Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Games Look Bad: HDR and Tone Mapping (2017)

ventspace.wordpress.com

Post reply on HN
Print view View original

181–190 of 219 posts

Re: Games Look Bad: HDR and Tone Mapping (2017)

#181

Post by mystraline »

dahart wrote: post #175
mystraline wrote: post #156

Earlier quoted context omitted.

> Our eyes absolutely blow out whites in the sun. Do you have a new technique to decode eye-brain perception in terms of how we perceive visual signals? Do you have a paper indicating how you make this claim for everyone?

Do you really need a paper? It’s well known that looking at the sun does damage to rods and cones, because it far exceeds their response range, long before perception gets involved.

'In the sun' != 'at the sun'

And you completely miss what I'm asking too.

Chemical reactions in the rods and cones are only a small portion of vision processing. The rest is in the brain, with a great deal of various processing happening, that eventually comes to cognition and understanding what you see.

And parts of the visual cognition system also synthesize and hallucinate vision systems as well, like the vision hole where the optic nerve meets the eye. But cognitively, the data is there smeared across time and space (as in a SLAM algo putting the data where it should go, not what is measured).

Re: Games Look Bad: HDR and Tone Mapping (2017)

#182

Post by SirMaster »

theshackleford wrote: post #164
SirMaster wrote: post #126

This seems pretty irrelevant now. This article is from 2017 which is before we had proper real HDR support in Windows 10 and much better HDR support now in Windows 11. And before we had OLED gaming monitors which can actually now display good HDR at 1000+ nits. This was definitely during a transitional phase with mostly fake HDR techniques that needed tone-mapping. Now we have real HDR that doesn't need tone-mapping,…

> And before we had OLED gaming monitors which can actually now display good HDR at 1000+ nits. It’s worth pointing out these monitors for the most part can not sustain it or achieve it at anything other than the smallest possible window sizes, such as the 1-3% window sizes at best. > Now we have real HDR that doesn't need tone-mapping, or only a small amount of tone-mapping above the display peak nits point. For the…

> It’s worth pointing out these monitors for the most part can not sustain it or achieve it at anything other than the smallest possible window sizes, such as the 1-3% window sizes at best.

Sure, but the parts of the image that are anywhere near 1000 nits are usually quite small and are things like muzzle flashes or light fixtures or centers of explosions, or magic effects etc.

https://www.rtings.com/monitor/reviews/asus/rog-swift-oled-p...

This is OLED gaming monitor that came out 2 years ago measures 904 nits on a 10% sustained white window.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#183

Post by charcircuit »

ferguess_k wrote: post #141
pradn wrote: post #140

Earlier quoted context omitted.

A large section of the gaming public sees photo-realistic games as serious, and prefers them for high-budget games. It's a rat race for devs though - its just incredibly expensive to create high quality models, textures, maps. I've been playing Cyberpunk 2077, and while the graphics are great, it's clear they could do more in the visual realm. It doesn't use current gen hardware to the maximum, in every way, because…

I think chasing photorealism also hurts the modding community, which hurts the players. No ordinary modding community could push out photorealistic contents in a realistic span of time. I think that's why we are seeing less and less mods nowadays comparing to the late 90s and early 2000s. For FPS, HL2/Doom3 is probably the last generation that enjoys a huge modding community. Anything above it pushes ordinary modders…

HL2/Doom3 have built in mod support, so I don't think it's fair to compare it to games that don't have mod support.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#184

Post by tmtvl »

markus_zhang wrote: post #102
tmtvl wrote: post #65

Earlier quoted context omitted.

Photorealism is a bad idea if your movement engine isn't good enough to handle the character walking around on uneven terrain. For racing games or flight simulators or such it is less of a problem, but seeing a regular person being absolutely flummoxed by a knee-high wall is massively immersion breaking. It's something I really noticed when playing Disaster Report 4, where the people look amazingly realistic but some…

I think it's like porn. Not sure about you guys, but for me soft-core always looks better than HD hardcore. Soft-core encourages imagination and conveniently covers any body part that is a bit far from perfect. And that's why I always think ladies who wear just enough clothes are way more sexy than nude ladies. Hopefully this doesn't offend anyone.

I get it, I prefer seeing two bears be tender and affectionate rather than just 'bend over and spell run'.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#185

Post by teamonkey »

dahart wrote: post #146
markus_zhang wrote: post #96

Earlier quoted context omitted.

I’d argue that immersion has little to do with graphics, even for FPS. Actually I had more immersion in some text adventure games than in some AAA games — and not out of nostalgia because I never played the said text adventure games before. I’d agree that certain degree of graphics helps with immersion, but photorealistic graphics only offers cheap immersion which turns off the immersion centre in the brain — Ok this…

I recall a paper from GDC many years back that studied the perception of immersion and they measured and ranked maybe a dozen different factors. Graphics and visuals were surprisingly low on the list. The number one thing was the player’s sense of identity and clear understanding of their goals. Players tended to correlate realism with high immersion too.

Oh that sounds really interesting, I’d like to track it down.

Was it this one? https://www.gdcvault.com/play/1015464/Attention-Not-Immersio...

Re: Games Look Bad: HDR and Tone Mapping (2017)

#186

Post by Timon3 »

PaulHoule wrote: post #69
ranguna wrote: post #62

Earlier quoted context omitted.

I'm not sure you were being ironic, I find the witcher 3 and elden ring beautiful

I think Elden Ring is a little ugly but still a world I want to experience.

While the graphics aren't as good as some other modern titles, the world and art design make up for it ten times over. There are a bunch of locations that could be paintings, especially:

- The first steps in Limveld

- Liurnia of the Lakes (from Stormveil)

- Leyndell

- The first look at the Scadutree

- Cerulean Coast

- Stone Coffin Fissure

- Enir Ilim

I can't remember another property with a similar diversity of incredibly beautiful and imposing areas.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#187

Post by theshackleford »

SirMaster wrote: post #182
theshackleford wrote: post #164

Earlier quoted context omitted.

> And before we had OLED gaming monitors which can actually now display good HDR at 1000+ nits. It’s worth pointing out these monitors for the most part can not sustain it or achieve it at anything other than the smallest possible window sizes, such as the 1-3% window sizes at best. > Now we have real HDR that doesn't need tone-mapping, or only a small amount of tone-mapping above the display peak nits point. For the…

> It’s worth pointing out these monitors for the most part can not sustain it or achieve it at anything other than the smallest possible window sizes, such as the 1-3% window sizes at best. Sure, but the parts of the image that are anywhere near 1000 nits are usually quite small and are things like muzzle flashes or light fixtures or centers of explosions, or magic effects etc. https://www.rtings.com/monitor/reviews/…

> Sure, but the parts of the image that are anywhere near 1000 nits are usually quite small and are things like muzzle flashes or light fixtures or centers of explosions, or magic effects etc.

Sure, but plenty of things are bright enough in combination at varying window sizes that combined the panels have to drop down significantly. So you might get 1000 nits for a muzzle flash but ~200nits at best for a “bright sunny day.”

The problem is way too many people (I’m not suggesting you) don’t realise this and just think they are “getting 1000nits!”

>https://www.rtings.com/monitor/reviews/asus/rog-swift-oled-p...

Yes, I own this display and it’s one of the better ones for brightness which is why I grabbed it.

However even on the latest firmware, It has a bunch of issues including with colours in HDR unfortunately. It also has incredibly aggressive ABL. Still a great display, but with more limitations compared to the TVs than I’d like still. They’ll get there though hopefully in few more generations.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#188

Post by teamonkey »

windward wrote: post #73
Arainach wrote: post #61

Earlier quoted context omitted.

>Control Did we play the same game? Some of the best lore-building and environmental theming around, paired with some cool mechanics? Sure, the combat got repetitive but this was hardly something to "just sell GPUs"

Nevertheless, it was commonly used for showing off (cloudy, particle-y) raytracing.

Control was one of the first big games to come out after Nvidia’s first line of GPUs with raytracing hardware (RTX 20xx) and one of the first games to use those hardware features. That’s why it was used as a showcase (there was probably a deal between Remedy and nvidia to make this happen, not sure).

It was a good looking game at the time, but remember it originally came out on PS4/Xbox One and that version did NOT have raytracing.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#189

Post by dahart »

teamonkey wrote: post #185
dahart wrote: post #146

Earlier quoted context omitted.

I recall a paper from GDC many years back that studied the perception of immersion and they measured and ranked maybe a dozen different factors. Graphics and visuals were surprisingly low on the list. The number one thing was the player’s sense of identity and clear understanding of their goals. Players tended to correlate realism with high immersion too.

Oh that sounds really interesting, I’d like to track it down. Was it this one? https://www.gdcvault.com/play/1015464/Attention-Not-Immersio...

That’s definitely in the same realm, but not the one I was thinking of. I believe I’m thinking of something maybe 10 years earlier, it had multiple authors, at least one woman, and some of the authors were psychology researchers who were into games. I’d wouldn’t be surprised if this is a theme and avenue of research that has come up many years at GDC.

Re: Games Look Bad: HDR and Tone Mapping (2017)

#190

Post by dahart »

mystraline wrote: post #181
dahart wrote: post #175

Earlier quoted context omitted.

Do you really need a paper? It’s well known that looking at the sun does damage to rods and cones, because it far exceeds their response range, long before perception gets involved.

'In the sun' != 'at the sun' And you completely miss what I'm asking too. Chemical reactions in the rods and cones are only a small portion of vision processing. The rest is in the brain, with a great deal of various processing happening, that eventually comes to cognition and understanding what you see. And parts of the visual cognition system also synthesize and hallucinate vision systems as well, like the vision h…

What, exactly, is relevant about the perception and cognition systems if the signal from rods and cones is clipped or distorted? By ‘blown out’ we are talking about the rods and cones being saturated and unable to respond meaningfully. Your question doesn’t make sense, and I’m neither making claims about nor arguing over what happens in the perceptual system to bad/saturated inputs.

I don’t know what you mean by ‘in the sun’ != ‘at the sun’. I’m the one who said ‘in the sun’ and I was talking about staring at the sun. I’m not sure what your point is, but if you’re trying to say that a game render of looking at the sun is different than the experience of actually looking at the sun, then I wholly agree. A game will (rightly and thankfully) never fully recreate the experience of looking at the sun. If you’re trying to defend &carlosjobim’s claim that human vision doesn’t have an absolute upper luminance limit, then I think you need to back that claim up with some evidence.

Post reply on HN