“A separate line of research has reported that fast eye movements known as saccades allow simple modulated LEDs to be observed at very high rates. Here we show that humans perceive visual flicker artifacts at rates over 500 Hz when a display includes high frequency spatial edges. This rate is many times higher than previously reported.”

Why I modulate LED PWM at 2+ kHz

https://www.nature.com/articles/srep07861

Humans perceive flicker artifacts at 500 Hz | Scientific Reports

Edward Morbius April 19, 2018 08:07

I’d long realised to 50 / 60 Hz flicker of flourescents was well within human perceptibility, let alone strobing artifacts such as from CRT gun scans.

LED Brake lights have proved far higher frequencies can also be detected. Prominently.

Lenz Grimmer April 19, 2018 08:20

+Lars Marowsky-Brée can probably confirm

Lars Marowsky-Brée April 19, 2018 08:21

Yes. Flickering lights suck. Especially if they’re not all perfectly synced.

Lars Marowsky-Brée April 19, 2018 08:25

It’s also why Phillips Hue and everything else sucks.

Michael K Johnson April 19, 2018 08:32

+Edward Morbius Cadillac running lights are the worst. Big, bright, and low frequency. I think recent models may have improved but there are a lot of them out there. It’s not so much the brake lights; it’s when they are PWMing them to be the running lights that it’s a problem…

Galen Johnson April 19, 2018 08:42

So… this line makes me question some aspects of the study…“The median contrast sensitivity curve as well as the individual curves for ten subjects is shown in Figure 1.” This just seems like they either only had ten subjects or picked the ten that best supported their hypothesis. However, flicker sucks…no matter the frequency.

Michael K Johnson April 19, 2018 12:15

+Galen Johnson well, you don’t need a sample size of thirty to say that an effect does exist in at least some people. It would take more than ten to be confident that an effect didn’t exist, though!

Rick Troth April 20, 2018 07:17

Wow … It’s not just me? Other people see that stuff too? Cool!

Eugene Crosser April 21, 2018 06:33

I used the default MCU clock of ~1 MHz, and the counter ceiling 1024, making PWM frequency about 1 KHz. I actually can see flicker when I shift my line of sight quickly with the LEDs in the field of view. But not otherwise, and it does not bother me.

Still, this is a limitation: in order to make smooth fade in and out, duty cycle needs to change exponentially. With 2x step, it is only 10 steps of brightness, not enough to make it smooth. In the end I implemented steps by √2 (except the first one that has to be 2x).

Imported from Google+ — content and formatting may not be reliable