Ecstatic movement: NaturalMotion on Euphoria

Ecstatic movement: NaturalMotion on Euphoria
Will Freeman

By Will Freeman

September 18th 2013 at 10:00AM

Develop asks the team behind this refreshed animation solution what it means for developers

[This feature was published in the September 2013 edition of Develop magazine, which is available through your browser and on iPad.]

After years in development, NaturalMotion’s ‘dynamically synthesised motion’ character animation engine Euphoria has be made available as a final productised middleware for games developers.

Delivered as an optional addition to the popular Morpheme tech, it lets studios use CPU to generate character movement on the fly, by simulating the motor nervous system, body and muscles of individuals in a game, meaning they can better respond to the world around them with natural movements and reactions.

Keen to learn more about the middleware, which is currently focused on next-gen, Develop caught up with NaturalMotion’s senior developer support engineer Peter Cavanagh and VP of sales for EMEA Paul Topping as Gamescom sprung to life.

Why would a developer need to embrace the ‘intelligent animation behaviours’ approach Euphoria offers?
Paul Topping: There’s a whole bunch of reasons to use Euphoria. Obviously, there’s going to be a lot of complexity with the next generation, and even with a lot of storage space and a lot of processing power, a team can only have so many artists. Euphoria takes over a lot of the less exciting animation work, with really nice, high quality and fully synthesised character animations. In that way, with next-gen, it will really smooth the production process.

And it will provide really good quality, unique animations every time a game is played, so it gives the player constant variety.

Can you elaborate on that? How will Euphoria actually impact the experiences consumers have playing a game?
Peter Cavanagh: It means players will see more unique moments. They won’t see repeated animations. Euphoria provides so many idle variations and different ways for a character to fall down or get up. Elsewhere, if a player plays a game for hours, or over and over, they will see the same thing with characters, which can spoil the game. With Euphoria, every time is unique; that could be special to the player in certain moments; making their game unique to them. There could even be more emotion, and there will be much more immersion.

How have you prevented prefabricated behaviours taking creative control away from studios, or giving Euphoria-powered games a uniform look or feel? If Euphoria is controlling the game here, couldn’t that take away control from the developer?
Topping: The behaviours do tell characters what to do, but through the parameters available these reactive behaviours are just a framework a developer puts direction into.

Each one of them has numerous ways they can be customised and tuned, and a lot of them have the ability to be modified by extra supplementary animations. They can all be tuned to be what a developer wants them to be, to make a character, for example, strong, weak, timid or forceful.

Cavanagh: And you can cheat the system with various forces and stronger constraints, so if you really want to do something, you still have the control to do that.

Topping: That’s a really important point, because you need to give control to the developer, and give them the power of creative direction. That’s what Euphoria does; it provides something more akin to a director’s tool than what traditional keyframing allows you to do.

What kind of studio is this technology for? It’s clearly next-gen focused for now.
Topping:
Eventually, I think it will be accessible to smaller teams. It is increasingly less complicated to use. It still likes to have access to a reasonable amount of processing power, to be honest, but games like Clumsy Ninja – which Apple demoed last year – show it can run on an iPad. So, yes, Euphoria is a technology for top-end games, but it should increasingly be suitable for smaller teams.

How is the tech made available to studios?
Topping: It does basically come with Morpheme, and it does need PhysX as a physics engine, because the characters are using physics for these movements. That’s the set up at the moment. So now there are two versions of Morpheme. One comes without Euphoria, which is licensed as it has always been, and then there is a supplemental charge for the version with Euphoria. And that’s available now.

And how many of these behaviours exist within the technology?
Topping: They number in the tens, but the point about them is that their strength is in their combination. You could compare it with LEGO bricks, were a relatively low number of brick shapes let you create near infinite different models.

What about the types of games Euphoria is most suited to?
Cavanagh: It is really for any game with dramatic moments you really want enhanced by natural movement. It will help you make those moments more detailed and believable.

Topping: It’s also really well suited to games with things like highly destructible environments, where you don’t always know what state your environment will be in. Euphoria lets characters react to worlds where a character may be moving many objects around or affecting the scene. We certainly think it will be popular with those making shooters, and it will really work well with sports games with a lot of character collision. And even for modern driving games.

IK might be perfect for the driver inside a game’s car, doing the steering, but maybe you’ve got a character hanging off the roof of a vehicle waving a machine gun around: Euphoria would be perfect there. Similarly, imagine a game where a character is thrown from a bike or skateboard; that would be a place where Euphoria would be ideal.

You mention IK there, which is part of Morpheme and many other tools. Considering the job Euphoria does, is it a replacement for IK?
Topping: No, not at all. It’s a complimentary technology. There’s a limited amount of resource on any piece of hardware, however fast and powerful it is, and you want to balance out what you do where.

If you want a character to open a door, there’s no point doing it in anything other than animation with IK, because processor-wise, things like two-bone IK are relatively cheap in terms of cycles, whereas running a full physics rig is harder on that front. IK also has many uses where there is a lot of locomotion. And for things like climbing a ladder, you’d probably use IK systems to make the character hit its handholds. But if you want high quality reactive behaviours, you want Euphoria to make them look fantastic.

www.naturalmotion.com/middleware