Surface Pro & Pro 2 Eye Strain

[javispedro]

So are you saying that it is not ignorant to suggest that people who do suffer from the PWM effect, should use eye drops, when that suggestion is based on lack of knowledge, that there really are people who suffer from such a problem?

The comment may be completely fine, if, for example, it comes from having had a gazillion patients who claimed to have eyestrain and eye drops fixed the problem. If you have proof that eye drops don't work on you, then fine, feel free to correct him and call someone who just tried to help you "ignorant".

Now people have been using displays since the eighties at work and at home, but only during the last couple of years LED or AMOLED displays with PWM have spread wide, in correlation with eye drop commercials, whichs were almost nonexistent 10 years ago, at least where I live. If they advertise, people must also use those more.

i.e. in my experience....

Eye strain and computers is certainly not a new topic from the 2000s. There is plenty of peer reviewed publications, and, yes, at least a A/B test:
- http://dx.doi.org/10.1016/j.survophthal.2005.02.008
- http://dx.doi.org/10.1007/s00417-003-0845-z
- http://dx.doi.org/10.1016/0003-6870(83)90223-5 (from 1983 no less)

They all conclude that "dry eyes" is the most common cause for eye strain after prolonge computer use. So yes, I can say with plenty of confidence than PWM is not the cause of eye strain for most people, despite "my experience".

Regarding eye strain because of flickering, as said, there are some cases indeed, which I know of because of complains that happened when they forbid classic lightbulbs. In fact it has been proven that on specially sensitive persons lighting flickering at less than ~70Hz can even cause seizures. For example:
- www.ormenww.com/assets/flickering.pdf
- http://dx.doi.org/10.1109/ECCE.2010.5618050

 

[maxx7]

The comment may be completely fine, if, for example, it comes from having had a gazillion patients who claimed to have eyestrain and eye drops fixed the problem. If you have proof that eye drops don't work on you, then fine, feel free to correct him and call someone who just tried to help you "ignorant".

Yes - as said, if you offer to help someone without having the slightest understanding of the issue at hand, you are generally regarded as ignorant.


i.e. in my experience....

Yes, this is a forum and anecdotal evidence is the very characteristic to a forum discussion. Without if everything would require a peer reviewed study.... very little progress.

Eye strain and computers is certainly not a new topic from the 2000s. There is plenty of peer reviewed publications, and, yes, at least a A/B test:
- http://dx.doi.org/10.1016/j.survophthal.2005.02.008
- http://dx.doi.org/10.1007/s00417-003-0845-z
- http://dx.doi.org/10.1016/0003-6870(83)90223-5 (from 1983 no less)

Yes, all of these studies have been floating in the different PWM discussions, but none of the studies recognize the PWM problem. If there were studies that did that, then for all the countless people who have been to several eye doctors, it would have been so much easier to just grab a couple of studies to show the doctor that a real problem exists. I personally have complained this problem to a couple of very competent eye doctors, but none of them recognize that there is such a problem or have ever seen anyone else with such a problem. The PhD actually said that computer monitors are nowadays so good, that it's almost like watching the nature, that there should be no eye strain what so ever.

They all conclude that "dry eyes" is the most common cause for eye strain after prolonge computer use. So yes, I can say with plenty of confidence than PWM is not the cause of eye strain for most people, despite "my experience".

Yes, you can say what you want, but if you would understand this problem, you would understand that dry eyes is a symptom of the PWM problem, not a cause of it. And as said, a lot more people have dry eyes nowadays... thus the suspect is the PWM, which just has not been widely recognized yet.

Regarding eye strain because of flickering, as said, there are some cases indeed, which I know of because of complains that happened when they forbid classic lightbulbs. In fact it has been proven that on specially sensitive persons lighting flickering at less than ~70Hz can even cause seizures. For example:
- www.ormenww.com/assets/flickering.pdf
- http://dx.doi.org/10.1109/ECCE.2010.5618050

Again you are referring to a different problem. Your links discuss the LED flicker without PWM, where the flicker is <165 Hz. The PWM problem just starts to occur >165 Hz, up to at least 600 Hz.

So not helpful at all, this same topic has been revolved with the likes of you in several of the forum discussion. People who understand the problem get very frustrated when someone comes to the discussion and starts to suggest solutions like eye drops, glare protectors or that this problem is related to low frequency flicker. I personally had no problems with CRT displays that flickered from 50 to 120 Hz. Also a display that flickered 165 Hz was almost problem free, only when I did really overuse it, it caused slight problems. The Samsung LED amoled displays with 240 Hz and the Surface Pro are the worst offenders.

 

[Wibin]

The issue is indeed with pulse width modulation. I get terrible headaches since flashing lights affect the brain, not so much the eyes.

I have multiple laptops which have intel HD graphics card and I've removed the PWM issue by adjusting the frequency. I have compiled the driver hookup to adjust pulse width modulation for bot 32bit and 64bit os, .NET 4.0 required. Setting it to 4000hz is within manufacturer specs and should relieve you from headaches:

http://donovan.fi/files/intel_HD_pwm_adjuster.zip

Note that I couldn't get it to work on some of the newer drivers, but most drivers from 2013 should be ok.

Please give feedback whether this works or not since I've been planning on getting a surface too.

 

[GoodBytes]

How does your program work exactly? Is the source code available?

 

[Wibin]

How does your program work exactly? Is the source code available?

Yes, the source code is available, although not by me!
You can find it here: http://pastebin.com/6iirMbfc

I have modified the parameters slightly, but other than that the functionality is the same. It calls memory mappings available through the driver that adjust the frequency. So it's driver dependent.

 

[GoodBytes]

Awesome! Thanks.
I was mostly curious on how it was done.

To know this, to me I think this is from someone who worked at Intel on the GPU drivers.

 

[Wibin]

Awesome! Thanks.
I was mostly curious on how it was done.

To know this, to me I think this is from someone who worked at Intel on the GPU drivers.

No problem!

I'm not sure if this is the one I checked, but here is one (of many?) whitepapers on the topic: http://www.intel.com/content/dam/ww...s/lvds-dynamic-backlight-brightness-paper.pdf
It has some examples of the driver registers which need modification. Again, I'm not sure which versions are supported, which are not, but it may be that later driver versions will end in error.

Most of the time, it seems like it's working - ie. no error message, but when you set the frequency to something like 100 hz (a really bad flicker) the display still looks the same. In other words, the driver and the tool I built function, but the actual backlight implementation is not listening to driver commands. This also happens if the display is current limited instead of pwm, or a mixture of this (like surface 3). Disclaimer: I have not tested it on surface pro 3, so it may or may not work for the less than 50% brightness use case.