (virtually invisible to the human eye), since nearly all (>99%+) pixel transitions, including overdrive artefacts, are now kept unseen by the human eye, while the backlight is turned off between refreshes.

Using a pursuit camera, BlurBusters were able to measure and compare the differences in motion blur between different refresh rates, with and without LightBoost enabled. In the best case scenario, a blur of 1.4ms (120Hz refresh with LB enabled) is 92% less than 16.7ms (60Hz without LB). This also corresponds to the amount of motion blur perceived by human vision, and pursuit camera photography.

The comparison graph was created from motion tests & oscilloscope measurements, including PixPerAn and Blur Busters Motion Tests, and cross-checked with TFTCentral oscilloscope tests (for 120Hz LightBoost = 100%) - see following section.  Accuracy +/- 0.1ms, assumes TN panel technology with optimized RTC, motion tests done with VSYNC ON with frame rate matching refresh rate.
 

    
Our Own Tests and Findings

We have recently reviewed the Asus VG278HE monitor (manufacturer rated 2ms G2G response time, 144Hz refresh rate support) which provides us with the opportunity to test this system for ourselves. We followed the process outlined later in this article to set up LightBoost activation in 2D desktop mode. We used the 'hack' method (see following sections) to test how easy it was to set up and following the steps worked fine and was quick and easy to activate. This allowed us to carry out tests with our oscilloscope system, PixPerAn and other visual tests outside of games. The screen is tricked into thinking its running in 3D mode when it isn't, but LightBoost remains active from the backlight. LightBoost only works with refresh rates between 100Hz and 120Hz (not the full native 144Hz support) since 3D and LightBoost is designed to work at 100Hz through 120Hz at the moment (e.g. 120Hz results in 2 x 60Hz, for each eye).

On the VG278HE once you activate this tweak the desktop changes appearance slightly. The image appears a darker and colours have changed a little to the naked eye. The brightness control from the OSD is not available any more and is instead a control over the LightBoost intensity. This controls how bright the display is and can also have an impact over the motion clarity.

An easy way to tell that this is activated on the VG278HE is the appearance of a "3D-mode" red label at the top of the OSD menu as shown above. You will notice that most of the other OSD options are now disabled when LightBoost is active as well.

The LightBoost control option is now available as well within the OSD 'Color' section. Contrast is still available as an option, but colour temp, RGB channels etc are now greyed out.

Interestingly the Trace Free option is now locked as well when LightBoost is enabled.
 


Oscilloscope Tests

We took some measurements with our oscilloscope and photosensor equipment. Above is the resulting oscillograph with LightBoost enabled at its maximum setting. Each horizontal grid unit is 20ms. We can confirm from this that the LightBoost is operating at a 120Hz frequency in sync with the refresh rate of the screen.

If we zoom in on the scale a little you can see the peaks and troughs more clearly. Each horizontal grid unit is 5ms in this graph now.

The base line of the graph is the time when the backlight is turned off by the strobing of the LightBoost backlight system. The upper peaks are the time for which the backlight is turned on for each cycle.

As you lower the LightBoost brightness setting from the OSD menu, the upper peaks become shorter, as the backlight is on for shorter amounts of time (accounting for the reduction in the perceived brightness). The period for which the backlight is off becomes longer. The feature still operates at 120Hz, but there are now longer dark periods and shorter bright periods (only be a few milliseconds keep in mind).

It is possible to measure the length of time for each "on" period and each "off" period at each of the LightBoost settings. For the purposes of this we can zoom in on the scale of the graph to allow for more accurate measurements as shown below.

On the above graph each horizontal grid is now only 1.25ms. We measure from the point at which the brightness starts to change and increase (red line) to the point at which it changes again and decreases (blue line). This is the "on" period.

Likewise we can take the measurement for the "off" period as shown above. The length of time at each of the 10 LightBoost settings is listed in the table below. Keep in mind this is an example based on the Asus VG278HE screen.

LightBoost Setting
(Higher is Brighter)

Time "on"
(ms)

Time "off"
(ms)

100%

2.375

6.000

90%

2.250

6.000

80%

2.250

6.125

70%

2.000

6.250

60%

2.000

6.375

50%

1.875

6.500

40%

1.750

6.500

30%

1.625

6.625

20%

1.625

6.750

10%

1.500

6.875

0% (Off)

2.500

5.875

As you can see, when you lower the LightBoost setting each step results in either a shorter "on" time, a longer "off" time or both. This leads to a progressively darker image, but also clearer motion if you are more sensitive to motion blur. We will look at the motion blur tests in a moment. The reduced brightness may be an advantage or disadvantage, depending on your needs.  For people who prefer a reduced brightness in an excessively bright LED monitor, it also provides another means of dimming the display. The lower LightBoost settings produce a much dimmer image than non-LightBoost mode configured to a brightness setting of 0%. Although LightBoost is a stroboscopic backlight, there are no PWM motion artefacts to create a distraction, due to the single-strobe-per-refresh nature of LightBoost. This is an advantage if you are not sensitive to PWM or CRT flicker, but is sensitive to the PWM artefact problem.

LightBoost also functions at 100Hz.  The strobe lengths are proportionally longer during this mode (about 20% longer).  However, 100Hz LightBoost is not tested in depth for this article.

At the lowest LightBoost (LB) setting in the menu it seemed that the brightness boost was quite significantly reduced. You could notice a drop in this to the naked eye as well and you can see that the peaks in the graph are now much lower. LightBoost is turned off but the backlight is still being strobed on and off. The grid line here are 5ms, and there is a 2.5ms "on" time and 5.875ms "off" time. While the "on" time is longer than LB = 10%, the voltage is lower on the graph signifying that the actual brightness is lower.  There seems to be no real reason to use LightBoost="OFF" instead of LightBoost="10%"; since the shorter strobes of LightBoost=10% results in the clearest possible motion.

 


Motion Blur Tests

Now that we've looked at how the LightBoost (LB) backlight is working we wanted to test the impact it had on motion blur in practice. We used some of the tests in the PixPerAn tool to help here.

With LB enabled you could certainly notice a difference in practice, even in the basic moving car test of PixPerAn. The image moved as smoothly and quickly as it had before (in normal 120Hz mode without LB enabled) but tracking of the image with the naked eye was much easier. As you followed the image it remained sharper and more easy to follow, and there was less smearing and blurring to the naked eye.

As a side issue which is linked to this specific screen it does seem that with LB enabled the overdrive impulse seemed to have been turned up to its maximum level. We no longer had access to the Trace Free setting in the OSD to turn this back down, and so the moving image did show more overshoot than we'd seen in our other tests (with TF of 60). This was a result of the pixel response times and overshoot caused by the aggressive overdrive impulse necessary to synchronize the LCD panel with the strobe backlight. The default contrast setting on the VG278HE was 80% and to compensate for the overdrive artefacts during LightBoost, we you can reduce the contrast down to 60% which helps to eliminate some of the overdrive artefacts, but also reduced the screen brightness. In fact from our review we'd already changed to a contrast setting of 70% during the calibration stages. Remember that if you adjust contrast, you may also need to re-calibrate your display. Of course the artefacts here are a result of the specific panel and the overdrive impulse being used, and so will be different from one display to another anyway.

Even regardless of the overdrive artefacts, the actual perceived motion blur was much less we felt. We were suitably impressed, to be honest. User reports are that the LightBoost motion blur reduction is even better on some other compatible screens and the VG278HE shows some faint after image artefacts as we've seen above, although somewhat reduced if contrast is reduced specifically on the Asus VG278H/HE. Even though these artefacts were visible, the improvement to perceived motion blur was significant to our eyes and we were pleased with the results.

The shorter strobe lengths at lower LightBoost settings produced a slight further improvement in motion clarity. Some subjective observations can be made, but may depend on how sensitive you are to motion blur. We visually tested the 100% LightBoost setting versus the 10% LightBoost setting, to get the biggest possible comparison. In PixPerAn, this is just about noticeable in the "I NEED MORE SOCKS" text when scrolling at high tempos (Tempo 8 and higher). At LightBoost=100%, the text is fully readable (which can't be said for non-LightBoost LCD) but retains about one pixel's width of motion blur. At LightBoost=10%, the pixels are clearer; the one-pixel gap between the letter "N" and "E" in "NEED MORE SOCKS", becomes very sharp. For most people, it is extremely difficult to notice the difference between different LightBoost settings except for brightness. However, some users reports have remarked they clearly see a difference, while others could not. At this point, we are starting to reach diminishing point of returns that depends on the person. A compromise LightBoost setting value can be chosen, such as 50% or 60%, as a balance between brightness and motion clarity.

Progressing into some of the other tests in the software we tried the 'readability test' in which a row of random letters is scrolled across the screen. You can control the speed of this and the idea is to identify the fastest scrolling setting possible while the text remains readable to the naked eye.

First of all we tried this test at 120Hz without LightBoost. The maximum level we could reach sensibly was 8 in this test on the VG278HE. After this, anything faster just became a blurry mess and you couldn't focus on the text or make out the individual letters as you tracked the movement across the screen.

With LightBoost enabled there was another marked improvement and again we were impressed. Even at the maximum setting of 30, you could pick out the individual letters clearly as they scrolled across the screen and as you tracked the movement with your eyes. It was moving too fast to really tell what they were, but they were visible and not lost in a blur. Some users have reported that they were able to pass a PixPerAn score of 30 completely, by also turning their head while reading the text, especially when using LightBoost=50% or less. The backlight strobing did seem to be doing a very good job of making images sharper during motion and reducing blurring. Note: the above image is just an example of the text clarity at a setting of 30.
 

We also conducted some motion blur tests using the BlurBusters motion test tool (coming soon). This includes a moving image test a bit like that in PixPerAn as shown above. With the screen running at 120Hz without LightBoost we saw smooth moving images running at a high 120fps frame rate. However without LightBoost there was still noticeable motion blur to the user. When LightBoost was enabled the blur was reduced dramatically again, and it was much easier to track the moving object across the screen where it remained sharper and clearer as well. Comparing the LightBoost settings (100% vs. 10% again), it is also observed that the three pixel eyes of the alien in the moving UFO at 960 pixels/second have about one pixel of motion blur at LightBoost=100%, while the three eyes of the alien are razor-sharp at LightBoost=10%.  



Colour and Image Tests

We had already observed the visual change in the image when you first switch the LB feature on. We took some measurements and reports using the X-rite i1 Pro spectrophotometer with LB disabled and enabled but after a calibration had taken place. First of all we had calibrated the screen based on the steps and settings in our VG278HE review.


Asus VG278HE - Calibrated Settings, LightBoost OFF

 

  
 

 

Calibrated Settings
LightBoost Off

luminance (cd/m2)

120

Black Point (cd/m2)

0.15

Contrast Ratio

799:1

As you can see this was the optimal calibrated state with gamma, white point, luminance and colour accuracy all corrected nicely. We had a static contrast ratio of 799:1 as well after this process. We then switched LB on and took a measurement again without any further calibration process being completed.


Asus VG278HE - Calibrated Settings, LightBoost ON

 

 

Calibrated Settings
LightBoost ON

luminance (cd/m2)

88

Black Point (cd/m2)

0.13

Contrast Ratio

670:1

After LB was enabled the image became noticeably darker (even with the LB setting on maximum) and appeared to be a slightly different temperature. This was confirmed with the i1 Pro where luminance had dropped a fair amount to 88 cd/m2. White point was also cooler now at 7777k, moving it 20% away from the calibrated target of 6500k. Contrast ratio was also impacted somewhat with a figure of 670:1 now achieved. It seems that with LB enabled the strobing backlight causes:

1) A reduction in screen luminance
2) A slightly cooler colour temperature by about 1277k (20%)
3) A reduced contrast by about 129:1

We wanted to look at the luminance range a little more with LB enabled so took measurements of luminance, black point and the resulting static contrast ratio at each setting as you got progressively brighter.

LightBoost Brightness Setting

Luminance
(cd/m2)

Black Point (cd/m2)

Contrast Ratio
( x:1)

Off

48.21

0.08

603

10%

51.52

0.08

644

20%

55.22

0.09

614

30%

59.00

0.09

656

40%

62.64

0.10

626

50%

66.18

0.10

662

60%

69.48

0.11

632

70%

73.13

0.12

609

80%

76.66

0.12

639

90%

80.23

0.13

617

100%

87.14

0.13

670

You can see the the luminance range of the screen is now much lower with LB enabled, ranging up to a maximum of only 87.14 cd/m2 and with an adjustment range of only 38.93 cd/m2. The average contrast ratio was 634:1. The limited brightness range of the screen was an obvious limitation of enabling LB.

 


What You Need

To take advantage of this technology for your own gaming needs you will need a few ingredients.

 

1) Monitors – you need an LCD monitor which supports the LightBoost technology. Most new NVIDIA 3D certified monitors now have LightBoost support and it is normally listed in the spec and product description. From user reports, the best-performing LightBoost monitors are the 1ms rated monitors, because of reduced crosstalk between refreshes. Ghosting tests have shown less than 0.5% leakage of frames between refreshes; and there is less 3D crosstalk on these panels than for polarized 3D.

Models known to feature LightBoost at the time of writing are:

Updated 17 Sept 2013

2) Graphics Card - Presently, you need an NVIDIA graphics card which can support 3D Vision, because LightBoost is an NVIDIA technology. You should preferably have a GeForce GTX 680 or faster; something that can do at least 120fps at 120Hz in many games. The technology can work with lower end graphics cards of course but frame rates would not be optimal in fast gaming. The LightBoost effect diminishes and disappears at half frame-rate (60fps @ 120Hz).  Repeated refreshes lead to a momentary sample and hold effect, so frame rates should ideally at least match the refresh rate in order to minimize motion blur. LightBoost also functions at 100Hz as well, to allow 100fps at 100Hz which is easier for newer video games with more demanding graphics.

 


 

3) NVIDIA 3D Vision kit – you don’t actually even need the 3D Vision kit (glasses and emitter) to get the LightBoost working, but of course if you wanted to play any 3D gaming then you would need that! It does also make life a bit easier since it is easier to enable LightBoost if you have the emitter.

 


Samsung 120Hz Monitors Note

 

Samsung's 120Hz monitors have gone the other route for 3D gaming, being aimed more at AMD users and not being certified for NVIDIA 3D vision. However, some of these screens do feature a LightBoost-style strobe backlight feature, although this is not advertised within the spec. Most Samsung 700D, 750D and 950D models seem to have this feature (BlurBusters have tested the S23A700D, S23A950D, S27A750D, S27A950D) so can be used in a similar way to reduce motion blur. The good news is that the Samsung strobe backlight works with both AMD Radeon and NVIDIA graphics cards. The bad news is that according to reports the input lag is worse than official LightBoost monitors, so this is not as useful for competition online gaming as official LightBoost monitors. See BlurBusters for more information about Samsung monitors and how to enable the strobe backlight feature for 2D blur reduction.


 

 
How to Enable LightBoost

There are two different sets of instructions, depending on whether you have the NVIDIA 3D glasses and emitter or not. If you have a 3D glasses emitter, then you can follow instructions that are already provided by the vendor.

Option 1 - Standard Setup

Choose these instructions if you have a LightBoost monitor, and you have obtained the shutter glasses emitter, or it’s built into your monitor (e.g. Asus VG278H). The procedure is simply to set up stereoscopic 3D normally, launch the game in stereoscopic 3D, and then use a ‘Control + T’ command in a game to turn off stereoscopic operation.  At this point, you have LightBoost in 2D and you can play normally with no glasses and with the benefits we’ve already discussed. The detailed procedure is as follows:

  1. Install the latest graphics card driver from NVIDIA

  2. Make sure your glasses emitter is connected to your PC and installed

  3. Go to Control Panel > Display > Adjust Resolution.

  4. Verify “Enable Stereoscopic 3D settings for all displays” is enabled.

  5. Go to NVIDIA Control Panel (system tray > NVIDIA icon), and select “Setup Stereoscopic 3D” from the left hand menu.

  6. Click “Test Stereoscopic 3D” and follow the vendor instructions to test it, at a refresh rate of 120Hz.

  7. Start a video game. The game should be running in stereoscopic 3D.  This makes sure LightBoost is running, since it is automatically enabled during stereoscopic 3D.

  8. Hit Control + T inside the video game to turn off stereoscopic operation. This is an official key-press combination provided by official NVIDIA graphics drivers.

  9. The LightBoost strobe backlight is still enabled, and you don’t need the 3D glasses as the 3D function is turned off.

Disadvantages:

 

 
Option 2 – Unofficial ‘Hack’ Method -
Updated 19 June 2013

There is also a method you can use to trick your system into thinking you have an emitter connected and without the need to launch into 3D games first. We tried this method and found it quick and easy to set. We won't go into the details here but you can follow the steps described at BlurBusters for more information.

 

A useful walk through video has been provided above. This 'hack' method has recently been updated and improved and now requires no registry tweaks, .reg or .inf files. A new LightBoost utility is in the works and expected to be available in July which will make LightBoost usage even simpler, allowing you to enable and disable LightBoost with a simple key-press. Keep an eye on BlurBusters for more information.

 

  
Strobed Backlights for Other Uses

Scanning and strobed backlights are also found in many newer HDTV's. However, many of them create far more input lag than computer monitors due to them combining motion interpolation along with scanning backlights. That said, a scanning backlight does benefit live fast-motion material. There is much less motion blur elimination for movies and low frame rate material (non-interpolated) though.

The best material for scanning and strobed backlights, is material that has a frame rate matching the refresh rate, and contains no source-based blur. This include fast live action taken using video cameras utilizing a fast shutter speed and without over-compression (which can introduce blur). Examples include fast pans in sports like football, hockey, NASCAR racing, ski racing, bike racing, Red Bull air races, etc.

Video games running at high frame-rates also benefit significantly, as already covered in this article. Other computer operations, such as smooth-scrolling of text and web browsers, can also benefit noticeably from a scanning/strobed backlight that allows crystal-sharp moving text.  It is also worth noting that LightBoost has far less input lag than the vast majority of these existing HDTV-based scanning backlights.

 

   
Disadvantages and Issues

While scanning backlights and the new LightBoost strobe backlights can certainly help with motion blur reduction, there are some issues which go along with them which you should be aware of.

Flicker

Because the backlight is being constantly strobed on and off there is of course the chance that the user will see flicker from the screen. As you switch the feature we detected a faint flicker to the naked eye but it wasn't particularly obvious or distracting we felt. For general desktop and day to day use you may not want to have LB enabled as it is not always useful at the Windows desktop. For games the moving images are changing so fast that you're unlikely to see any flicker anyway and the benefits of the reduced motion blur are likely to out-weigh any flicker.

Reduced Brightness

This may or may not be a disadvantage, depending on whether you were already reducing the brightness of your monitor in the first place. As we've already seen from our tests, the brightness of the display is impacted quite significantly when LB is enabled. Even at the maximum LB setting we only achieved a luminance of ~87 cd/m2 from the Asus VG278HE. Without LB enabled the backlight could be controlled up to a much higher 263 cd/m2. Some users (including 3D Vision Blog) have found alterations to the contrast control can help somewhat so it's worth experimenting if you are using one of these screens and want to try the LB feature for motion blur reduction. To reduce motion blur, the dark periods of the image need to be sufficiently long and in doing so, the overall perceived brightness is greatly reduced. It is probably still adequate for most users to be honest, and we expect that as LightBoost 3D displays develop we will see brighter backlight systems used which will help overcome this limitation more and more. Also keep in mind that contrast ratio seems to be impacted by a small amount when LB is enabled.

Impact to Colours

We've already seen from our tests that LB seems to impact colour temperature but about 20% (on the VG278HE at least), making the image overall a bit cooler. Furthermore without access now to the OSD controls in most areas, this may be hard for some users to correct. Picture adjustments are still accessible in NVIDIA Control Panel, and calibration with a colorimeter tool can help overcome this and correct offset gamma and white points, but if you're wanting to switch between LB off and on for different uses, it may become a pain having different profiles required. Some manual by-eye calibration using various online or software based tools may help get a screen looking more to your liking perhaps.

 


Further Reading 

 


Conclusion

We went into this article with an open, but quite sceptical mind about motion blur reduction backlights. We'd tested several scanning backlight systems in the past and never been very impressed at all. We've talked in this article about the different techniques used and what a good scanning backlight system would need to look like to offer decent results, and older attempts in the desktop monitor market had certainly never really delivered in our opinion.

We expected similar mixed results from the new LightBoost method which we'd read about with interest. We weren't expecting much but all in all we were suitably impressed to be honest with the performance we saw. There's a few limitations with the technology in its early days at the moment (limited brightness and possible flicker mainly), but there are certainly some big benefits for anyone wanting to improve their gaming experience on an LCD screen. We were very impressed by the improvements it brought to moving images, providing greatly reduced motion blur in practice and making eye-tracking much clearer and easier. This brings obvious advantages for any kind of gaming or fast moving content, where the hold-type nature of LCD's has really been a problem for manufacturers for a long time. We always recommend fast response time screens for high end gaming, preferably with 120Hz+ refresh rates to bring about obvious smoothness and frame rate benefits. The fact that many of these screens support stroboscopic backlights is an added bonus and really does add another level of performance improvements. If you've got a LightBoost capable monitor, give this method a go as in our opinion it really is worth it.
 

If you have enjoyed this article and found it useful, please consider making a small donation to the site.