4 ms·

It’s a solved problem already. Various IK systems will tackle it properly and some games will have your feet IK follow the terrain and hands IK push up against

by dagmx 3y ago

It’s a solved problem already. Various IK systems will tackle it properly and some games will have your feet IK follow the terrain and hands IK push up against surfaces. However it’s a trade off of performance (the constant ray casting and IK solves aren’t free) but also responsiveness. Many games opt for faster locomotion and responsiveness instead. At the end of the day, it’s a technically solved issue and has been for years, but as with every single thing in game design, it’s a choice and trade off.

modeless 3y ago
IK doesn't "solve" the core problem of realistic character motion. Sure you can tweak foot positions to stop them from sliding but the farther they get from the original animation data the worse the result will look. It's no panacea. There are new approaches to generating realistic character motion being shown at SIGGRAPH every year, getting better and better, but the best ones are largely too expensive for games to adopt. Plus, for player-controlled characters, at a certain point you run into a fundamental tension between control responsiveness and physical realism of animations. I recently saw a great article on the topic which even comes with a WebGL demo: https://theorangeduck.com/page/code-vs-data-driven-displacement#footlocking https://theorangeduck.com/page/code-vs-data-driven-displacem...
- dagmx 3y ago
  IMHO IK provides enough to counter the issues the person I was replying to was mentioning. Now of course you can go above and beyond that for more realistic motion and , but having worked in the space, I would Serguei that is also solved without requiring recent research. Most recent motion research is about novel motion generation but the general problem of dynamic adjustment to authored animation has been solved by several systems already. But just like my original comment, there’s both a development and runtime cost to all of this.
  - dagmx 3y ago
    Ugh autocorrect changed suggest to Serguei but too late to edit my comment now.

by dagmx 3y ago

modeless 3y ago

IK doesn't "solve" the core problem of realistic character motion. Sure you can tweak foot positions to stop them from sliding but the farther they get from the original animation data the worse the result will look. It's no panacea. There are new approaches to generating realistic character motion being shown at SIGGRAPH every year, getting better and better, but the best ones are largely too expensive for games to adopt. Plus, for player-controlled characters, at a certain point you run into a fundamental tension between control responsiveness and physical realism of animations. I recently saw a great article on the topic which even comes with a WebGL demo: https://theorangeduck.com/page/code-vs-data-driven-displacement#footlocking https://theorangeduck.com/page/code-vs-data-driven-displacem...

dagmx 3y ago
IMHO IK provides enough to counter the issues the person I was replying to was mentioning. Now of course you can go above and beyond that for more realistic motion and , but having worked in the space, I would Serguei that is also solved without requiring recent research. Most recent motion research is about novel motion generation but the general problem of dynamic adjustment to authored animation has been solved by several systems already. But just like my original comment, there’s both a development and runtime cost to all of this.
- dagmx 3y ago
  Ugh autocorrect changed suggest to Serguei but too late to edit my comment now.