Why Don't More Games Use This!?

[Soviet Heavy]

Valve's Source Engine has a piece of technology that I find amazing. Specifically, it is a phoneme editor that matches lip movements to whatever script and recorded dialogue is put into the game. The best example of this is probably in the Meet the Team Videos for TF2. Compare the English version to the Russian Version of Meet the Heavy.

That's almost perfect lip syncing.

So why don't more games use this technology? If it has been around for years now, why haven't more developers added it to their games with multiple languages? A game like The Witcher 2 could benefit greatly from this, since the lip syncing in that game is closer to the original Polish than the English translation.

 

[Jimmy T. Malice]

The Source Engine actually does that? Wow. That is awesome.

 

[dessertmonkeyjk]

It might be due to the dedicated program Valve has just for the lip syncing of characters. I think it's called FacePoser but I don't know if it's only functional with the Source engine but it does likely require the model itself to be setup properly before use.

Still, if similar tools like this were available then I would definitely be questioning why it's not commonly used if not for budget reasons.

 

[Loop Stricken]

It looks cool, yes. First time I've noticed it, what with me not making a habit of watching videos in languages I can't understand.

But why don't more people use it? Possibly because, thanks to a quick Googling, Valve don't seem to distribute the library in any SDK.

 

[Zhukov]

Wait... that's done automatically? Huh. I actually didn't think that was possible. Assumed they did the various facial animations by hand. Cool.

As for why other people don't use it... buggered if I know. Is it widely available?

 

[hoboman29]

They don't because they're afraid to put out a game that looks like it didn't come out of the game-a-tron 9000. Because consumers neeeeevvveeeerr want to play good games

 

[valleyshrew]

I remember being amazed at this when it was patched into the original counter-strike a long long time ago and I suddenly noticed people talking on mic's character's lips were moving with what they were saying... It's pretty awful graphics by today's standards (e.g. heavy rain, la noire), and would only work for a cartoony game. Guarantee half-life 2 didn't use this.

 

[Scrustle]

Wow I had no idea there was even technology to do that kind of thing! That's awesome! Seems like a total no-brainer to use it in everything!

 

[Valdsator]

I remember being amazed at this when it was patched into the original counter-strike a long long time ago and I suddenly noticed people talking on mic's character's lips were moving with what they were saying... It's pretty awful graphics by today's standards (e.g. heavy rain, la noire), and would only work for a cartoony game. Guarantee half-life 2 didn't use this.

The GoldSrc engine games (Half Life 1, Counter-Strike, etc.) did not use this technology. They instead made the mouth move automatically with the volume of the sound, which allowed Valve to easily implement it in multiplayer games. The Source engine that Half Life 2 and other Valve games use handles this differently. A program reads the sound file, and also some text that the developers write (which is what the person is saying), and using this information it's able to automatically choose from different mouth movements for each part of that one sound file. So, a different movement for the letter L, S, etc. Every Source game uses this technology, and it looks great.

 

[Soviet Heavy]

Wait... that's done automatically? Huh. I actually didn't think that was possible. Assumed they did the various facial animations by hand. Cool.

As for why other people don't use it... buggered if I know. Is it widely available?

Well, so far the only engine I know that uses it is the Source Engine. But the Source Engine is seven years old now. I don't know why more developers aren't trying to implement this into their own engines.