What about the face over IP stuff?

Yeah, the face over IP stuff. So just so everybody's aware, it's intrinsically linked to voice over IP.

Because the face comes from the voice.

That's correct, actually. And the way we actually couple the data together, it's the way we synchronize the face and the voice together. So let me give the classic, I guess, Sean explanation where I want to give you all that stuff beforehand. So VoIP itself has a certain amount of data throughput. But to give you an idea, the actual face data on top of that is about one-tenth of what the actual VoIP data is, because when we're sending the face stuff, we're embedding it on the same data stream as the VoIP stuff. And it's nice because it's not that much overhead that actually connects into the VoIP transfer protocol. So we don't send JPEGs or video or images or anything like that to the client. We approach it under the same mentality as we approach procedural planets or anything like that, which is we send the instructions to the client for how to rebuild what's happened on your client, basically. So we send a whole bunch of data over, the client rebuilds it. But it's all connected to that VoIP data. So what's more important for FOIP is that VoIP works. And what's important for VoIP to work is that the group system works. And that basically is chat, group system, and all the rules that go around that. So the FOIP stuff is more or less complete in terms of me sitting here with a webcam driving my face. That works, hooking up to the character creator or whatever. We'll have some fun with it. But we're just waiting on some group system stuff, which is happening for this release. So as group and VoIP come online, FOIP comes in with it at the same time.

I went to AT&T Shape, which was an expo held at the Warner Brothers Studios about three or four weeks ago. And Faceware was there, and they had the demo going, where you're in the bar. It was cool to watch other people, not Star Citizen in any way, shape, or form, sit there and just be amazed at how well it tracks.

And there's new stuff coming up right now. I'm getting a new build from Faceware later this month, and one of the big things that they solved was the calibration issue. The calibration issue, what I'm talking about, is you have to manually calibrate. So that means you've got to stop for a frame, look to the center of the screen and hit a button to calibrate yourself to the neutral. So they're bringing online this new system that auto-calibrates. So basically, you set up your calibration once, you just say calibrate. And as you're talking, it's going to find your neutral and then it's going to map that to the face. This was kind of the biggest quality problem that we ever had, because you could lose your calibration as you were sitting there talking. If you lost your face and then came back in, sometimes the neutral wasn't the same. So this is a super nice thing that'll come in before it actually rolls out to the player. But the next step is just having some good settings and these kind of things. But I don't want to beat that to death too bad.

Are people going to find ways to get them to say calibrate, so it fucks up?

Yeah, and you could calibrate yourself, like cross your eyes and do a silly face or something like that and calibrate yourself to that. So that's why it's nice to have the auto calibration. Because then we don't have to rely on the neutral state of the face.