Alyosha2001 I have a Canon 80D that can do at least 1/4000 shutter speed.
Temporal Dithering Sensitivity - My Solution
- Edited
JTL There is also a method to determine PWM frequency using a black image with a vertical white line and a DSLR using a pan shoot. Details of the test can be found here, under the testing and interpretation section http://www.tftcentral.co.uk/articles/pulse_width_modulation.htm
In essence, you take a picture of the line while panning the camera, and then count the resulting lines if PWM is present. Else there should be just a blured line (if no PWM)
Alyosha2001 Been there, done that.
Alyosha2001 Check my other thread on how to build homemade oscilloscope and you dont have to do any more of these stupid tests
Ive done them too and wasted a lot of time. They rarely give any real proof and just confuse you more.
martin Very interesting DIY! I have access to an electronic osciloscope, but this seems very handy, except the part with overvoltage
ditherig 1.8 is out, it still enables default dithering back aftter the screen goes off and back on, using Intel HD Graphics 530
- Edited
Alyosha2001 I dont understand one thing. If the notebook displays are 6bit with FRC (dithering) to make it look 8bit and instead of actual 256k colors seem like they have 16,7 million, how does turning on ditherig doesnt make the display look like crap all of a sudden?
Is there any way to find email on that guy who did the ditherig? Cannot find it anywhere.
- Edited
martin Basically, "ditherig" just turns off the Intel chip CALLING for dithering. Whatever the display has to do in order to fake the colors still happens, irrespective. The Intel chipset has dithering locked on during the output phase at all times, whatever the output device is.
The issue here is that both the chip AND the display are dithering/flickering/PWM'ing/etc. There are so many sources of flicker/pain that it's hard to eliminate them all.
When I once talked in the nouveau* IRC channel with one of the devs, he said they enabled temporal dithering for notebook displays by default because notebook displays don't dither and so they need software dithering. He might be wrong or his knowledge might be outdated though. But if he's right, the color drop should be obvious and result in visible banding.
* (an alternative open source NVIDIA driver)
martin I've attatched 2 pictures on november the 17th, take a look at them, in my case the banding (when the dither is off) is obvious
- Edited
Alyosha2001 Thank you I checked, its very obvious. However, did the eyestrain then disappear when the dithering was disabled?
@Gurm That is exactly what we have to do - eliminate the known issues one by one in order to solve this and work pain free, or work at all. In my case I really need a damn laptop by now, but I am getting an oscilloscope first to measure whether if it bothers me, its PWM. Then ill just go down that road from there.
I am also doing a nice run through all possible doctors and examinations, but more of that in the health thread.
I couldnt measure any PWM on one iphone I !have! to use for work and that gives me the worst strain ever (iphone 6s), so maybe the better oscope will finally pick it up.
I am still conflicted on why symptoms that PWM gives us would be the same for other, unknown cases where PWM isnt present.
martin Yes, the difference is worth it, I prefer the banding all the time.
As far as I know, iphones do not use PWM, but a very interesting similar matter I encounter with the mirrorless Sony NEX5 and a6000. If I use them even for very short period, like taking 2 or 3 pictures, a very sudden dizziness occurs. And as far as I could tell, their LCD's are PWM free. On Canon 550D (T2i) I did not encounter this. I suppose it's only the dithering, but it's like some are affecting more than others.
Unlike the PWM, I don't know how to measure or acknowledge the dithering.
martin Martin, my feeling here is that we are sensitive to some kind of ... "flicker". Flicker is a term that could cover - strobing, pixel marching, pulsing, or any of a variety of other ills. If I take two screens, one which hurts and one which does not, there tend to be two kinds of differences. Either the one that hurts seems, when you look closely, to "flicker" (i.e. static images are subtly moving) or else it is painfully sharp. In that case, I suspect blue light but can't rule out a third, unknown factor.
What I can say is that we've come closer to hitting the nail on the head in some cases here. We now know - in the vaguest terms - what android phones are doing wrong, and it's color spaces. We know what Windows 10 is doing wrong, and it's image compositing. We don't know WHAT in that layer is going wrong, but we know what the problem is. We're getting there.
9 days later
quantumzipper Any updates? Did the S7 work out for you with the screen protector? What about the Dell XPS? Downgrading Intel driver?
23 days later
hard to know the right "dithering" thread, but thought this was helpful from the displayCAL forum. (I havent seen much talk here of using monitor calibration software)
How do I know if a graphics card is capable of dithering?Looking at a smooth grayscale gradient in an image viewer with color management turned off(!) is usually a reasonable test. I.e. first look at it with the videoLUT set to linear (no calibration) to make sure banding is not in the image itself or introduced by the display (some panels are less than 8 bit and use dither internally to generate intermediate steps). Then load your calibration. If it still looks smooth, the graphic card either dithers, or ouputs a higher bit depth (both should suffice to reduce or eliminate calibration-related banding artifacts).
4 days later
So, the dithreing is done either in the LCD panel itself, either by the graphics card? Are there any panels that do not dither by default? Can we make a list of them? Is there any true 8 bit panel? Is there a test to spot the dithering simmilarly to the pwn one? (Seems like although it affects a lot of people, not many are aware of it)
Alyosha2001 So, the dithreing is done either in the LCD panel itself, either by the graphics card?
Correct. On laptops the LCD panel doesn't have much electronics of its own so it's often done by the graphics card.
FRC is dithering, which is often used in desktop monitors, and if your using a panel with FRC and a graphics card that is dithering, it can cause issues with image quality.
Is there a test to spot the dithering simmilarly to the pwn one?
Very hard to do I'm afraid. I have some ideas that involve a microscope and high speed camera ($$$$)
JTL FRC Very interesting article! I do happend to have a Dell with eIPS, and it does not use PWM for backlight, but gives me some kind of headache. The bad part is theese panels have dithering of their own and cannot be turned off in video drivers.
Still I think the best way to spot the dithering nowadays is banding, until some prooves a 8 bit FRC-free panel exists
Alyosha2001 BenQ VA panel?
More expensive Dell 2 or 4k monitor?
JTL I don't know, but I would like to. I also have headache with Sony mirrorless LCD's, with old Canon I don't