Инсталирайте Steam
вход
|
език
Опростен китайски (简体中文)
Традиционен китайски (繁體中文)
Японски (日本語)
Корейски (한국어)
Тайландски (ไทย)
Чешки (Čeština)
Датски (Dansk)
Немски (Deutsch)
Английски (English)
Испански — Испания (Español — España)
Испански — Латинска Америка (Español — Latinoamérica)
Гръцки (Ελληνικά)
Френски (Français)
Италиански (Italiano)
Индонезийски (Bahasa Indonesia)
Унгарски (Magyar)
Холандски (Nederlands)
Норвежки (Norsk)
Полски (Polski)
Португалски (Português)
Бразилски португалски (Português — Brasil)
Румънски (Română)
Руски (Русский)
Финландски (Suomi)
Шведски (Svenska)
Турски (Türkçe)
Виетнамски (Tiếng Việt)
Украински (Українська)
Докладване на проблем с превода
Правила и насоки за дискусии
Докладване на публикацията
Once again, the OP wasn't correct about his claims of ugly women. LOL wut.
And even if a game had a woman that was "Ugly." it doesn't mean anything about pandering or wokeness, or anything like that.
Otherwise Samus is woke, and that's a damn silly claim to make. :P
Or the original Lara Croft.
I'm just saying, why do people make up blatant false claims and then get people to somehow agree with them without any evidence to make the claims realistic?
I just don't understand it.
Censorship does happen from time to time, but that's neither here nor there, and has nothing to do with "ugly" women.
Good grief.
Alright. That is fair enough.
UE5 kind of makes uncanny Valley turn into a canny plateau
last time i checked, outer worlds looks just like new vegas but slightly better. New Vegas has always been an ugly game anf thats ok
Just very brown. :P
I'm not sure if it's the uncanny valley effect you take the most issue with, or the attempt at creating "realistic" characters. Either way, if you want for now, you can resort to playing only "cartoony" styled games, (Excepting of course those Japanese games you dislike.) avoiding the more realistically styled ones, at least until graphics increase to the point of being photo realistic.
He's not complaining about the uncanny valley at all. Not sure how you got that from him complaining about ugly people.
We're still not at the Uncanny valley yet, but we're getting closer.
If you want to "cherry pick" then you can do that in many games, also when the characters are all supposed to look gritty, but whatever suits peoples agenda I guess.
Like this ie.
https://steamcommunity.com/sharedfiles/filedetails/?id=1889793851
I am sure many considers them "ugly" instead of understanding the setting, engine limitatins and that these were cherry picked
I'd have to say that your standards for hyper-real characters are a bit too high to be even partially satisfied with anything but a very few, very limited, sorts of games.
It is entirely possible to have hyper-real 3D characters in a game.
But, balancing that against everything else necessary to reach that level of art direction ends up adding far too much in terms of assets/processing dedicated to it and not enough for... "gameplay."
***
Is it worth it?
Not usually, no.
^-- The most meaningful answer.
***
One's "game" may only be moving a hyper-real character around one room. And, the level of work needed to do that in a completely believable way... It's not worth it and may not even be achievable.
Some things can only be achieved with certain very intense bits of engine mechanics and, even then, we're still a bit behind in being able to reach "perfection." (Hair is a great example. Even though we do know and have good physical modeling of all its meaningful characteristics, it is still the "three body problem" of 3D animation/art.)
There is a very real need to balance processing requirements, memory load, asset footprint and work dedicated towards producing those "hyper-real" assets.
Is "hyper-real" necessary for gameplay? Not usually, no. So, does a studio develop a huge part of the work to producing "hyper-real" given they still need to produce "gameplay?" No.
As a 3D Hobyyist, I can produce accurate representations of "real." But, there's a low-level problem, there - To be truly hyper-accurate, one would need a heck of a lot of custom topology in different 3D models. Every single one of those would necessarily be "its own model" and take up a separate footprint instead of being something of an instanced load, just with different values for rigging, scales, certain vertice positions, etc.
This is why one often sees the same body mesh being used in all characters, but maybe with different head geometry in older games. In newer games, one may see a shared base-mesh, but just with higher resolution and more variability in shapes and some rig scaling, which is very lightweight too.
I have an interest in photogrammetry, 'cause of course I do. And, along with that, human anthropometrics. 'Cause, of course I do. (I love human development stuff across an entire spectrum of specific subjects. It's fascinating to me.)
These are being applied today to achieve "hyper real" characters in 3D simulations and even, in certain instances, in gaming.
One can absolutely scan a human figure and derive a 3D model from that. The resolution is only limited by our tools and processing power. The most common way to do this has long been using... the Microsoft Kinect... No lie. :) But, now replacements are being made for that for this use, so it's not so terribly difficult to find hardware. When Microsoft stopped the Kinect program, they immediately orphaned a lot of photogrammetry/anthropomettic researchers.
One can, with increasing capability, even scan a human figure and render that point cloud into a workable, riggable, animateable, 3D mesh. It can be done and has been done. There are companies out there selling commercial-grade meshes that are truly animation ready that are as exact a photo-real scanned representation as is currently, reasonably, possible. An appropriate production house can even creat unified meshes from this with blendshapes/morphs for custom characters. (Even a hobbyist can do it, though the results aren't always that great. :))
There are "scientifically" accurate models derived from human anthropometrics that qualify as "hyper-real" inasmuch as their intended use. (Usually, ergonomics, safety, and military equipment design, some in certain clothing design/etc.)
Intended use and the resolution necessary to facilitate that must be balanced.
But, reducing the memory load needed for "hyper real" can be difficult. It's most common to have a shared base-mesh among all characters so that at least it's a shared load in terms of a general original point-cloud/geometry/topology and then differences are just additional mathematical variables applied to that mesh.
But, the further down the hyper-real rabbit hole one goes, up to and including photogrammetry and custom topology or even additional high-density unique meshes, the more processing power and memory load such a thing takes.
Physicality Based Rendering (PBR) has come a long way in reproducing "hyper-real" materials. Cloth can actually now look like "cloth." Skin can truly look like "skin" with Subsurface Scattering, even. And, very realistic reflectance and specularity are much more easy to achieve as a result. Human Hair, sweat, detailed facial movements, vellus hair (like the fine hair on your arm) and very subtle skin/muscle movements... are all still the stuff of Hollywood Studios and not even they can do all of them "hyper real" all the time. (Grand Moff Tarkin can be spotted as a 3D fake immediately by someone who knows what to look for.)
IMO - Unreal Engine 5 currently represents the present "best case" engine/renderer for gaming with "hyper real" characters. But, even that is still a bit distant from hyper-accurate, hyper-real, characters.