Pixel Crushers Forum

Hello,

I have started using an asset named Look Animator in unity, and it seems to be complementing the dialogues I create with DSU very well. I use Lipsync and UMA to animate the lips and face expressions, and look animator really helped to make the characters more realistic.

The only problem is that when there are multiple NPCs, managing who looks to whom during the conversations become really complicated. I am doing some work around to maintain this now but it is not very sustainable in a lengthy game.

I was wondering if it would be possible to consider adding a future integration with that asset, so that, for example, the speaker and listener would look at each other, or for multiple NPC conversations, everyone could look at the speaker but the speaker could look at a specific NPC.

Thank you for considering it.

That sounds like a good idea. I'll check into it for possible integration. This Look Animator, right?

Yes, that is the one. Here is a snapshot of what it does in the game. I really like the Skyrimasque feel it adds.



One scene that I am working on has 4 characters talking on a breakfast table. Using it really got complicated along with DSU for that particular scene, but I believe i will have difficulty for any dialogue involving 3 or more conversants.

Okay, I can add support for that. I'll probably add two options: a component that you can add to characters to make them look at the current speaker, and a sequencer command that you can use to control Look Animator more manually if you want to override the component. Any thoughts? What would be most useful, and with the easiest workflow?

I think it is a great idea to add a component to make them automatically look at the current speaker. I talked to the developer, and he adapted the component that enabled a head-look to be triggered for a set number of seconds for the character the component was on, to make it active as long as you are inside the trigger. This seems to use a function called MomentaryLook, so the main look target inside the component doesn't change, and seems to be manipulated by code.

In a multi-npc dialogue, usually you want all conversants to look at the character currently speaking, and, the current speaker to look at the previous speaker. So it would be ideal for that component to do both things: make character look at the current speaker, and if the character is the speaker, make him look at the previous speaker (if that is possible). And for the exceptions (where the speaker may need to address someone else), a sequencer command for that would be perfect.

Thank you for looking into this so promptly.

Thanks for the info. I'm traveling this upcoming week, so it might be a week before I get it implemented.

Have fun!

Thanks!

Hello,

I wanted to check if you had any updates on this integration.

Thank you.

I just happen to be working on it right now! It'll be in the upcoming version 2.2.0, which I'm aiming to release very soon.