So I have finally created my own oscope from an external sound card, downloadable software and a bit of soldering.
I did tests on my old laptop (acer 5820tg), that does have PWM confirmed via video-camera test but that does not cause me any issues. PWM measured on the Word document white background. System is Windows 7.
Below are the results:
100% no PWM as expected -
90% PWM -
About 50% PWM -
And lowest brightness (0%) setting PWM -
As you can see, at 100%, you only see (probably) the refresh rate of the screen frequency and a flat line.
Once the brightness is lowered, PWM is present at 2000 Hz. The amplitude is very nice, indicative of a fade-out/fade-in of the lights. The model is supposed to have LED backlit display, but the amplitude seems like its older CCFL.
This PWM does not bother me at all. My guess is the high frequency, but most importantly combined with the amplitude. I have measured external display that bothers me, dont have any photo, however the freq there was measly 200Hz and there was no fade-out/fade-in, the line went immediately down, indicative of purely stroboscopic light.
I have a theory that needs to be tested and also temporal dithering would have to be discarded as an issue before that, but that is that if you are one of the people that already have a problem with PWM, maybe the frequency doesnt matter, and as long as the amplitude is shitty with no fade-in/out, you will get headaches. If its 2000Hz or 100 000Hz then might not matter.
Would explain why even the new macbooks give me problems. Other guess is the temporal dithering. Now if we only could find a way how to measure/test for that ( @JTL, do you have some solid proof of that? I remember you mentioned when your macbook 15 was started in safety mode, dithering was off and it was no-issues all of a sudden, is that still valid?).
If you want, I can post DIY for this little device with detailed instructions so everyone who wants can build it.