DLP Smoothness Perception versus OLED

[BFI]

Moved to a new topic at Chief's request:

Why do I find motion noticeably smoother on a DLP projector than OLED TV, both with 16.7ms persistence and input lag? I'm aware OLED pixel response is almost as fast.

I've wondered about this for a long time, because I know my eyes are too slow to see the wheel's individual colors (though recording a DLP at 1000fps reveals them, so maybe they are visible to me subconsciously?). To be clear, I'm talking about mouse movement feeling much smoother (to use the best comparison), and not persistence blur which feels - technically is - the same.

[Chief Blur Buster]

Moved to a new topic at Chief's request:

Why do I find motion noticeably smoother on a DLP projector than OLED TV, both with 16.7ms persistence and input lag? I'm aware OLED pixel response is almost as fast.

I've wondered about this for a long time, because I know my eyes are too slow to see the wheel's individual colors (though recording a DLP at 1000fps reveals them, so maybe they are visible to me subconsciously?). To be clear, I'm talking about mouse movement feeling much smoother (to use the best comparison), and not persistence blur which feels - technically is - the same.

More datapoint to muddy the water: DLP also uses temporal dithering (flashing pixels on and off rapidly at 1920Hz-2880Hz, ala per-pixel PWM), which only becomes easily visible if your camera is filming at more than 2x that rate (e.g. 5000fps camera).

The type of OLED you were looking at (AMOLED, WOLED, QD-OLED) can produce dramatically different look and feel for some human eyes; depending on WHAT they are sensitive to, as no two humans sees perfectly identically to the others. Both the human vision and/or the human brain or the system working together (the natural pickiness factor; much like stutter vs tearing vs brightness user preferences/priorities)

Let's klik off the checkboxes, shall we:

OLED
- Fast pixel response
- No temporal dithering
- Global PWM behaviors may exist (depending on OLED)
- Often has odd subpixel structures
- Often a smaller size

DLP
- Fast pixel response
- Temporal dithering (pixels flashing on and off rapidly on a per-pixel PWM!)
- Temporal color (color wheel)
- Might have soft focus
- No subpixel structure
- Often a big size (wall screen)

You have what is considered an interesting outlier observation in human perception, which may actually be a perceptual trigger caused by something else (e.g. screendoor trigger, lens soft focus trigger, subpixel structure of some OLEDs).

Eliminating some factors (screendoor, lens focus, subpixel structure) requires GPU scaling to intentionally run a resolution 1/2 to 1/3, e.g. running a 4K display and projector at 1280x720, and comparing them at scaled resolutions. This eliminates the screendoor effect, and allows a more apples-vs-apples motion comparison that is undistracted by various spatial factors (screendoor, subpixel structure) or adjacent-native-pixel temporal factors (DLP per-pixel 2880Hz PWM that is non-global, for purpose of temporal dithering), and adjusting projector distance / viewing distance to achieve the same screen size or same FOV.

One theoretical explanation (of MANY!) is the colorwheel provides a defacto "colored GPU Motion Blur Effect" (See www.testufo.com/mousearrow at certain settings may show rainbow effects on some DLPs that make it feel "smoother" than not giving it a colorful motion blur).

There are many perceptual triggers lurking in displays that can amplify/attenuate a different perceptual trigger, sometimes "optical illusion" style (normally placebo but with legitimate "something-else-optical" trigger). Other times it's a new perceptual discovery that requires a new science paper.