As long as mocap won’t be used for every animation of every humanoid character, forever
its_probably
I already exist.
Bobsofa
If you get out of the chair once in a while and go outside, you might find one /s
Sarcasm aside, AW2 looks pretty good. It’s getting close. But Mocap needs to happen more. LA Noire was an interesting step, too.
MichaelMJTH
The uncanny valley is an incredibly large gap to cross. There is still a significant amount of research going into more believable human characters, simply because there is a significant amount of minute details that go into crossing the uncanny valley beyond just standard character animation and mocap. It’s uses go beyond just games and entertainment, but also robotics.
There are so many micro movements the human body does on a regular basis, loads of tiny muscle movements in just our faces to create expressions. And even after doing all that animation work, we’d still need to work on the interplay between humans and other objects, or how light interacts with skin via sub-surface scattering, etc. Suffice to say making believable humans models and movement is incredibly hard, because humans have evolved over thousands of years to recognise their own kind better. We’re innately able to pick up on flaws quickly.
Before I started university, as part of one of my A-Levels I actually wrote a short research paper on this topic and interviewed some experts in the field. When I think back to what I learned back then, it’s astounding how far we’ve come in just the last 15 years.
5 Comments
Forever.
As long as mocap won’t be used for every animation of every humanoid character, forever
I already exist.
If you get out of the chair once in a while and go outside, you might find one /s
Sarcasm aside, AW2 looks pretty good. It’s getting close. But Mocap needs to happen more. LA Noire was an interesting step, too.
The uncanny valley is an incredibly large gap to cross. There is still a significant amount of research going into more believable human characters, simply because there is a significant amount of minute details that go into crossing the uncanny valley beyond just standard character animation and mocap. It’s uses go beyond just games and entertainment, but also robotics.
There are so many micro movements the human body does on a regular basis, loads of tiny muscle movements in just our faces to create expressions. And even after doing all that animation work, we’d still need to work on the interplay between humans and other objects, or how light interacts with skin via sub-surface scattering, etc. Suffice to say making believable humans models and movement is incredibly hard, because humans have evolved over thousands of years to recognise their own kind better. We’re innately able to pick up on flaws quickly.
Before I started university, as part of one of my A-Levels I actually wrote a short research paper on this topic and interviewed some experts in the field. When I think back to what I learned back then, it’s astounding how far we’ve come in just the last 15 years.