The compression really helps the scenery to look realistic… in fact, I’m not sure that this isn’t a gag.
This will be the case for many years, possibly decades yet.
Without doing a full bottom-up kinematic simulation of every joint in a person, it’s incredibly difficult to make characters move well and also be able to respond interactively to an environment. It’s the hardest problem in video game development.
The best solutions right now are a sort of clumsy mix of blended trees of animations with some localized inverse kinematics for placing hands and feet on surfaces. This is a top-down centralized approach that is used because it’s much cheaper than trying to go bottom-up with simulated bipedal movement and it’ll likely be decades before we can do anything else.
Buy some real estate in the uncanny valley. We’re all gonna be living there a long while.
I always liked the idea of a game that pushed this envelope as cheaply as possible. Say, a game set in a Westworld type environment, a robot theme park where everything is supposed to move like that and they never see a human character. And in that respect the player is perfectly manipulated into everything being uncanny in an expected and justifiable way, so the illustion of reality is completely optimized.
But then you put in one absolutely perfectly-rendered and animated human character. Perhaps using trickery to avoid all the problems you describe. For example, distance, lighting, fog, fixed perspectives/ cutscene. Anything to make that sudden appearance, of a perfectly convincing human character, ice the player’s spine.
This is a great idea. All the best games (and movies/TV for that matter) find strength in the weaknesses of the medium rather than fighting against them. I guess this would only work a few times before the idea was used up, but I like it.
Unfortunately, that doesn’t mean that we would stop seeing games coming out after the idea was tired…
My bigger problem is the contrast between a realistic environment and bumping a table doesn’t knock things over (usually). Everything is rock solid. It’s disconcerting far beyond animatronic humans.
This is another constraint we’ll be living with for a very long time. Full physics simulation of the environment is possible, but too expensive and causes too many issues to be worthwhile. If someone knocks over a table, now everyone has to navigate around it, for example, and whole areas of navigation are trivially blocked, ruining the game. In real life, someone would clean it up, but again, simulating that is expensive and not worth the thousands of hours of engineering that such a moment would require.
These areas of video gaming that seem stalled progress-wise are in a kind of local maximum that is nearly impossible to escape. It certainly creates a funny dichotomy, wherein we have literally photorealistic environments with raindrops bouncing off every surface, yet stiff-legged robots still bump into motionless furniture. It’s a dichotomy we’ll be living with for a good long while though. I wouldn’t bet on games doing better in these areas in my lifetime.
A big part of the issue is that video game development doesn’t scale very well. We’re pretty stuck at around 30 engineers per team and about five years of development time for a game. Going any bigger gets too expensive and difficult to manage. This is more of a cultural problem than a genuine one. NASA managed to coordinate 400,000 engineers on Apollo, so it can be done. Video game companies run by bros and their bro college roommates who were hired because they’re bros are not going to solve that, though.
If only that were true. It’s 15 billion, and counting…
incredible environment, but the character jank seriously lessens my enjoyment. Then again, I’m the kind of person who loves the opposite kind of style - Valheim with it’s low poly models and crazy shaders and effects wins every time in my book.
Take my opinion with a grain of something, I also love dithered effects in 3d environments.
I personally don’t find much appeal in the attempt at being photo-real. It’s come a long way in the last decade but it still isn’t perfect and as mentioned the characters and odd physics or lack thereof further ruin the possible immersion. Give me a more stylized world with strong art direction over this stuff and really nail the physical interactions and I’m much more happy.
‘Psst! -pot, hash, ‘cid’…no, that’d be COD Berkeley. Amsterdam game needs heroin pushers.
Berkeley, 14. I’m walking and get asked if I want ‘pot, hash, cid, meth’ and I didn’t understand so I asked and he repeats louder and louder until we are all yelling it down the street.
Amsterdam I got ‘goofballs’ and I also said ‘what’? Both dangerous. Try some real life to get your blood pumping.
As someone who regularly walks around this area - not enough bikes.
The only thing giving it away as not being Amsterdam is that there aren’t more bikes. Nobody’s riding by, and there’s still spaces left on the railing across the canal, and ENTIRE SLOTS unused on the bike racks. And where are the bikes chained to other bikes?
cough money laundering
I’m thinking way back to games like The Last Express or XIII. Great art direction for both.
The problem with animal motion in video games starts when the designers simplify the skeletal models. It’s easy to think of the femur and calf as two solid bones pivoting around a single point at the knee. But the rounded ends on the bones (condyles) are not perfectly round, so the motion doesn’t pivot around a single center point – the pivot point slides slightly from front to back. The calf is not one solid bone, but two bones that are joined together with fibers, and they twist slightly as you rotate or tilt your ankle. And the spine is definitely not a straight line that connects the neck pivot point to the pelvic pivot point.
Because the models omit those small organic details, the motion capture is reduced to fit the simplified model. And so the resulting slightly weird gaits put the characters in the uncanny valley.
Can this be fixed? Sure! They could use ultra-detailed skeletal models that would permit the finer motions. And with modern graphics processors, it’s not like gamers don’t have the CPU power to render all those details. They’d probably have to redo all their motion capture again, and capture a broader range of people so they don’t all walk the same. And they would especially need to include more disabled people, because if you want the world to look real you better put a real population in it. But it’s doable.
gameplay is also one of those. the graphics quality generally increasing while story choice, branching and interactivity decrease. there’s a sweet spot in the middle somewhere that is getting better. but it seems slow progress
even women are sometimes a bridge too far