I wonder if there are differences in lagom test, comparing XL2546K 240Hz non strobed 4.16ms MPRT Vs 240Hz (and 100Hz with VT tweak to take 240Hz scanout) DyAc+ Premium 0.5ms MPRT. To check if short MPRT and longer time betwhen refresh, help hide pixel walking, like it do with motion blur.
For curiosity, i tested with the small screen of the old Galaxy S3 smartphone, AMOLED 60FPS/Hz VSync On non BFI. Pixel walking in all boxes, even with slow scroll speed. This confirm than even with a boost in GtG is not fixed.
Even knowing than pixel walk is not caused by slow GtG speed, and is related with voltage control, for this test over 240Hz monitor is very relevant GtG speed. I take box 1, one pixel distance betwhen colors, alternate pixel black (0-0-0) and grey (128-128-128).
http://www.lagom.nl/lcd-test/inversion. ... on-all.png
Rtings XL2546K AMA Off
I selected Off because is overshoot-free and have the best dark average. (0-0-0) to (128-128-128) is equal to (0%-50%) in this table, and value is aprox. 10ms, and (50%-0%) 1.2ms
Traslating ms to FPS. The max. framerate reaching target color is:
(0%-50%) 1000 / 10 = 100 FPS
(50%-0%) 1000 / 1.2 = 833
Even fixing pixel walk, GtG can't maintain color precision beyond 100 FPS. The more scroll speed over 100 pixels/s , the less color accuracy and contrast.
This is very important in one FPS focused monitor, where fast mouse turns are constanty done, overpassing 100 px/s constantly. Imagine a game with a texture made of 128-128-128 grey text over black background, and move mouse at 960 px/s
Testufo scroll text 240 px/s:
https://www.testufo.com/framerates-text#pps=240&count=1
(0%-100%) 1000 / 5.9 = 169 FPS
(100%-0%) 1000 / 1.8 = 555
120 px/s should be Ok, but 240 px/s overpass 169, in theory even with DyAc+ Premium, white color following black text should be light grey. And more longer and dark trail, the more the speed is increased. You can confirm ?