Comment by numpad0
3 days ago
It's not the matter of viewing angles. Distances.
Try calculating distances between eyes to the edges of especially ultrawides, for which curved displays are pretty common. Standard recommended distances between head to display is 50cm(20").
Optimal viewing distance for most media (including text found in programs and websites made in the "Vista" and "Retina OSX" era to today, but also all movies and TV shows cut for modern 16:9 displays) is between 30 (SMPTE rec.) and 40 (THX rec.) degrees.
THX's recommendation is based purely on the viewing angle of the fovae (the inner part of the retina that is "high res"), and trying to optimize full coverage of it (ie, pixels on the screen should not fall outside of the fovae).
Microsoft, during Vista, and Apple, during the evolution of OSX, both standardized font sizes as a little larger to make all text sizes comfortable at the 30 to 40 degree range. 30 degrees is * 1.6 diagonal size, 40 degrees is * 1.2 diagonal size.
So, if you have a standard 24" 1080p monitor, that is 28.8 to 38.4 inches, not 20.
SMPTE and THX did not change recommendations for 4k, as the view angle of visual media (ie, the focal length in movies/TV shows) did not change, and text doubles in (pixel) size (but not apparent size) to accommodate it; ergo, do math as if you're on a 100% DPI display. Ergo, 24" 4k would be the same.
Also, for completeness sake, 27" 1440p are rare, but a bit more common with gamers, and their math works out to be between 32.4 and 43.2 inches. This is assuming you adjusted your DPI to 133% and/or you're purely focused on movies/TV and games; if you consume only text and stay at 100% and never view media, you may wish 24.3 to 32.4 inches instead.
I'm not sure if I understand your thought process. I'm talking about distance errors between center of screen and edges of the screen, not singular distance. Frankly you can just paste this to ChatGPT. It's correct enough for this type of topics.
> let distance between head to display 20 inches, screen width 20 inches as well, tell me how to calculate distance errors between center of screen to edges of screen
^ this yields a figure of 2.36 inches among few kBs of padding data
> does that mean the eyes need to be refocused when there's distance error of 2.36 inches, ok to go step by step for this
^ this yields a figure of 30 um among few kBs of padding data
> does that mean the eyes need to be refocused when there's distance error of 2.36 inches, ok to go step by step for this
^ this yields such elaborate responses as "Estimated DoF is about ±1 to 2 inches around the focus point", "Yes, the 2.36-inch distance error is right at or slightly beyond the typical depth of field of the human eye at 20 inches.", "The effect is subtle, but in precision tasks or long durations, your eyes may notice the strain"
Which are all more than correct enough. btw the math is mostly just basic Pythagorean theorem so not hard to follow.
>I'm not sure if I understand your thought process. I'm talking about distance errors between center of screen and edges of the screen, not singular distance.
Yes, I covered this above, but maybe it wasn't clear: this is only a concern on low quality polarizers on normal DPI LCDs. Mostly due to sub-pixel text rendering in combination with certain forms of eye problems (nearsightedness and astigmatism are two I know of, but aren't limited to that), your angle of view through the polarizer changes enough that it causes eye fatigue via color fringing.
Another issue is non-IPS/IPS-like screens: gamma and hue shift happens on traditional TFT and MVA/PVA screens, which leads to eye fatigue.
The fix for TFT and MVA/PVA is going to IPS. The fix for shitty polarizers is either good polarizers or exclusively HiDPI monitors (all 4k monitors of any size seem to have exceptionally good polarizers, in comparison of the past 20 years, even if the panels themselves are sometimes mediocre) or just getting rid of LCD altogether and going to MicroLED and OLED.
I have not observed extreme color fringing or LCD-like gamma/hue/brightness defects on off-angle viewing on MicroLED or OLED.
And for full disclaimer, I will state: some MicroLEDs and OLEDs do not have standard RGB layouts, which may lead to a different (although consistent) sub-pixel color fringing effect: DirectWrite in Windows only understands (V)RGB and (V)BGR, not any other alternative format which a lot of MicroLEDs and OLEDs seem to be experimenting with. Freetype CAN do others, but Gnome, KDE, and other desktops lack ability to describe new formats via GUI. This is less of a concern on HiDPI, but Windows is allergic to making greyscale AA a normal thing again.
So, to reiterate:
* Hue/gamma off angle: Curved can help, but so can using IPS or non-LCD
* Brightness off angle: Curved can help, but IPS is extremely mild, and non-LCD doesn't have it
* Sub-pixel text off angle: Curved can help, but good polarizers on IPS _or_ HiDPI IPS or MicroLED or OLED are better
* Sub-pixel text on-angle: Curved can't help, HiDPI is only way out.
So, given eye fatigue is what we're trying to cure, all the other technologies (which many curved monitors also employ) seem to be a better option.