Originally Posted by Paps.pt
EDIT: only one problem: the test at 120hz without light boost was made on a TN monitor and not on a LPS like the qnix. Isn't it different?
Correct, but it looks nearly exactly the same.
Also see LCD Motion Artifacts 101
CallSignVega confirmed this, when he ran tests on an overclocked 120Hz monitor. He confirmed the motion blur is similiar to non-LightBoost TN, albiet slightly worse due to slower pixel transitions of IPS. (somewhere between the 60Hz photo and the first/blurrier 120Hz photo, albiet slightly closer to the first 120Hz photo)
The sample-and-hold effect makes motion blur mathematically identical on all sample-and-hold "X Hertz" monitors, because each frame is displayed for exactly the same amount of time. On all 120Hz LCD monitors that don't strobe their backlights, 1/120sec is a mathematical constant.
The speed and accuracy of pixel transitions, adds a minor variable, however. The difference in pixel transitions for IPS versus TN, the motion blur will be slightly worse on QNIX than for a TN with LightBoost disabled. However, it's very subtle (QNIX ~40% less motion blur versus TN ~50% less motion blur), since the sample-and-hold effect creates the majority of motion blur. Pixel transitions is NOT the major creator of motion blur on modern LCD's, because pixel transitions take less than a half frame to complete. (The high speed video
also proves this).
Another good article is Why Do Some OLED's Have Motion Blur?
, which has excellent scientific references
listed at the bottom, to scientific papers explaining the sample-and-hold effect (eye-tracking-based motion blur). Even 0ms instant pixel response displays have motion blur, due to sample-and-hold. As you can see, the vast majority of motion blur you see on modern LCD, is all caused by eye-tracking, not the speed of LCD pixel transitions. This is fixable by adding more Hz, or by adding black periods between Hz. (aka shortening the frame length -- ala CRT flicker effect in eliminating motion blur)
So, it's really:
Photo of 60Hz -- applies to all 60Hz LCD's with continuous shine backlights
Photo of 120Hz -- applies to all 120Hz LCD's with continuous shine backlights
I've actually posted the best-case-possible scenario (fastest non-strobed LCD 120Hz motion blur).
Assuming perfect motion at fps=Hz (frame rate matching refresh rate), to eliminate stutters from contributing to blur....
Tracking-based motion blur is directly proportional to the length of time a frame is displayed for.
60Hz LCD = 16.7ms continuous shine = baseline
120Hz LCD = 8.3ms continuous shine = 50% less motion blur
120Hz LightBoost "100%" = 2.4ms backlight flash = 85% less motion blur
120Hz LightBoost "10%" = 1.4ms backlight flash = 92% less motion blur
(The LightBoost strobe backlight flash length is adjustable, by changing the OSD setting for "LightBoost", see TFT Central's oscilloscope tests
We consider a 60Hz the baseline for motion blur, a frame being continuously shone for 16.7 milliseconds (1/60sec)
The 120Hz frame ratio of 8.3ms:16.7ms is 50/100ths of the original motion blur (50% less)
The strobe flash length ratio 2.4ms:16.7ms is 15/100ths of the original motion blur (85% less)
The strobe flash length ratio 1.4ms:16.7ms is 8/100ths of the original motion blur (92% less)
The mathematical relationship between motion blur and length of visible frame, is pretty accurate when pixel transitions are far less than one refresh long. The pursuit camera photographs confirm this mathematical motion blur equivalence.Edited by mdrejhon - 5/31/13 at 9:43pm