TCL NXTpaper 60 Ultra - testing deep dive! Transistor leakage? Dithering?

Z3R0Gravitas

I made a big ole Reddit post going into the details ad imagery of my testing and subjective impressions. (Everyone should be able to see that, right?)

But just a couple of select questions from that, for you experts here:

a) Would we call these blips (and the low brightness noise) transistor leakage?

b) Was anyone able to definitively see (and record) dithering on this screen..? That was then definitely gone after using the ADB commands?

My observations were equivocal; 480fps microscope video on Reddit. (Don't think I can embed here?) Stills:

simplex

A) no, its LED driver unique shematic. It drop current each refresh tick. In mostly screens there are slow fluctuation equal to refresh rate (scheme voltage drop each refresh tick)

B) temporal impossible to detect optically. Spatial could be detectable, but you need to clearly see subpixels which is not possible with TCL used coating

Z3R0Gravitas

Thanks for the reply.

simplex LED driver unique shematic

Is this an industry term? It Googles up more general LED driver design considerations, for me.

But I think you're pointing out that it's not an issue with screen-embedded transistors, like in OLEDs. But instead with the separate LED backlight dimming/power system?

I've only half read a bit around this… Does the spike (ahead of the drop) indicate an engineering design attempt to compensate for a drop in brightness, as a frame refresh draws power (and drops the system voltage)?

b) Are you speaking casually, with "impossible"? Because Nick seems to be showing it here in his review (left image, Color Paper mode). If I turn up the YouTube video quality and my laptop brightness.

I didn't actually test this mode very thoroughly. (Because I'm not interested in using it.) Maybe that's a grey-tone he has displayed too?

simplex

Z3R0Gravitas draws power

Yes. Each transistor charge subpixel with necessary level of energy to control light amount. This lead to voltage drop. And only LED driver circuit scheme can smooth it, or peave "as is"

simplex

Z3R0Gravitas speaking casually

Temporal FRC hard to detect coz you can mistake it with pixel-invertion. For example your panel 6-bit and need to draw 167 value of green: it shine 160 value 2 frames, then 170 value 1 frame, then 180 value 1 frame. It is impossible to see difference in subpixel light between 160, 170 and 180 values in camera, they will shine same amount of light. Idk what ADC need camera, to show you so little difference of light ( more darker green in 1 frame, more shiny in next 2 frames etc )

Spatial FRC always has repeatin patterns

simplex

This might be userful for understanding output form ripple which is visible in oscillograpf

https://www.romstore.ru/system/storage/download/FR9889%20FITI.pdf.gAwJFblT7ubf4TkgHJ9loFpDF06rCCdw