Originally Posted by ovawe
I see. i thought about programming software that does black frame insertion throughout Windows in any game in any application. So black frame insertion at 144 Hz. It could work on linux as well.
If we do black frame insertion at 144 Hz, then the Flickr would be at 72 Hz I think So mame's black frame insertion works at 60 Hz on the 120 Hz monitor. If that's the case then I see why black frame insertion isn't as ideal as strobing backlight.
Yep, black frame insertion would allow the same motion blur of 144Hz, during 72fps operation.
It won't be able to shorten the frame sample length to less than 6.9 milliseconds (1/144sec) since it's not stroboscopically shortened shorter than that, but it would allow 72fps to have exactly the same amount of motion blur as 144fps -- meaning getting 144fps fluidity without the GPU requirements.
First, read TFT Central's Motion Blur Reduction Backlights
and Blur Buster's Article About Sample-And-Hold
, as well as its scientific references, to understand the relationship of eye-tracking-based motion blur being equal to the length of the time the frame is displayed for... Which means you can reduce motion blur by either more Hz, or by stroboscopically shortening each refresh (or each frame) with black periods between refreshes (or each frame). Black frame insertion is a good technique but it has limitations -- software-based black frame insertion usually only reduces motion blur by 50%, except when you combine black-frame insertion *simultaneously* with LightBoost.
60 Hz regular LCD -- baseline -- 16.7ms continuous display
120 Hz regular LCD -- has 50% less motion blur than regular 60 Hz LCD -- 8.33ms continuous display
144 Hz regular LCD -- has 60% less motion blur than regular 60 Hz LCD -- 6.9ms continuous display
120 Hz LightBoost at 100% -- has 85% less motion blur than regular 60 Hz LCD -- 2.4ms strobe length
120 Hz LightBoost at 10% -- has 92% less motion blur than regular 60 Hz LCD -- 1.4ms strobe length
For your suggestion:
144 Hz regular LCD running at 72fps (no black frames in between) -- has a 6.9ms+6.9ms = 13.8ms of continuous display (two frame repeats)
144 Hz regular LCD running at 72fps (black frames in between) -- has 6.9ms of continuous display (one refresh per frame, shortened by black frame)
So an equal black frame insertion (1:1 black frame insertion) reduces motion blur by 50%.
It won't be superior to LightBoost, but it would reduce GPU requirements.
More benefits would be for 120Hz with black frame insertion, to allow perfect 60fps LightBoost within emulators, like is already being done for MAME -- see http://www.blurbusters.com/mame
. Creating a special resident software like you describe, would allow every single 60fps emulator to benefit fully (without any modifications to the emulators) -- for complete motion blur elimination while using 60fps software like emulators. 60fps software is more common than 72fps software. So I think 144Hz BFI isn't as valuable as 120Hz BFI (while LightBoost is enabled).
This is because:
120 Hz LightBoost LCD running at 60fps with black frame insertion and LB=10% equals 1.4ms frame sample length for each 60Hz frame = you get motion blur elimination benefits for 60fps emulators. A 60 frames-per-second emulator gets 92% less motion blur when you combine LightBoost *AND* black frame insertion. That's because the black frame insertion is doing the job of supressing every other strobe (blocking it with a black screen), so your eyes are only getting 60 strobes per second, getting the benefits of 60 Hz LightBoost necessary for 60fps emulators. So instead of just getting 50% blur reduction (2x less blur), you magically get the full 92% motion blur reduction! (12x less blur). That's the real magic: The combination of black frame insertion AND LightBoost. That's why software based black frame insertion is WAY MORE valuable at 120 Hz when combined simultaneously with LightBoost -- it essentially gives you 60 Hz LightBoost.
. You won't be getting the same blur reduction benefits when using black frame insertion with 144 Hz.
The ratio of 16.7ms:1.4ms means a LightBoost display has a 12x shorter motion blur trail (92% less motion blur). If the motion blur was 12 pixels on the 60 Hz, the motion blur is only 1 pixel on the LightBoost 120Hz (10%) for the same speed motion.Edited by mdrejhon - 4/20/13 at 10:59pm