Petition to Intel

diop

Great.. so nice to know Intel cares.

The only way around this entire situation IMO is if we can find insider dev's/contacts to develop a 'bootleg' driver - I'm getting more convinced all these changes causing problems is mostly at the driver level.

For example, you all say you have problems with W10 and that is valid, but you can only use a driver from say 2015 onwards when the problems were already present even on past operating systems.

Is there a way to slipstream or hack a <2011 driver to work on a modern chipset/OS? We just need conclusive proof somehow that it is the driver at fault.

Methinks there is something else at play regarding all these changes in the last decade... /tinfoil 🙂

JTL

diop @hpst Is there a way to slipstream or hack a <2011 driver to work on a modern chipset/OS? We just need conclusive proof somehow that it is the driver at fault.

No, but with enough effort I could compare the changes of a previously "known good" driver and 1 version newer that causes strain (if possible) and work from there.

The test I propose uses a lossless capture card to compare the display outputs of a "known good" source vs a "bad" source using software that I will have to write to compare the pixels over an interval. Should be fairly easy to spot dithering that way and convince Intel and Nvidia that we aren't crazy.

Plsnostrain

Thats sad they locked the tread. Unfortunately I bought an acer ryzen laptop and it gives just as much strain. So it just not only Intel.

Soreeyes

Plsnostrain

Yeah, AMD has been known to cause strain for years just like Intel and now (as of around 2015) Nvidia too. It has been a bigger issue with Intel though since 90% of laptops are Intel based.

martin

JTL Maybe its time to start a kickstarter for the card then, Ill chip in if you point me.

KM

Does anyone have problems with Intel drivers under latest Linux? Which currently is 4.18 or 4.19 (in development) . If you type "uname -a" in a console you can see your Linux version. I searched the Intel thread and found no usable Linux information.

martin

martin I have started a new thread on intel asking for the results to be published. https://communities.intel.com/message/571538#571538

JTL

martin Soon, just have a few more things to sort out first.

deepflame

martin Would be also interesting to know how the TUV tested it. Maybe they did not analyse the screen with a magnifying lense to check for dithering e.g. (?).

JTL

martin Very polite well written post.

Gurm

I'm keenly interested to see what TUV did or didn't find. Any one of us can, in a double-blind scenario, pick out which driver is which simply by spending a few seconds/minutes looking at the same display and a/b. The fact that there is "no measured difference" just tells me that they are measuring the wrong things.

JTL

Gurm in a double-blind scenario, pick out which driver is which simply by spending a few seconds/minutes looking at the same display and a/b. The fact that there is "no measured difference" just tells me that they are measuring the wrong things.

Agreed

Gurm

JTL And, since everyone is now doing it, I'm guessing whatever it is was SUCH a commonplace/boring/whatever change that they just can't conceive that it is a problem.

diop

Gurm Since everybody is now doing it, let's say e.g. dithering - why are they all of a sudden? Did old portable devices and PC's have banding? I don't remember so myself.

andc

diop

There are multiple possibilities when it comes to dithering. Some of them:
- frequency of dithering could change, so it’s no longer acceptable
- dithering could go from pixel-level to subpixel level or the other way
- dithering can be made on specific color shades, based on my tests they are usually darker, maybe old drivers didn’t use dithering to such extent
- our eyes could get worse over time so we can’t cope with dithering anymore

@JTL wrote somewhere that OSes/drivers seem to be poor at understanding how many colors a screen can handle. Maybe it used to be tuned better in the old systems, screens might use more varying technologies which make it harder to recognise and it enforces dithering everywhere.

Gurm

I think it coincided with the push to make screens that could display "16.7 million shades of chartreuse". The option was to start producing 8-bit screens OR to start cranking out 6-bit screens with dithering. And I think that everyone started compensating for 6-bit screens, because manufacturers would report the screen as 8-bit when it wasn't. Even early retina displays misrepresented themselves. I remember a period where screens would report as 10-bit, which is ridiculous.

andc

That's my guess too - actually colorful screens are good from every business point of view - OS, graphic cards and screens/devices themselves will sell better if they have wider gamut, so it must have been mutual aggreement, that more bits = better, even if these bits are not so tasty as the real ones 😉

Gurm

There was a short period where laptops were shipping with super-wide-gamut RGB displays, but that was few and far between. Now, they are ultra-low-power, ultra-bright, 6-bit screens with dithering. It's a trend that will likely subside ... eventually.

Soreeyes

Gurm

Average laptop screen quality seemed to nose dive massively when LED backlighting became the norm.

Actually all screens in general did. Fortunately, It's slowly crawling its way back up.

LED backlit screens can look amazing but IIRC the LCD screens with the best gamut are still all CCFL backlit (this may have changed with recent developments it has been a while since I've looked it up). It's why some companies still did CCFL backlit (very expensive) monitors for "professional use" until very recently.

JTL

Soreeyes There are some GB-r LED backlit wide-gamut displays but they ain't cheap.

https://www.bhphotovideo.com/c/product/1005477-REG/nec_pa302w_bk_sv_multisync_30_gb_r_led.html

Tempted to pick this up as a present for myself after I finish my Macbook dithering research and do a review for you people. Worse case I resell it on the open market.

« Previous Page