tfouto

I'm sorry, I missed your previous comment. I captured HTC 10 screen only with Iphone 8+ camera, but it seems to miss any flicker, probably due to some kind of algorithm in iOS to make videos more fluid, I couldn't make it reveal the same things I observed while using HTC 10 to capture these videos. I'll probably buy another HTC 10 in coming months once battery in mine dies (it's at 60% capacity at the moment), so it might be interesting to capture both devices one with another.

Regarding microscope - it's an old soviet semi-toy microscope called "Naturalist" - exactly this one: https://www.ebay.ie/itm/Microscope-NATURALIST-Vintage-PLASTIC-CHILDREN-TOY-USSR-Box-and-Accessories/123554698744?hash=item1cc46f15f8:g:9x0AAOSwMNxXanjk:rk:12:pf:0
Unlike most current toy microscope this actually works and allows me to get decent zoom-in to see individual pixels.

I think blue light will be on every recording since it's present in gray tones and camera capture can be a little bit different from the source due to encoding / compression. I think some minor flickers on my videos are probably compression artifacts, but there is definitely an obvious alternating flicker pattern clearly visible too.

5 months later

nice findings! thanks for your great work. indeed i´ve got eyestrain when using htc10 and iphone8, but also on ipad 10.5 2017. But i will now stay away from displays with FRC!

displays are getting cheaper and cheaper, you have discovered another reason. Don´t buy this sh*t.
I´ve already send Apple several mails regarding their cheap displays...no answer yet.

9 months later

Weve done similar tests now under a powerful industrial microscope with some extra tools to get rid of intereferences like PWM and screen refresh rate. I can confirm some of the findings here - iphone 4s which was good for me showed no subpixel flicker at all. Iphone 7 shows quite a strong flickering. I also noticed flicker in apps, where the app layout and fonts doesnt flicker, but what the app loads (website browser) does. We will record our findings and eventually make an article and post it here. It is possible that PWM is a separate flicker from subpixel flickering (temporal rendering, dithering) and screen refresh rate. If these frequencies meet in a weird way, it produces flicker of way lower frequencies than what is safe. Also those may not be a whole screen occurence, but localized in different parts of the screen.
@andc Could you write me an email specifying what exactly you did in those tests? mjanas555 (at) gmail (dot) com

    4 days later

    martin This matches the info on the patent I posted a while back. So it’s somehow creating a ‘pseudo 15hz’ flicker or something similar. Anybody setting their monitor to even 24hz will tell you how horrid that is.

    It would be good if this research could expand to PC’s - use a known good GPU/PC/Monitor but load good and bad OS/Driver combinations to see if a trend emerges.

    Surely the labs that make this technology would have inspected everything at a microscopic level before mass manufacture? This is to assume it is the panel at fault, which I still largely believe it isn’t.

      diop Yes it does and no I dont expect they would. It seems to be a somewhat unfortunate combination of features that span monitor, OS and driver issues. It might even be a coincidence of the industry, where the left hand doesnt know what the right is doing. Seeing some industrial cases in the past years (c8 pollution by dupont for example), I higly doubnt anyone checks anything, yet on microscopic level. Or that they care even when they know to do anything about it if its not cost effective. This issue is not life threatening (unless you wanna suicide from it, which I dont blame you if you do), so its easy to disregard on industrial level compared to what other horrid stuff came from large industries.

      • diop replied to this.

        martin It is not life threatening, but I would say it is certainly life limiting. I still think the industry have simply mass-imposed dithering (at driver level/hardware level) in new tech as a shortcut to avoid producing expensive 8bpc+ monitors, and to hide any signs that the display is of a lower quality. It wouldn’t surprise me if our expensive phones are using sub-par or older display tech, which relies on dithering to keep the perceived colours 99% perfect across the board. In many ways it could be said that dithering is an RGB standard in effect as it almost guarantees the same perceived color range as 10bit, minus the cost.

        What I don’t understand after all this time is why there isn’t an off switch. Even when directly speaking to graphics developers, they either don’t know or don’t care about it. The freedesktop Linux post on another thread has been open for weeks, I simply want a black and white answer to.

        What dithering is currently used on consumer devices? Can it be disabled? Somebody had to have created this pixel movement therefore it can be disabled in just a few lines of code.

        4 years later

        martin Were you able to post your findings from the powerful industrial microscope? I'm tempted to purchase a slow motion camera to see if I can identify why some displays work well and others do not.

        dev