[ad_1]
Why don’t we see more games utilizing the facial animation techniques from LA Noire?
[ad_2]
View Reddit by realgeorgelogan – View Source
[ad_1]
Why don’t we see more games utilizing the facial animation techniques from LA Noire?
[ad_2]
View Reddit by realgeorgelogan – View Source
Expensive as shit (Team Bondi didn’t even fully pay the studio that did this), and for me… LA Noire had a veeeerrrrryyyy uncanny valley feel, it is an art/graphic design that does not fit every game.
Same reason LA Noire 2 wasnt made.
On a serious note; doing that (even to that level at that time) will have taken a serious amount of time and energy, which is why the game mechanics are based round it. For most games that dont rely on such mechanics it would be expenditure that they dont need to make, so therefore an extra dint in the profit margin.
Because it’s expensive. People who say it “looked like shit” don’t realise that it looked amazing for its time and nowadays it would look so much better.
But it’s expensive, you’re hiring actors just for animations, alongside them reading the lines (or some other actors do the voice acting). It takes a lot from the budget, so it’s just easier to have animators do the job, since they are already making other animations.
Everyone’s gonna say it, so allow me to parrot:
IT’S REALLY FUCKING EXPENSIVE!
Plus, Ninja Theory always does it better. You seen videos of Enslaved: Journey to the West? Now that’s how you fucking do MoCap.
​
Andy Serkis is an international treasure.
It’s was inefficient and basically sank the studio. There are better ways to get similar results.
This has been asked lots. The actual technique used in LA Noire was to put an actor’s head in a fixed position (they couldn’t move it) and use a ton of cameras to record them as they attempted to say their lines without moving their heads. The software recorded video and attempted to record the 3d morphing of their faces. Animators and motion capture would later be added in for the bodies.
Nowadays it is better and cheaper to only use motion capture and a few key facial textures.
Did you play The Quarry?
Goddammit I loved LA Noire. Not just the facial animations, the whole vibe. There’s no real experience in gaming quite like it – that I’ve found anyway.
That game taught me that I suck at reading facial expressions.
Because it was technically a very expensive flop. And companies with money prefer to keep it.
Cost, development time, and it’s return on investment isn’t very high. Facial expressions isn’t exactly a selling point in a video game.
Because it’s expensive and we have WAY better options now. Motion capture is the norm for AAA studios and it’s becoming more and more accessible and makes animators’ lives a lot easier.
Took up too much memory and was limiting for the actors to the point of being distracting. You have to sit perfectly still in a chair, cannot move your head in any direction, and somehow convey every emotion required including making faces for when you nod or shake your head, without nodding or shaking your head.
Since then the tech has evolved into what we see with Naughty Dog’s prestige stuff, the cameras instead of being pointed at you are attached to you and the end result blends in with the rest of the graphics you’re going for.
My guess is that the technology, while it is extremely detailed and incredibly advanced, costs a lot of money. Not only is the technology expensive, but you also have to hire good actors and that costs a lot more.
If you want an example of why the technology doesn’t seem to exist anymore, a decent example would be the recent Call of Duty game that came out. You’ll notice how the facial expressions are incredibly accurate to the actors portraying their roles.
It was expenaive and failed. Also it struggled with teeth
I wish they’d make another LA Noire 😔
Because it was expensive and a nightmare to replicate for results that really haven’t aged that well. Modern facial capture is much better.
I worked at Depth Analysis (Team Bondi’s child company that created the motion capture software in LA Noire).
I think people are wrongly blaming cost. Cost would have been similar to regular mo-cap if it had ever been reused, rather than being a one-off project.
The problem with this technique is that it lacks any artistic control. You’re recording a mesh, not just some control points. That means you have to have a character that looks exactly like the actor — it won’t work for anything other than photo-realism. Most games want a degree of stylisation. And changing lines for lip-sync in other languages or rewrites is a pain because the meshes don’t want to line up.
Ultimately, a head-mounted camera and a face covered in dots (typical modern mo-cap setup) can get a similar resolution and can be done at the same time as body-capture and can be applied to any model and works better with typical animation flows.
Is that Ken Cosgrove?
AI modelling should vastly help facial animation in the next 5 years
I never quite liked it.
Since they recorded faces and bodies seperate, there was always that weird uncanny valley from the disconnected body language.
Because that degree of face mocap is not only labor intensive, but hideously expensive.
Team Bondi bankrupted themselves to make the game happen.
I’m at least somewhat responsible for the black and white mode in this game.
I was at a work conference in the gaming industry and we were the first people to see the first trailer before it was released publicly a few days later.
I asked one of the people working on the game if it would have a black and white mode. He said they weren’t planing to at all and nobody had thought of that. It shipped with black and white and then a year later I saw him and asked him about it and he remembered me and said I was the reason it was included.
Too expensive.