Jessiah, one thing to consider is that CCFL-backlit monitors also use PWM for dimming. It's much less annoying than PWM on an LED monitor because the light doesn't go out completely on each cycle with CCFL like it does with LEDs. It's possible that this is bothering you. I don't know how plasma screens are dimmed. Also, does f.lux help you at all? For me it greatly relieves blue-light-related eyestrain (but not flicker-related eyestrain, of course).
This is crazy and awesome at the same time:
Sorry to post here, but i think you guys would find it interesting, even if not related with our case...
Yes, both are same, actually if you want to make persistent the settings you set with nvidia-settings you need to save them to your Xorg.conf (nvidia-settings tool can nicely generate the Xorg.conf). Please note that all this works only with proprietary(binary) nvidia drivers.
Regarding Intel, as I said before I tried once a notebook and although according kernel log panel reported itself as 6bit and Intel GPU enabled spatial dithering the screen was very bad to look at. It just came now to my mind that the default PIPECONF_DITHER_TYPE_SP can still be dynamic (what nvidia dynamic2x2 calls), so it could make sense to try PIPECONF_DITHER_TYPE_ST1/2 or just disable it...
About blue-light: Lots of studies about the negative effects on Retina and particulary the Outer Retina Layer of blue light.
People who feel pain in the eyes might be related with the negative effect of blue-light on the retina.
Now I want to expose a different approach blue-light has on human retina: The Photosensitive ganglion cell
“A non-rod non-cone photoreceptor in the eyes of mice, which was shown to mediate circadian rhythms, was discovered in 1991 by Foster et al. These neuronal cells, called intrinsically photosensitive retinal ganglion cells (ipRGC), are a small subset (~1–3%) of the retinal ganglion cells located in the inner retina, that is, in front of the rods and cones located in the outer retina. These light sensitive neurons contain a photopigment, melanopsin, which has an absorption peak of the light at a different wavelength (~480 nm) than rods and cones. Beside circadian / behavioral functions, ipRGCs have a role in initiating the pupillary light reflex.
In humans the retinal ganglion cell photoreceptor contributes to conscious sight as well as to non-image-forming functions like circadian rhythms, behaviour and pupil reactions. Since these cells respond mostly to blue light, it has been suggested that they have a role in mesopic vision. Zaidi and colleagues' work with rodless coneless human subjects hence also opened the door into image-forming (visual) roles for the ganglion cell photoreceptor.
Most work suggests that the peak spectral sensitivity of the receptor is between 460 and 482 nm. Steven Lockley et al. in 2003 showed that 460 nm wavelengths of light suppress melatonin twice as much as longer 555 nm light. However, in more recent work by Farhan Zaidi et al., using rodless coneless humans, it was found that what consciously led to light perception was a very intense 481 nm stimulus; this means that the receptor, in visual terms, enables some rudimentary vision maximally for blue light.”
“Compared to the rods and cones, the ipRGC respond more sluggishly and signal the presence of light over the long term. They represent a small subset (~1-3%) of the retinal ganglion cells. Their functional roles are non-image-forming and fundamentally different from those of pattern vision; they provide a stable representation of ambient light intensity.
- Photosensitive ganglion cells innervate other brain targets, such as the center of pupillary control, the olivary pretectal nucleusmidbrain.pupil
The photopigment of photoreceptive ganglion cells, melanopsin, is excited by light mainly in the blue portion of the visible spectrum (absorption peaks at ~480 nanometers). Thephototransduction mechanism in these cells is not fully understood, but seems likely to resemble that in invertebrate rhabdomeric photoreceptors. Photosensitive ganglion cells respond to light by depolarizing and increasing the rate at which they fire nerve impulses. In addition to responding directly to light, these cells may receive excitatory and inhibitory influences from rods and cones by way of synaptic connections in the retina.
In 2007, Zaidi and colleagues published their work on rodless, coneless humans, showing that these people retain normal responses to nonvisual effects of light. The identity of the non-rod, non-cone photoreceptor in humans was found to be a ganglion cell in the inner retina as shown previously in rodless, coneless models in some other mammals. The work was done using patients with rare diseases that wiped out classic rod and cone photoreceptor function but preserved ganglion cell function. Despite having no rods or cones, the patients continued to exhibit circadian photoentrainment, circadian behavioural patterns, melatonin suppression, and pupil reactions, with peak spectral sensitivities to environmental and experimental light that match the melanopsin photopigment. Their brains could also associate vision with light of this frequency. Clinicians and scientists are now seeking to understand the new receptor's role in human diseases and, as discussed below, blindness.”
PWM and blue-light
I think there are people here in this forum with different symptoms and causes. There are the flicker PWM/dithering affected and the blue-light, and other more complex. Of course it can people which are affected by both.
I think I am not effect by PWM alone. I had a LCD Monitor long time ago with PWM that never caused me much of a trouble.
So I think it’s possible and helpful to identifying the symptoms and maybe causes:
- Eye Strain (tireness not pain), headaches, migraines are more related to flicker/PWM.
Increasing PWM frequency helps.
- Eye pain with no headaches and no migranes are more likely related alone with blue-light. Increasing PWM frequency it’s the same.
Blue-light only sensitive people should have no problem with PWM. Also people lowering the brightness of displays should get less pain. LED’s have a maximum brightness values too high (blue-light LED’s), so lowering means less blue-light. Of course this is related with ambient light. It’s different having a screen outside on a sunny day or inside the house in the dark. Nevertheless lowering the brightness to a very low level should help.
Now blue-light (led’s displays) and PWM at the same time might cause problems. Lowering brightness to 33% on a no- PWM monitor means the brightness is always at 33%. There is no fluctuation on brightness levels. On a PWM monitor on the other hand, the light will be at full brightness 33% of time. So during the 33% of time it will be a high value to the eye, even if the perceived effect on brightness is constant. Nevertheless, 33% it better than 100%, of course.
Changing the PWM to higher differences should help only PWM people. People with blue-light sensitive should feel no difference even if it’s a really high frequency. 150hz its no difference then 80 khz in regards the quantity of blue-light entering the retina. During the same time there will be 33% of maximum brightness despite the number of times this value will be achieved.
So for people with blue-light sensitivity its better a non PWM monitor then a PWM monitor, but the main cause it’s the white-Led blue light high energy.
PWM/flicker with high blue-light could also cause stress on pupillary control due to intrinsically photosensitive retinal ganglion cells sensitivity to blue-light or other unknown causes.
Of course, this is just pure speculation…
The Intel dithering parameters seem to be in the Linux kernel now, not in the Xorg configuration files (in contrast to Nvidia, though I'm not sure that Nvidia drivers actually pay attention to the Xorg configuration file parameters wrt dithering). So adjusting these would seem to require recompiling the kernel, which is annoying but not that hard.
BTW for anyone else who may think we've gone off our rockers and are way off topic: the point of all this is to establish what is causing the kind of eyestrain we experience. If we can do this on Linux we would at least know what we are dealing with. Assuming dithering is a major cause of eyestrain, I would like to see an accessibility option to turn off dithering on Macs. Even better would be a way to select which kind of dithering is used (none, spatial, temporal, spatiotemporal etc.). Of course Apple is not known for giving users lots of low-level control, but here it's really critical.
OK I have two monitors of the same type: NEC231wmi which is ccfl backlit.
One of them is hooked up to an INTEL graphics card the other one is connected to ATI card.
The NEC231wmi is an 8-bit LCD panel (I assume dither free) and it has 180Hz PWM.
I overclocked both monitors to 81Hz matrix level by using a Custom Resolution Utility (CRU) .
On top of that I used the intelPWMcontrol program on the monitor with INTEL graphics card.
The same program cannot be used on ATI graphics card.
So here are the differences:
81Hz matrix level with PWM 180Hz bothers me. (On ATI and INTEL)
81Hz matrix level with PWM 600Hz bothers me much less. (I achieve 600Hz PMW with intelPWMcontrol)
81Hz matrix level with PWM 2000Hz bothers me much less. (I achieve 2000Hz PMW with intelPWMcontrol)
But what I do next is:
I apply f.lux on the ATI and it bother me much less with PMW at 180hz.
I apply f.lux on the INTEL and it bother me much less with PMW at 2000hz.
So more and more I am convinced that the blue light contributes a lot to our suffering. However I fully agree with mvanier that flicker is the fundamental cause of all. Mvanier please let us know how your research goes with linux and for this I would like to thank you.
Kvoth thank you very much for pointing out f.lux.
I also need to get in contact with Jessiah and I ll do that asap.
Thank you all.
PS: I have also tried out the Ipad mini without retina thinking that it would give me less trouble then the Ipad with retina. I was hoping that it would behave similarily to Iphone 4s with which gives me no trouble and oh boy I wass wrong. Ipad mini gives me same symptoms be it Retina or Non Retina. Retina however gives me symptoms faster and stronger.
ArtechokiQ, thanks so much for your detailed experiments! I agree that blue light can make a bad situation (caused by flicker) worse; in fact, I'd be very surprised if this wasn't the case. I am curious: do you notice any difference between PWM at 600 and at 2000 Hz? I'm also very interested that you managed to overclock your monitor to 81 Hz! Do you notice any difference between this and 60 Hz?
Note that the newer "flicker-free" monitors from Eizo and Dell don't use PWM at all except at low brightness (< 20%) and even then it's in the 8000 Hz + range. To me this is evidence that the industry is paying attention to people like us.
I am still experimenting between different PWM rates. But even at 600Hz the image on the computer monitor appears visually more solid than at 180Hz.
Here is the link for the overclocking software and explanation on how it works:
The difference is noticable but even more so if PWM is increased. This suggest to me that PWM may after all be the major culprit of our troubles, more so than matrix level flicker. I dont know where I would rank dithering but my feeling is that the dithering adds little less to discomfort than above two. Nevetheless flicker is flicker and most of it is made worse by blue light. Of this I am certain.
I am seriously thinking of trying out the EIZO foris FG2421, It is 120Hz without strobing and 240Hz with strobing (the second one can be turned off). Plus no PWM above 20% brightness as confirmed by prad.de. Also it has lower blue intensity as well. The only thing that prevents me from buying it is the 8-bit panel with frc to get 10-bit. Frc produces dithering I believe!
PS: Also notice that 180Hz is multiple of 60hz (But so is 600Hz!) so there could be more going on than what we think!
Oh one more thing that I know helps a lot is sleep, which some of us have a hard time achieving lately!
what flux makes besides blue-light is reducing brigthness. Blue-light contributes a lot to brighntess on white LED-brightness, so flux dim's the monitor at the expense of blue light. So the flickering of PWM, transition of black-white-black will be less intense (less contrast flicker). And the lowereness of brightness is achieved via software. Instead reducing the monitor brightness, increases the black period of the PWM.
Have you tried to use the 81Hz matrix level with PWM 180Hz at full brightness? It should not use PWM at full brightness. There you would be exposed to blue-light at the max, but no PWM. If you feel confortable it's just PWM that bothers you, if not it's both.
But as i said in my previous post, i think that blue-light and pwm at the same time are bad to our eyes... But what i see more and more, is that new devices, new ipad's new tablets are increasingly brighter and brighter. It's a trend. And almost always at the expense of blue spectrum of led's.
Happy holidays to everybody.
I have turned off f.lux and tried Intel graphics at 81Hz matrix level with PWM 180Hz and PWM 600Hz at full brightness seperately
PWM600Hz feels different, I am more relaxed and much much less annoyed. But for some reason when I turn on f.lux I feel even better. However give me a week or so because sometimes my symptoms come with a delay as if my nervous system tunes in at a certain resonance and it takes time to resonate back thus producing symptoms. I need more time and I will definietly get back to you guys. I want to exclude the positive psychological effect that I can have sometimes when it comes to change. Sometimes I get excited by seeming success but then oops I was wrong. We will see.
Back to the drawing board.
I've been using the surface pro 2 a lot this week an yesterday morning was enoug for me. My eyes were very irritated, like I'd have an eye inflammation. I simply had to stop using any displays. This morning my eyes are still a bit sore, but I've been now browsing the net with this galaxy s2 an my eyes do not get irritated.
I was hopeful that as the Surface pro 2 does not seem to have PWM when at full brightess, it would not cause eye strain. But my eyes have not been this irritated since I struggled with the Dell 2411 display a year ago. The dell also did not have PWM at full brightness, but it still caused bad eye strain.
So, either the PWM is so high Hz that it does not show in the DSLR tests, or then there is yet another reason for the eye strain. Could be dithering, but the Surface pro 2 has such a high dpi that I would think that the dithering should not cause problems. Blue light it is not, as I have 3 led displays that absolutely do not cause problems.
on microsoft communty forum:
http://answers.microsoft.com/en-us/surface/forum/surfpro2-surfhardware/surface-p ro-2-screen-flicker-blinking/5f85ea81-76fe-4029-80ae-bb7770c5cace?tab=question&s tatus=AllReplies&status=AllReplies%2CAllReplies%2CAllReplies%2CAllReplies%2CAllR eplies
Do not suggest a workaround for a broken product.
I have the brightness set to 100% because there is PWM flicker if it is not 100%. My eyes get extremely irritated if there is even a hint of PWM flicker in a display backlight.
But just so you know, I was able to install the Intel driver straight from the Intel website and now it does not flicker anymore. It as fixed the stuttering scrolling and the general jerky, blinking behavior of the screen.
Maybe Microsoft tried to get too much battery savings and the result is completely stuttering and flashing display.
I'd return the device otherwise, but where I live, one can't just take purchased productc back to the store like in US.