Improved scanning backlights coming? Now possible for LCD to have less motion blur than CRT! - AVS Forum
Forum Jump: 
 
Thread Tools
post #1 of 8 Old 09-13-2012, 05:27 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
Hello!

For high end LCD HDTV's, manufacturers utilize scanning backlights (Sony XR 960, Samsung CMR 960, Elite HDTV etc), usually in conjuniction with motion interpolation, to simulate "960Hz" clarity of motion from a 60 Hz signal. This greatly improves quality of motion during fast-action sports. However, I'd like to see scanning backlights be done without motion interpolation (At least as a configurable option). Removing motion interpolation from the scanning backlight equation, should be a configurable option (in OSD menus) that improves image quality, for times such as playing FPS video games, or watching fast-action sports.

Now, I've recently done research and found that it's technically possible to build a home-made scanning backlight for relatively cheap -- (using an Arduino and ribbons of LED tape), hacking open a cheap HDTV or monitor and replacing its backlight, one Arduino pin connected to VSYNC, and a software adjustment for timing (adjust phase of scanning sequence to match input lag). No other modification of the monitor's electronics is needed, the Arduino only needs to know VSYNC timing, for a home-made scanning backlight that allows an LCD panel to have less motion blur than CRT.

Historically, in the past, some computer monitor makers have attempted scanning backlights. For example, BENQ's AMA technology in an over-priced display:
ama_diagram.jpeg
(Refrence: TFT Central).
However, this does not improve motion blur very much because it's bright more often than dark. Observe the graph image above -- the dark period is extremely short.

Today's high end LCD HDTV's -- e.g. Elite HDTV, Sony CMR 960, etc -- the ones that simulate "960Hz" out of a 240Hz frame-interpolated signal, are much better than the above graph today. These displays use a 75%:25% dark:bright scanning mode in a backlight, where the LED's of the backlight is dark 75% of the time, and bright 25% of the time, to reduce motion blur by an additional 75%. The LED's in these displays need to be incredibly bright to compensate for longer dark periods. These existing high-end HDTV's simulate "960" by using 240Hz motion interpolation, combined with the scanning backlight (75%:25% dark:bright) to multiply that by another 4 (i.e. reducing motion blur by another 75%)

However, we've now reached a technological point where it's now possible to eliminate 90%-95% of motion blur from an image, using suffiiently bright LED's (e.g. 10 times brighter than is required for continuous illumination). We thus, now approach the possibility to have LCD superior to CRT in motion-quality elimination -- and WITHOUT using motion interpolation (which is hated by some).

Summary:
1. LCD's are now fast enough to finish refreshing before the next frame (requirement of 3D LCD's)
2. LED's are now bright and cheap enough (requirement of extra brightness needed in ultra-short flashes in scanning/strobed backlight)
3. Today's 120Hz LCD's, means that flicker of a scanning-backlight, will not bother _most_ people. (3D LCD's brought us 120Hz LCD's)
4. Controllers for scanning backlights are now cheap (it can be done with an Arduino)
5. Scanning backlights make it possible for LCD blur to be *better* than the LCD's own response speed.
6. Finally, scanning backlights can be made configurable in on-screen menus (e.g. turn off scanning and make it behave conventionally), if you only want to use it during videogames or fast-action sports. ("Game/Sports Mode") if you dislike CRT-style flicker at other times. One could still include motion interpolation, but permit it to be turned off during game mode, while keeping the scanning backlight.

LCD blur can be less than CRT, since LCD pixel response no longer matters (once you meet pre-requisite #1 above)
For example, a single 8ms refresh (1/120th second) for a 120Hz display, can be enhanced with a scanning/strobed backight:
2ms -- wait for LCD pixel to finish refreshing (unseen, while in the dark)
5ms -- wait a little longer for most of ghosting to disappear (unseen, while in the dark)
1ms -- flash the backlight quickly. (1/960th or 1/1000th second -- or even 1/2000th second!)

Here, you just bypassed LCD pixel response as the factor in motion blur. Heck, you could strobe for only 0.5ms instead of 1ms -- and you get sharper motion on an LCD than on a CRT, imagine that! The scanning backlight flashes briefly once per frame (e.g. a single 1/1000th second flash or 1/2000th second flash per frame, per backlight segment). The scanning of a scanning backlight illuminates already-refreshed LCD pixels, while waiting for other parts of the LCD to refresh. Flicker fusion and persistence of vision (exactly the same principle as using a CRT display), does the rest for you -- motion is crystal sharp. A 1/2000th second flash (one brief flash per native refresh, e.g. 60 times per 60Hz, or 120 times per 120Hz) would be "simulated 2000Hz".

The scanning backlight would be designed to look similar under a high-speed camera as a CRT scanning:

Notice how about 1/8th of the CRT is brightly illuminated. For a 1/60th second frame (16ms), that corresponds to a phosphor decay of 2 milliseconds. Surpass that, and you're already becoming better than this specific CRT. To be able to do this with a LED scanning backlight, you need to be able to illuminate only 1/8th of the backlight at a time -- and illuminate the backlight 8 times brighter to compensate for the longer dark period.

The shorter the illumination, the sharper the motion is.
A scanning backlight that's dark 50% of the time, reduces motion blur by 50%
A scanning backlight that's dark 90% of the time, reduces motion blur by 90%
(There's no limit to how much further motion sharpening you can do, the limiting factor is the amount of brightness in LED)
etc.

The requirement for extremely bright backlight illumination, before today, is why scanning backlights have not yet been able to surpass CRT yet. But we are already quickly getting there. LED's are now used in projectors! Today, LED's are sufficiently bright and cheap enough nowadays. Although the Arduino version uses LED strips, the high end TV manufacturers prefer individually-controllable LED's, to permit local dimming (which is great too). Local dimming can still be combined with scanning backlight too. For shorter pulses of LED's, you need even brighter LED's, which certainly adds cost. However, if all you want is less motion blur than a CRT display (without worrying too much about black levels), it actually is nowadays relatively cheap thanks to cheap LED tape (one can get 600 LED's in a 5 meter LED tape, for only $11 off eBay -- 50 watts, more than enough to super-brightly-illuminate a 24" computer monitor). Prices for LED's are still falling. We're already at the point where it's practical for an ultra-high-speed scanning backlight (90% dark, 10% bright), required to permit LCD's to have less motion blur than CRT. The cost of LED's to make a backlight 10 times brighter, has fallen to less than $100!

I am pretty interested in seeing TV manufacturers adopt at least inexpensive scanning backlights (perhaps simply by taking advantage of cheap mass-manufactured LED ribbon tape, putting a few thousand of LED's behind a 40" LCD for less than $100 cost), even if local dimming is not used. LED ribbon tape is cheap nowadays, permitting backlights to be 10 times brighter than today -- sufficiently bright for high-speed scanning backlights (90%:10% dark:bright cycles)

The advertising of better-than-CRT-quality motion (WITHOUT motion interpolation, and WITHOUT noticeable input lag), would certainly help sell these LCD displays, and still be much cheaper than OLED panels. A good, useful technology to consider while waiting for better technologies on the horizon. And of course, it is onscreen-menu configurable, so the scanning can be adjusted for a flicker-versus-motion tradeoff. Properly done, the flicker/scanning effect would look and feel exactly the same as a CRT.

Has television manufacturers considered high-speed scanning backlight technology (90%:10% dark:bright, or better), now that the appropriate LCD and LED technologies have now made it possible to exceed CRT in sharpness of motion?

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
Sponsored Links
Advertisement
 
post #2 of 8 Old 09-14-2012, 03:25 AM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 254
Some of the newer PC monitors actually do just this, with sub-4ms response times (may actually be sub-2ms?) and backlight scanning. John Carmack has been posting about them on his Twitter feed recently. (some time in the last month or two I think)

I’m fairly sure that Sony offers this with the “Impulse” backlight scanning option on 2012 sets as well. I don’t think it uses any interpolation, and does exactly what you describe, though I don’t know if it introduces any latency, and being a television rather than a monitor, it will still have at least 30ms lag.

Even my lowly 480Hz HX909 virtually eliminates motion blur with backlight scanning, though it unfortunately does not have the option of doing so without interpolation, and while persistence-induced motion blur is eliminated, there can still be some motion blur from the panel response time—some colour transitions appear to be over 16ms. (gaming monitors are focused on response times over quality colour/gradation)

I don’t think you can do exactly what you propose, because even if you make the displays bright enough, flicker will be unbearable.


I only use my display for three things though: a PC monitor, gaming (technically still a PC monitor, I suppose) and film.
For the monitor and gaming use, I need minimum latency which means I can’t make use of the backlight scanning. (but being a UV2A panel, even without scanning motion blur is kept to a minimum for LCD)
With film, I can, but I can’t tolerate 24p without interpolation anyway (native 24p doesn’t look smooth to me at all) so I don’t mind the requirement. But of course I would prefer to have the option to avoid it.


But yes, I hope this is trend that continues forward, and more manufacturers will start adding a low-latency gaming focused backlight scanning option. But as we still have delays of 30ms and up with televisions, clearly latency doesn’t seem to be a concern for them.
Chronoptimist is offline  
post #3 of 8 Old 09-14-2012, 01:09 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
Quote:
Originally Posted by Chronoptimist View Post

Some of the newer PC monitors actually do just this, with sub-4ms response times (may actually be sub-2ms?) and backlight scanning.
True, they've been doing this since 2006, but many of them only reduce motion blur by 25% or 50% (motion interpolation does the rest). What we want here, is a high-speed scanning backlight that's dark 90% of the time, to reduce motion blur by 90% without using motion interpolation. Are you aware of any 24" desktop computer monitors with better than 75%:25% scanning backlights (e.g. dark at least 75% of the time), scanning backlights that are able to reduce _at_least_ 75% motion blur or more?
Quote:
John Carmack has been posting about them on his Twitter feed recently. (some time in the last month or two I think)
Thanks. I've tweeted a link to my Home-made Arduino Scanning backlight mod, since it's possible to add a scanning backlight with only knowledge of the refresh rate (And manual adjustments for phasing / speed, to sync to a specific LCD's lag. Not too different from a calibration utility for adjusting PC-based 3D shutter glasses for minimum crosstaslk, properly timing the shutters to the LCD refresh. In fact, the VSYNC signal can come from many sources into the Arduino, either from a software program accessin the DirectX VSYNC, or tapping the monitor cable, or tapping a 3D shutter glasses signal, etc.)
Quote:
I’m fairly sure that Sony offers this with the “Impulse” backlight scanning option on 2012 sets as well. I don’t think it uses any interpolation, and does exactly what you describe, though I don’t know if it introduces any latency, and being a television rather than a monitor, it will still have at least 30ms lag.
Probably. I'm curious how many % of the time that backlight is dark. 50%? 75% or 90%?
A scanning backlight dark 50% of the time only reduces motion blur by 50%, while a scanning backlight dark 90% of the time will reduce motion blur by 90% (that requires 10 times brighter of LED's for the short 10% illumination, to keep the same brightness; the technology has finally been able to reach that point recently). Such high-speed scanning backlights (90% dark) have not been done yet, because of the extreme brightness required. But right now, ultra bright LED's allow this.
Quote:
Even my lowly 480Hz HX909 virtually eliminates motion blur with backlight scanning, though it unfortunately does not have the option of doing so without interpolation, and while persistence-induced motion blur is eliminated, there can still be some motion blur from the panel response time—some colour transitions appear to be over 16ms. (gaming monitors are focused on response times over quality colour/gradation)
Agreed (for many LCD panels), but 3D LCD's are finally able to eliminate nearly all of leftover "ghosting" because they have to clear the next refresh for active shutter glasses -- there is some left, but not much on the good 3D LCD panels -- The LCD makers have been forced to design the LCD to be as clean as quickly as possible by the end of a refresh, to avoid image bleed between the two eyes of shutter glasses. It will no longer be a limiting factor. Worse come to worse, it's just a very faint sharp ghost after-image lagging after a super-sharp CRT-style perfect motion on an LCD -- similiar to the sharp double image during 3D shutter glasses (and seen only on high contrast edges), but having no effect to the sharpness of bright edges in motion. (e.g. limiting factor would be how brief the illumination in a scanning backlight will be). LCD panels, thanks to 3D, will continue to improve in this department, but right now we're already at a point where it is now possible to have less motion blur on LCD than on CRT -- provided you have a sufficiently overkill of a scanning backlight needed to do so.

-- Solved problem #1: The above critical innovation piece (LCD's that nearly completely finish refreshing all pixels before next frame), was one of the missing pre-requisites necessary to allow scanning backlights to allow LCD's to have less motion blur than CRT. (given a sufficiently bright and fast scanning backlight that's dark 90% or 95% of the time, something not done on any consumer HDTV or computer monitor yet due to cost of the extreme brightness needed, until now.)

-- Solved problem #2: The other innovation piece is the LED brightness required (for a 10x brighter backlight needed for a scanning backlight dark 90% of the time). It can be done using less than $100 of LED's now. If you wanted to go "For The Win" and gain "2000Hz" simulation (simulate a CRT with ultrashort persistence phosphors), a backlight of sufficient brightness would be simply 20 meters (64 feet) of 6500K LED ribbon tape, a whopping 2,400 LED's, cut up into 32 segments of 24 inches each, and laid into a 24"-27" diagonal rectangle, for example -- that's 200 watts of LED's, though you'd only illuminate 10W or 20W at a time, which also happens to be the power consumption of a typical high-efficiency 24" monitor! For simplicity, you could start with one or two 5 meter (16 foot) LED ribbon tape. LED ribbon tape is as little as $11 for 600 LED's off eBay, but you want to pay extra for the high-quality 6500K with at least CRI 80 or better.

For 120fps gaming at 120Hz refresh with 'perfect CRT' motion, using a scanning backlight, each 1/120th second (8ms) frame would be:
2ms -- wait for LCD pixel to finish refreshing (unseen, while in the dark)
5ms -- wait a little longer for most of ghosting to disappear (unseen, while in the dark)
1ms -- flash the backlight quickly. (1/960th or 1/1000th second -- or even 1/2000th second!)
Any ultra-faint afterimage would be no worse than that crosstalk image seen during 3D active shutter glasses; it's not a limiting factor in motion blur as it won't affect the ability to improve the sharpness of the original edge in motion, during fast motions, during a high-speed scanning backlight.
Quote:
I don’t think you can do exactly what you propose, because even if you make the displays bright enough, flicker will be unbearable.
Properly engineered, It would be no different from a CRT at exactly same refresh. Since videogaming at 120Hz native refresh is already being done on PC gaming with high-end GPU's (GTX 680's and better, etc), the flicker is not bothersome, compared to 60Hz (for most people). Also, the scanning backlight can be flexible -- sync to any refresh between 60Hz and 120Hz. We already know 120Hz computer monitors are now available, and these are the ideal monitors to add scanning backlights to. Also, small capacitors can be added to each scanning backlight segment to simulate the approx 1ms "phosphor decay" of CRT. By this point, a high-speed scanning backlight looks almost exactly the same as a CRT, when pointing a high-speed camera at both. Also, one segment of a scanning backlight will always be lit up at all times, so the average number of photons hitting the eyeballs remains constant (this helps too -- scanning is 'slightly' more comfortable than full-screen strobing). Do you know how unbelievably bright CRT phosphors are, for that 1 millisecond time period? They have to be, because a CRT is black most of the time. Eye discomfort for a specific person will be exactly identical to CRT. If you hated CRT, you'll just turn off the scanning backlight. If you liked CRT for gaming, you'll love the scanning backlight. Note that most people were/are OK with gaming on CRT at high refresh rates such as 85Hz, 96Hz, or 120Hz (many 120Hz monitors can sync to any refresh rate between 60Hz and 120Hz, so choose a good compromise between flicker, versus GPU performance required to get full framerate out of your games, etc) -- and you can just turn off the scanning when you're done with the game, enjoying steady illumination while using Microsoft Word... Scanning backlights can be turned on and off, anyway -- or even tunable (240/480/960/1920 simulation) -- for flicker/quality tradeoff -- via onscreen menus or a utility app.

Quote:
I only use my display for three things though: a PC monitor, gaming (technically still a PC monitor, I suppose) and film.
For the monitor and gaming use, I need minimum latency which means I can’t make use of the backlight scanning. (but being a UV2A panel, even without scanning motion blur is kept to a minimum for LCD)
With film, I can, but I can’t tolerate 24p without interpolation anyway (native 24p doesn’t look smooth to me at all) so I don’t mind the requirement. But of course I would prefer to have the option to avoid it.
But yes, I hope this is trend that continues forward, and more manufacturers will start adding a low-latency gaming focused backlight scanning option. But as we still have delays of 30ms and up with televisions, clearly latency doesn’t seem to be a concern for them.
Agreed, we need improvement to low-latency gaming.

My point to my post is that we're finally at the technological point where scanning backlights can allow LCD to match (or surpass) CRT in motion sharpness, thanks to a confluence of factors I wrote in my original post. Past displayed used scanning backlights that only reduced 25% or 50% motion blur. Today's (expensive) displays have scanning backlights that reduce 75% motion blur. (e.g. Samsung CMR 960 does this to reduce 240Hz motion blur by 75%, to achieve "960Hz" simulated operation)

Now, it's time to design CRT-perfectly-matching scanning backlight (looks the same under high speed camera), or even simulate ultrashort phosphors (0.5ms), and do this without doing any motion interpolation. This requires a scanning backlight that can reduce 90% or 95% of motion blur. Again, CRT phosphors shined amazingly blindingly bright for that tiny time period, to compensate for the short persistence. Scanning backlight LED's have to be equally bright (e.g. 10 times brighter than today's backlight) in order to yield roughly the same average picture brightness as today's displays.

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
post #4 of 8 Old 09-14-2012, 01:58 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
For other readers, some useful/educational references:
List of References -- Including Scientific References

For the home-made Arduino scanning backlight project idea:
Home-made Arduino scanning LED backlight to simulate 480Hz or 960Hz in a 120Hz LCD?

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
post #5 of 8 Old 09-14-2012, 03:18 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
Endorsement from John Carmack on twitter. Sounds like I am on the right track!!! !!! !!!

Mark Rejhon @mdrejhon
@ID_AA_Carmack I'm researching home-made Arduino scanning backlight (90%:10% dark:bright) using 100-200 watts of LED's. 120hz.net/showthread.php…

John Carmack @ID_AA_Carmack
@mdrejhon Good project. You definitely want to find the hardware vsync, don't try to communicate it from the host.

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
post #6 of 8 Old 09-15-2012, 03:15 PM
AVS Special Member
 
borf's Avatar
 
Join Date: Oct 2003
Posts: 1,172
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by Mark Rejhon View Post

choose a good compromise between flicker, versus GPU performance required to get full framerate out of your games, etc

Yea 120hz is too hard on the graphics card for most games and 60hz flickers. I always believed 75-85hz scanning/bfi would work. A multisync scanning lcd would be very desirable to people tweaking performance for eye candy. I'd buy one.

Quote:
Originally Posted by Mark Rejhon View Post

the Arduino only needs to know VSYNC timing, for a home-made scanning backlight that allows an LCD panel to have less motion blur than CRT.

Sounds like you know what your'e doing Mark. I remember hearing that syncing the backlight with refresh was no easy feat for manufacturers.

Quote:
Originally Posted by Mark Rejhon View Post

For example, a single 8ms refresh (1/120th second) for a 120Hz display, can be enhanced with a scanning/strobed backight:
2ms -- wait for LCD pixel to finish refreshing (unseen, while in the dark)
5ms -- wait a little longer for most of ghosting to disappear (unseen, while in the dark)
1ms -- flash the backlight quickly. (1/960th or 1/1000th second -- or even 1/2000th second!)
Here, you just bypassed LCD pixel response as the factor in motion blur.

Pixels are in different states of transition when back lit? Theoretically that's why i always liked black frame insertion. Theoretically anyway. Can the above be done with any precision or does it just "happen" on the fly.
Quote:
Originally Posted by Mark Rejhon View Post

However, we've now reached a technological point where it's now possible to eliminate 90%-95% of motion blur from an image, using suffiiently bright LED's (e.g. 10 times brighter than is required for continuous illumination). We thus, now approach the possibility to have LCD superior to CRT in motion-quality elimination -- and WITHOUT using motion interpolation (which is hated by some).

Simply increasing the blank time does reduce blur but you also need to keep unique frames in there (from interpolation or graphic card or source) to avoid other problems. You need both these things, which is why i would experiment in the range of 75-85hz lcd, where unique frames are attainable. - nice thumbs up from Carmack. He must be a display junkie too.
borf is offline  
post #7 of 8 Old 09-15-2012, 08:25 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
Quote:
Originally Posted by borf View Post

Yea 120hz is too hard on the graphics card for most games and 60hz flickers. I always believed 75-85hz scanning/bfi would work. A multisync scanning lcd would be very desirable to people tweaking performance for eye candy. I'd buy one.
Ideally, I'd design it auto-sync to the timings of the VSYNC (measuring average time between the last 3 VSYNC pulses, and adjusting the scanning speed), so I'd work at any refresh rate. There are new 120Hz computer monitors now, and those will still be the best platforms due to their faster response that still benefits use with a scanning backlight, even when the displays are used at a lower refresh.
Quote:
Sounds like you know what your'e doing Mark. I remember hearing that syncing the backlight with refresh was no easy feat for manufacturers.
Thankfully, it's a lot easier now, due to 3D LCD's and Arduino. I created a schematic already; I am going to post it in a separate thread for discussion.
Quote:
Pixels are in different states of transition when back lit? Theoretically that's why i always liked black frame insertion. Theoretically anyway. Can the above be done with any precision or does it just "happen" on the fly.
That is correct; an LCD is continuously lit, so our human eyes are actually forced to endure the motion blur of the slow pixel state transitions. This YouTube shows the continuous illumination of LCD next to CRT:
.
Now, by turning off the backlight while waiting for pixels to change -- you hide the pixel state changes. On today's modern LCD's -- LCD pixels are already mostly finished refreshing already, slightly before the end of the refresh -- 3D made this a requirement. All we need is that very tiny millisecond to flash the backlight through. Worse comes to worse, there may be a very faint amount of afterimage -- but no worse than the crosstalk between eyes for 3D shutter glasses.

Bottom line: You're bypassing the pixel response as the limiting factor in motion blur, since you can use shorter backlight pulses than the LCD response speed, as long as you do the backlight pulses towards the end of a refresh. I just need to choose fine-enough granularity in the number of segments in my scanning backlight, at least 8 segments, maybe 12 or 16. (8 segments is far easier with existing Arduino outputs, though I may try for 12 segments, since I can still cram that into an Arduino using existing output pins)

A scanning backlight is technically the same principle as black frame insertion; except instead of flashing the whole backlight, you're flashing one segment at a time. This has several advantages:
1. LCD displays typically refreshes pixels from top to bottom (imagine it as if it is a CRT with permanently lit, non-decaying phosphors). You want to light the backlight only behind fully-refreshed LCD pixels, while keeping still-refreshing LCD pixels 'in the dark'.
2. Flatter power requirements. You've got larger power surges for a full-backlight flash, while scanning means one segment is always lit at a time, keeping power requirements flat.
3. Flicker fusion threshold is observed to be slightly lower with scanning than strobing. That's because the average amount of light hitting your eyeballs remains pretty constant, because something is always lit (one backlight segment at a time). More comfortable on the eyes, as a result.
Quote:
Simply increasing the blank time does reduce blur but you also need to keep unique frames in there (from interpolation or graphic card or source) to avoid other problems. You need both these things, which is why i would experiment in the range of 75-85hz lcd, where unique frames are attainable. - nice thumbs up from Carmack. He must be a display junkie too.
Yes. With my scanning backlight, my frames are always unique. So if there's 60Hz, the scanning backlight flashes each segment 60 times each per second. If it's 85Hz or 120Hz, then 85 or 120 flashes respectively. Obviously, the faster the flashing, the less annoying the flicker, so I agree that the 75-85Hz is best.

I'd be getting a "120Hz" 3D-ready LCD to test with anyway, because those are the ones that are the fastest at finishing refreshing the image by the end of the refresh (which has been a necessity to eliminate crosstalk between eyes during 3D active shutter glasses), even if I'd test them at various different refreshes. A higher refresh will also lead to less motion lag on almost all existing LCD displays, so that's another tradeoff to think of, when doing games.

It is known amongst experienced CRT/LCD gamers that 72fps@72Hz on CRT has less motion blur than 120fps@120Hz on LCD. I fully expect 72fps@72Hz with an LCD with Arduino scanning backlight, to have far less motion blur than 120fps@120Hz, greatly lightening the load on the GPU, too.

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
post #8 of 8 Old 09-15-2012, 08:57 PM - Thread Starter
AVS Special Member
 
Mark Rejhon's Avatar
 
Join Date: Feb 1999
Location: North America
Posts: 8,124
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 102
Update: I have now posted a schematic diagram in a separate thread, to attract comment from other Arduino enthusiasts.
http://www.avsforum.com/t/1429546/lcd-with-less-motion-blur-than-crt-schematic-arduino-scanning-backlight-960hz-1920hz-simulation-with-no-interpolation

Thanks,
Mark Rejhon

www.BlurBusters.com

BlurBusters Blog -- Eliminating Motion Blur by 90%+ on LCD for games and computers

Rooting for upcoming low-persistence rolling-scan OLEDs too!

Mark Rejhon is offline  
Reply OLED Technology and Flat Panels General

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off