AbRASiON wrote:OLED or bust people.
I agree!
OLED is a great technology.
But it's not done properly for perceived motion blur in all of them.
Even an instant 0ms response display can have lots of perceived motion blur -- due to sample-and-hold.For example, the PS Vita OLED screen has more perceived motion blur than CRT or plasma, because it's sample and hold.
I'll reiterate:
mdrejhon wrote:I have bad news... OLED is no good for motion blur if it's sample-and-hold version (e.g. PS Vita). Notice how [perceived] motion blur PS Vita isn't any better than a good 2ms TN LCD? (The colors ARE much better; but we're strictly talking about [perceived] motion blur here).
The good news is it's possible to impulse-drive (flicker) an OLED or Crystal LED -- which is what good ones do to eliminate [perceived] motion blur.
Sony's expensive "Crystal LED" prototype intentionally flickers, for the purpose of motion-blur elimination.
Sorry to be the bearer of bad news. The only way to eliminate flickers *AND* keep sample lengths short (e.g. 1ms) is to have 1000fps@1000Hz -- something that's not going to happen unless motion interpolation is used (and motion interpolation is not game friendly). This is why people say CRT 60fps@60Hz (1ms samples) has less [perceived] motion blur than LCD 120fps@120Hz (8.33ms samples). Sample lengths dictate [perceived] motion blur. (Note: sample length is not pixel persistence. Sample length is the length of time that a refresh is visible to human eyes for. LCD refreshes are generally static for the whole refresh).
Yes, I prefer OLED over LCD, assuming equal [peceived] motion blur ability (e.g. equal strobe length). The colors on OLED is so much better.
It's a misnomer/myth that pixel persistence is 100% the cause of [perceived] motion blur on LCD.Pixel response have been becoming shorter and shorter, and on fast gaming TN panels, is an insignificant part of a refresh. Pixel persistence is currently now the minority cause of [perceived] motion blur, and most of [perceived] motion blur you see is eye-tracking-based.
Even if pixel response was instant (0ms) on a sample-and-hold display, you still get lots of [perceived] motion blur due to the sample-and-hold effect (which leads to eye-tracking-based motion blur), because your eyes are continuously moving when tracking moving objects. Your eyes are at a different position at the beginning of the displayed refresh (sample start) and the end of the displayed refresh (sample end). That's retinal blurring -- just like moving a camera while taking a picture without its flash. The camera movement cause blurring in the photograph taken at a slow shutter speed (e.g. 1/60 sec). But if the whole scene flashes instead briefly (Xenon flash) then shaking the camera doesn't blur anymore (e.g. Xenon flash 1/1000sec). Likewise, the strobe driven nature of CRT, plasma, and LightBoost, shortens the visible sample (very quick start-to-end of a displayed refresh -- whether via higher refresh rate, or via large black period between refreshes) that prevents it from being blurred across your human vision.
More reading --
Science & References (skip first part, scroll halfway down)
OLED's that are sample-and-hold have the same issue. Remember, LCD 120fps@120Hz has more motion blur than CRT 60fps@60Hz -- and that is not because of pixel persistence but the sample-and-hold.
Mark my words, it is going to continue to be a problem with OLED monitors, because various solutions for eliminating motion blur are problematic:
-- Sample and hold means perceived motion blur (many hate LCD for that)
-- Impulse-driving means flicker (many hate CRT for that)
-- Motion interpolation means input lag (Not good for games or computer)
-- Higher native refresh rates (go all the way to the end of point of diminishing returns for 99.9% of population).
The only way to solve all the above simultaneously is insane native refresh rates and framerates. Well above 240fps@240Hz. A 240fps@240Hz sample-and-hold display would use ~4ms samples, which is longer than CRT impulses of 1-2ms (phosphor illuminate-and-decay cycle). So a 240fps@240Hz sample-and-hold display does not have as clear motion as CRT, which is impulse-driven at 1-2ms. To bypass all the way to the end of the diminishing returns WITHOUT flicker and WITHOUT interpolation, you need to equal CRT samples. (1ms samples = 1000fps@1000Hz, and 2ms samples = 500fps@500Hz) This is fully impractical (e.g. 1000Hz displays with GPU's capable of 1000fps won't be happening). So we have to go with compromises like impulse-driving or motion interpolation. We already know motion interpolation is not suitable for games due to lag, so that leaves impulse-driving as method to shorten samples without flicker (the CRT flicker effect).
It is worth saying you need extra (extreme) brightness for impulse-driving. CRT phosphor shine insanely brightly when they're illuminated -- so CRT's don't have a dim image. This complicates impulse-driving, since OLED's have long had brightness problems. Though Sony's expensive "Crystal LED" prototype display (not OLED) solve the problem by using over 6 million discrete LED's instead. That is, however, insanely expensive. Eventually OLED will be bright enough for CRT-quality impulse-driving, but until then, we're stuck with sample-and-hold OLED's as well as OLED's with long strobe lengths, which don't eliminate motion blur as well as CRT yet. Give it another decade, though.
The best OLED monitors should have a switch between sample-and-hold mode (solid image) and impulse-drive mode (best game motion). This would be the best of all worlds.