TL;DRLee Banyard says audio optimization is mostly about paring back the many sound sources and making the managing systems efficient, including thread safety for multiple cores, while keeping quality up amid changes from other disciplines.
Generally, on the audio code side of things, I think people don't necessarily realize the kind of modern problem with game audio is that there's so many objects and things in the game that can all make audio at any given moment. It used to be that you couldn't actually trigger enough audio in games in the old, old days when I started. And now it's just mostly that everything that should have a sound has a sound of some description. And it's actually paring it back. We spend a lot of time doing that, trying to make sure that the experience is optimized for the user. So it's not just in terms of speeding up code and things, it's actually making sure the systems that manage it all are optimal. And at the moment they're doing a lot of stuff with what they call thread safety, which is to make sure it works over multiple cores and stuff like that. That's the term at the moment, isn't it? Thread safety. It seems to be, hearing it from them. So a lot of it is making it more optimal; that is largely what they're doing in terms of optimization right now. But because it's obviously a running release, we're optimizing all the time as we go. And that's still quite difficult for us because we're downstream from so many other disciplines who can throw us a curveball in terms of what their systems do. So it's not just optimizing in terms of just getting it running, which half of it seems to be maintenance, it's keeping the quality up as well.
TL;DRThe audio team confirms they reworked the ambiences of Port Olisar, Levski, GrimHEX and other stations, with new team members doing passes, and also added some new music.
Yeah, I think there might be something strange with your brain or something, Jared. No, I think, like I say, we always... Phil, you can probably actually answer this better than me.
Yeah, we've done some reworks on the ambiences for pretty much all the locations. We've had some juniors join the team recently, and putting them through their paces is first and foremost putting them through ambience passes on new locations. One, it's great to get new ears on each location and stuff and move away from the status quo. And two, to train them up in the editor and get them used to all the tools that we use. And we try and embellish and improve all the locations as often as possible. So I think Port Olisar got a new pass, Levski, GrimHEX and some other space stations and stuff like that. We've also implemented some new music, so you might have picked up on that as well.
They're quite involved locations, and I think because they're quite high-value locations we just keep trying to revisit them and bring them up as much as we can every time. You're always finding stuff wrong; it's never really finished in terms of that.
TL;DRSoundSim is a conceptual device giving consistent rules for sound as an informational stream, still being worked out, and Banyard hopes it eventually becomes a player-controllable gameplay mechanic, though he can't commit to that.
Yeah, Jason Cobb has been doing a lot of stuff. He's based in Austin. The general concept behind that is, you could argue, a device that we use to give an excuse or reason for sound to be, and to communicate what it does within the context of Star Citizen. But it's partly introducing rules that are consistent for it. Because the general concept of it is, and there's technology that exists now for doing this actually, where you analyze your environs with either lasers or something crazy like that. They're actual lasers. But lasers in real life are a lot less intrusive than you think. You can actually derive sound information. Sound is meant to be an informational stream, to inform decisions that the player makes from moment to moment. So that's really what that mechanism is all about. And you can't really think of all the contexts in which you want to resort to it, or what should be SoundSim or what shouldn't. Is it on at this point? How much does it come from the ship? How much of it is an onboard device that each player has? So a lot of it is just working out conceptually as much as the actual system itself. There's a lot of dependencies and stuff with it. So I spend a lot of time thinking about this thing.
I mean, it's weird if you don't have it. What I like about it is that it will be something I hope will be within the control of the player. We'll be able to turn this off, and, I can't commit to this fully, but that's pretty much advice I would say about everything: I cannot confirm or deny. But I hope eventually it will be harnessed as a gameplay mechanism in its own right. Maybe it can be turned on and off for some sort of tactical or strategic advantage. We need to present these kind of elements and tools sometimes, almost for designers to say, "Oh, okay, that's something that we could actually use to fold into a mission or scenario," or something like that. And I think that would really open possibilities up.
I think it'd be really cool in a situation where you're in a big space battle and you get EMP'd or your ship gets disabled and then your SoundSim falls away. And then it's just you stuck with the sound of your immediate surroundings of the ship. You see all the stuff going on and you've lost that sensory input. And then you're really in trouble.
There was a bit we were thinking about, maybe making it so it's processed to such a high fidelity by your ship. And so that would give you almost a tactical reason to park up your ship near to certain action and stuff like that. But again, that's something I've talked about with designers, whether that's something they would actually go for. Maybe that's something that would be painting them into too much of a corner. I like thinking in terms of the sound having meaning like that and being able to actually present possible gameplay things to everybody, really.
TL;DRThe recording session was more free-form jazz than Stairway, using a ButtKicker actuator to make metallic objects resonate at low frequencies in odd ways, and Banyard found jazz keyboard isn't his thing.
It was more of a Jazz Odyssey. If we're going to swap references, Jared, then that was more of a... It's a Spinal Tap reference, Jared will know. And a lot of that was playing around with lots of different low-end frequencies to see where objects would resonate in interesting and sometimes just stupid ways.
It was a ButtKicker. TM. That's a line of essentially actuators, normally used in consumer-level stuff where you simply attach it to the other side of your sofa while you're gaming and you send sound information to it, and it adds yet more information to your physical body coming through that. It's just using it in quite an unconventional and unintended way. It's very good fun. It's quite tiring, strangely. Jazz keyboard isn't really my thing, we found out.
TL;DRThere is currently a hard limit of 256 simultaneous sounds set in Wwise, but playing that many isn't desirable because human perceptual bandwidth is limited, so the aim is to manage what the player hears.
You mean in the game or just in the universe in general, in reality? It's a very open question. In the game, right? I think there's actually a hard limit of 256 at the moment. I think I set that myself in the Wwise project. So that is literally how many you would. And again, that's one of those ones where people assume that more is better, and actually it kind of isn't. Like I said earlier in the show, we spend quite a lot of time working out what the player should be hearing from moment to moment. So actually playing 256 at once isn't really a desirable place to be. There's only so much perceptual bandwidth we have as humans, at least most of us. And that's why it needs managing a lot better a lot of the time.
Analog versus digital. Well, there's a lot of differences between them. There's persistence of vision and persistence of audition, I think they technically call it. They're quite different things. So we think about it quite a lot, actually; that's what I do.
That's the thing: you can have a really, really nasty picture but great audio, and you perceive it. But then if you have a great picture and nasty audio, then you can't engage with it at all.
TL;DRDeep 20 Hz bass is likely to come later, because stacking low-end content muddies the mix and they first need mix technology to control where bass comes from, such as resynthesizing it or sending discrete content to the sub.
Right. There's actually a good reason, I think, that a lot of games eschew that, because it's a lack of predictability of how all the sounds come together. It's really easy, I think, to get a lot of sounds... again, this relates back to the limit of sounds we have, what we mentioned before. If you actually have quite a few sounds playing at once that all have this low-end content, it creates a mush. It becomes very ill-defined, and you'll find that actually a lot of the... I think that's something that's going to come later, when we really turn on the bass, if you like, once we figure out where low end should come from and how to limit that in a more intelligent way. Because I agree, people associate that with it being impressive and powerful sounding. I just really don't want to essentially muddy everything at this point, particularly. It's partly mix tech that we have to bring online for that. We've looked into almost resynthesizing bass, like taking it out of sounds and then selectively putting it back in, which I don't think anyone else has been doing before. So there's plenty of stuff we can do in that area. And I'm keen to get it in, but it's just one of those things where it's more likely to be a bit further along than at the moment.
And also authoring discrete sub content, right? We've looked into that as well. So rather than just routing everything 5.1, 7.1, 9.2 or whatever, we can actually discretely send chosen information to the sub to make sure we don't muddy the mix, rather than have everything sent.
Again, that's another thing. If you put lots of low-end information in sounds that have been sent to stuff like environmental reverb and things, you get that in the reverb as well. There's not much room. It's a jostling room down at that end of the audio spectrum. It just takes extra management.
If we're doing anything clever with metering in Wwise, then you've got these sub-bass frequencies going on that give you a lot of amplitude information that might not even be perceivable.
Yeah, a lot of people don't have that stuff. A lot of our users are headphone-based, and actually we tend to get quite a bit of low end of a type through headphones, but you don't get that physical sensation as much, that kind of hitting you like a wave that you get from speakers.
TL;DRStar Citizen has a jukebox that Banyard would like to extend, but streaming player music to others is a legal issue he avoids, and a musical instrument system is only an idea he asks not to be held to, not high priority.
It's not quite the same thing. We've kind of got the jukebox, and I'd like to extend that further so that people can do that more. One thing that is a problem with that is streaming it out to other people. And that's a whole legal thing that I'm just going to back away from. I'm just not touching that. But yeah, you should be able to do that. Playing it in Media Player as a separate application entirely, it wouldn't be spatialized as if you were playing it in the world. So that's generally what people want to be able to do. And I know I would want to be able to do that as an alternative to having, say, the score. Really, they should be able to coexist quite nicely. I think you could do that if you had a specific device that represents your music device in the world, the equivalent of Alexa or something: "Play me this in my ship" as I'm trekking.
I want people to be able to have a musical instrument system and learn how to play in the actual universe as well. I've mentioned this, please don't hold me to that. I want people to be able to play gigs in venues. Disaster Area sort of stuff. But I just see it as a possible valid thing, if there's that many possibilities. I've talked about it before on one of the forums, but as yet, it's not high priority.
TL;DRPretty much all sounds are custom: library material is only source that gets layered and processed in a DAW alongside their own recordings, and machine learning for dialogue variations is experimental and a long way off.
Well, pretty much all of them. Very rarely would we ever drag and drop from a sound effects library. Typically what we do, at a very basic sound design level: if you want to design something, say like an ambience or something, we bring in lots of different source material and then track-layer and make sure it supports all the different frequency content bands, like lows, mids and highs and stuff. And then process it and render that out as a new file. So it's not like we'll get wind and then just cut that out and drop it in the game. We'd always take it into a DAW and combine it with other different sounds.
A DAW is a digital audio workstation. Like Reaper, Pro Tools, Cubase, Logic, all that kind of stuff. So yes, we have access to loads of sound libraries, but we purely use that as source rather than end-user material that we push out to the game.
We do a lot of our own recording of our own source as well, hence the freeform jazz metallic madness. Because I think that's important, to get source that has unique provenance, when you know you've got something that no one else has experienced. It's a combination of both things.
You get a high burn rating on some sounds. Everyone knows the Wilhelm scream, right? When we hear that, you just cringe. I'll be watching TV shows, films, and there will be certain sound effects that you just cannot use, not even as source, like servo stuff, some pressurized air stuff, there's a flame whip that's overused all the time. And even if people aren't aware that they're hearing this repetitive material, people will subconsciously know that it sounds like something else. So we always try and move away from that and try and record as much as we can.
Yeah, the brain is a lot less forgiving to repetition than the eyes. Interestingly, I think there was a question I saw about how do you ensure that wild lines and stuff aren't so repetitive. A lot of it, we're looking at ways to use, say, stuff like machine learning perhaps, to process dialogue lines so that we can almost procedurally do variations rather than have to go back to the source and back to the actor. And that's very experimental stuff that we're doing with that. And that's a long way off. But it's in our minds, that sort of thing; you've got to keep bearing it up.
TL;DRYes, ship and weapon sounds are meant to reflect swapped components, but how to do it is still up in the air because every combination must sound good together, possibly via the spectral-aware mix system.
Yes. It's always been on the radar to do that because you want it reflected, right? And especially when we're looking at revising some of the ship audio anyway. And I think if you want to hear the difference it makes to your ship, the ship is where you live, especially at the moment, apart from Port Olisar, obviously, Jared. Your ship is kind of almost an extension of you. And I think if you customize it, you want to hear that happen. It's a bit like if your car depreciates or something over time... I've only had one car in my life, and I got quite attached to just the general sound of it. And anything I did, if I got it fixed up, it was quite satisfying to hear the change being imparted to it. How we approach that is still slightly up in the air, because I don't want people to actually just create an audio mess, put a few different components together and it just becomes unpleasing. So we have to look at how we manage that again and how we limit that but still keep it good.
Yeah, I think about all the Need for Speed games that I played growing up. And every different modification you could do to that car, like you get a turbo, you hear a turbo and stuff like that. I fully imagine we're going to realize that in the game. And same with weapons and stuff. I know the Mass Effect series did a great job, and Dead Space, for having customizable weapons. And each customizable element changed the sound. But like you say, it's a difficult job because you've got to introduce new sound, but you don't want to just have a combination that doesn't work. We have to make sure every combination will work together.
Yeah, I think it kind of almost goes back to that mix system, where if we have something that knows the spectral footprint of these sounds, we'll know how they can come together and which one should win. And if they impinge upon each other, how one needs to duck down a bit. And recombining lots of elements in ways that we can't possibly predict is partly what this game is about. So we spend quite a lot of time on that.
TL;DRThe non-diegetic score won't be altered by in-world effects like grenades, though diegetic music and supporting effects would be affected, and dynamic music responds to intensity but shouldn't be made too reactive or reveal action early.
Well, it's interesting because we've got the split between diegetic and non-diegetic. Diegetic is in the world, so in the situation that you are. So you hear music from a radio, that's diegetic. Now, when you hear score, like you would hear in any kind of film, that's non-diegetic. So nothing in the game would really affect how that would play back with regards to sound quality, because then you're bridging the gap between the two. But certainly for anything that happens in the game, so if you've got music coming from a radio or PA and a flashbang or grenade goes off near you, that's an attack on the senses. So as well as how it would affect you visually, there'd also be supportive audio effects, taking out the top end, the typical thing, people bring up a high whine and stuff like that. But I don't think there's anything that would change playback speed or quality of playback. But obviously the game responds to action in the game. So if you're doing nothing, you're stood around in FPS, you'll get ambient stuff. And then if enemies appear and combat starts ramping up, all the music responds to intensity and mood. And that'll blend between different intensities of music and music suites and stuff like that.
Again, it's one of those things where if you make the score, the dynamic score, too reactive, it can devolve and dissolve into something of a mess. So you have to be quite careful and fussy about what you tie it to. So if you did tie it to certain things that are easy for it to be reproduced at a high rate, that can just lead to stuff that doesn't really give you anything good.
Some people like that. Some people like almost to get gameplay cues from the score, and I've noticed some games do that, and that's fine if that's the way they want to go. I just think that should be more something that the game audio does, as a proper real informational system. Obviously, these rules can be broken and bent depending on the situation, but my general rule is that there should be some other information stream, and then the music in the score can be apt to confirm that, perhaps. Because you just don't want it giving the surprise away before it really happens.
TL;DRBanyard likes the idea of an age and use-based degradation parameter on objects so wear changes their sound until repaired, and says they've pushed a serialized variable for it, without confirming a specific implementation.
It's actually something we talked about probably a couple of years ago or more now. I always thought it'd be great to have an age and per-use parameter attached to almost anything you can come into contact with. So even like doors, cupboards, how often you use your bathroom, whatever, that could be used to inform a certain parameter that would tie into depreciation, or wear and tear. Degradation. I'll name the parameter degradation. As long as it does the job, I don't mind. That's actually something that we've been pushing, a serialized variable, I think they call it technically. I'm not that code-y. I mean, that's fascinating, actually. You could have something that you use thousands of times, and you could actually have it subtly change over time and become something different. And then you go and get it repaired and it's back to pristine again. I think that sort of long-form kind of stuff is fascinating. It hasn't really been done much.
Absolutely, because I think there's a big simulator part to the game. Imagine if you don't maintain your car, then every now and then it's going to fail. So imagine you get into your Hornet and you've just been smashing it in the 'verse and you've not actually taken care of it, and then something happens: you get in your ship, false start, panel falls off.
It could apply to so many things in the game. It's particularly persistent if you have clothing that does that. That can wear down, and guns can somehow start to jam and sound gritty. It would have to reflect that in the actual jamming, but there's a lot of things we can do to indicate that, hey, this is on the way out.
TL;DRThat was the ButtKicker session, where they vibrated scrapyard metal at different frequencies to gather source for ships entering atmosphere or under duress, and they also do bespoke recordings such as weapons and plan more.
Well, I think that's pretty much it. That's the ButtKicker session. We booked out a studio, padded out the live room, and we actually went to a scrapyard just down the road, and we threw all this scrap in the back of a van. We turned up and were like, "We want to vibrate your metal. Is that cool?" And the guy was like, "Yeah, all right." Don't really know how to take that. So we were going around hitting things, dodging rats. And then we took it to a big studio, and we got a Reaper session with a MIDI keyboard, played that through a synth into the ButtKicker.
And we just played different profiles and vibrated different constructions of metal and stuff. And that was really just to try and capture different materials vibrating at different frequencies, which we can use as source to implement for ships when you're coming into atmosphere or whenever a ship is under duress. But we've done other recording sessions. I think a while back we did a cool gun recording session at Copehill Down, which is a military training facility, and we recorded .50 caliber machine guns, a .50 caliber sniper rifle called the Destroyer. Now, I was about six meters away from that when it was being fired, and the pressure wave that came off that was like compressing my sinuses. It was insane. But that recording session wasn't about getting the shot of the weapon. We put microphones all over the simulated Eastern European town, and it was to get the reflections of all the weapons, like the slapback.
But yeah, we try and get out and record as much as we can. Because sometimes, frankly, the scope of this game is so vast, there isn't stuff there. There aren't libraries that we need to actually draw material from. So sometimes we have to go out and record bespoke stuff.
Yeah, we're going to do more of it as well. It's something we want to do more of over the next year, or forever, really. It's definitely one of the perks of getting out and getting stuff, knowing this hasn't been heard anywhere else. Especially the weapon stuff. There's so much scope.
TL;DRNo single weirdest source is named, but a colleague recorded household objects like a plastic cup at very high sample rates to slow them into alien sounds, and Phil recorded himself choking for breathing sounds.
Weird sound source. Darren did a load of stuff with just very strange household objects, but he was recording at such high bit rates and sample rates. They were just stuff you wouldn't expect, and you went, "What is that sound? It sounds massive." "Oh, it's a plastic cup." That sort of thing. He went through that session and spent about a week at home just digging out all sorts of random strange stuff. I don't know what would be the weirdest sound source.
Well, that in itself is one step down the rabbit hole, because if you're recording small objects at 192k, it means you've got so much information that you can slow it right down. So you can take something tiny like a set of keys and just jingle them. But then because you're recording at such a high fidelity, you can slow it down and it sounds like huge, weird, alien tonal bells. So it's stepping into a spectrum that we've never really perceived before. And using those 100k mics, where it can pick up to 100,000 hertz. We can only hear to 20k and this is 100k. So it's all this frequency information that we can't perceive. Then we pitch it down, and there's all this stuff going on that we're never aware of usually.
And some of that was just really strange stuff. But you kind of forget what it is once you start hearing all these strange sounds. So it's probably something that he... I mean, Darren is one of the weirdest members we have anyway, and if anyone wants to find a weird prop to use, it would be him.
I think me personally, the weirdest thing I recorded was myself choking, for the breathing stuff. Nearly passed out a couple of times, nearly threw up a couple of times, legit.
TL;DRThey are working toward automatic raycast-based room detection combined with hand-marked bespoke locations and door propagation, while balancing realism against fun, such as not modelling long-distance gunfire delay one-to-one.
So this is slightly a crossover. We were talking about this before, actually: a technical question with how environmental sound, or acoustic effects, I think we refer to them as, work in Wwise and stuff at the moment. We're trying to come up with a way to make that automatic by having raycasts and stuff out from the player. So a system level, that word again, where the game knows what kind of room you're in and then adjusts automatically. I've done a lot of markup-type stuff before, where you manually go in and set out a room to sound of a certain size and dimensions and that sort of thing. I want us to be able to do that for certain bespoke locations, because I think humans can do a better job of that ultimately than maybe a system can, but that should be interleaved with a system that chooses the right preset or whatever for doing that, and perhaps layers them as well. And there's so much more work to be done in that area. It tends to be a bit more 2D at the moment. And I think the propagation system, which I think I mentioned already, once it takes all those acoustic effects and ports them properly through doors and stuff, it would actually bring a new level to the perceived realism of the game. It's quite a big subject, actually. I think we spend quite a bit of time thinking about that.
The trouble we face is we want to model how audio works as true to real life as possible. However, if you do a one-to-one model, then it might sap all the fun out of the game. So we have to balance that. So if you're getting shot at, the whole crack-thump thing: you hear the bullet whiz by you, and then you hear the thump way off in the distance. If we model that one-to-one, the distances we deal with in the game are huge. So if you've got a futuristic sniper rifle that can fire over, say, a 10-mile headshot, then you'd never perceivably hear the gunfire. You'd just hear the bullet go by. Some people might like that and stuff, but we have to make sure we balance the fun of the gameplay as well.
It's usually where Phil will say stuff like "rule of cool". I don't think there is a rule of cool. I think there's maybe a set of rules that may end up cool if you're lucky.
TL;DRSome work like source gathering and manufacturer branding can start early, but sounds are mostly implemented once a ship is flyable, and Banyard thinks ship audio is currently too thruster-led and keeps needing rework as the game changes.
Well, we can do appreciable audio work in all stages. So as it comes through whitebox, we can basically start gaining source for the different types of thrusters to give that ship character. But it's not until it becomes actually flight-ready that we can actually start implementing the sounds and actually play. We almost use the ship as a controller. So we actually pull these sounds in and we try and manipulate it to make sure it sounds right and stuff like that. So really a lot of the work gets done as and when the ship becomes flyable.
You try and get in and start thinking about it as early as possible, at concept stage. We want to do more work in the area of branding, that sort of thing, so that different manufacturers have a different signature to them in some way. So there's some work we can do outside of ships being ready: if we've got some idea of a brand identity of different manufacturers, we can start work there and keep evolving that too. So in a way, it's another one of those things that's ever ongoing. And it's another big job. The ship pipeline is probably one of the most consuming ones; there's a lot of ships in our game. And we're always reviewing them. We just had another review session just the other day, which had some interesting conclusions, I think. And we're going to improve those again. In my own opinion, it's a little bit too thruster-led at the moment. It's one of those things where you go down a certain path trying to get that right first, and then you bring that back and start to emphasize other elements again. That's kind of where we are.
Again, I think that comes from simulator versus fun. You want to make sure it's not too literal and not using too much material, but we know there's a context of hearing that sound in atmosphere when it's going to be in space and stuff like that. So they constantly get iterated upon. But there's dialogue assets, UI assets, thruster SFX assets and stuff like that. So there's loads of stuff to work on.
And it's hard to get it all right when it's still very much in flux further upstream. I think a lot of people don't necessarily realize that. It's something I keep banging that drum on. It's frustrating for us as well. If the game was finished, finished, finished, then you could make so much more progress on the audio in quite a little time. But the reality is it's getting better everywhere else all the time. We have to try to keep up with it.
And we're not always dealing with normalized figures. So it's not always zero to one. They might just say, okay, this ship now can go twice the speed. We rebalance all the audio then.
TL;DRVoice commands are hard to integrate well and would need to be reliable and user-trained, but Banyard says it needs more exploration and hopes the game gets something like it.
Yeah, it's one that actually came up quite early on, and we looked into the VoiceAttack thing as well. And it's a hard one to actually interleave properly into the game, because I think the reason why VoiceAttack works so well is it's very, very customizable. There's a lot of UI devoted to that. And it's just a lot of engineering, I think, to maintain a system like that. It's something I think we need to explore more, because I think it does have its advantages. And it is kind of fun. I totally get the idea. How much we need to be actively involved... I think, as you said, it's almost like a gameplay or a system design thing, because it's another control mechanism, and it won't be suited for everything. For it to work, I think it has to be trained properly by the user. It's very difficult to get stuff to work just out of the box. I think that's evidenced by how much money has been spent by the likes of Amazon, Apple, et cetera, on getting that stuff to work in the real world. If you're relying on it in particularly twitchy situations in the game and it fails you, it's like, well, that's a mission that's been put out. You've got to fail at some point with it. And I'm very aware of that. So we'd have to get it pretty solid before we would roll something like that out. But I hope we do. I can definitely see... this is what we associate with the future in a way, and we kind of have it now, but we'd expect it to be far more accurate in the time that we're trying to model. But it's a tricky one.
TL;DRThe related voice and face-over-IP work is ongoing but partly out of their hands because of dependencies on other systems and teams; they're supporting it and expect features to roll out together eventually.
Yeah, it's part of VoIP, I suppose FOIP, Face over IP. It was kind of folded into that. It started off as pretty much a VoIP thing and slightly bloomed out. It's still an ongoing thing. It's slightly out of our hands, because there's interdependencies with other systems and other in-game applications that we're not working on directly. So we're supporting as much as we can and trying to push it through as much as we can from our end. And a lot of our tech is in place. There's other sort of gnarly stuff, I suppose, that we can't necessarily do directly. That's kind of what we're waiting on.
It's just a bit of a beast. Because the voice is the main player interaction with it, people assume it's audio-owned, but there's so much going into it which is on so many different teams and so many different systems. Obviously the voice interacts with the system, but there's so much under the hood that takes into account different disciplines. But I think it'd be a great feature to be able to fully flesh out: to be able to speak in the world, in the game, and have your voice sound as if it's propagating in that space. We've got a tannoy system. We could have a microphone in your ship on the bridge, and you could press a button and then your voice could be propagated through different PA speakers in the ship, like an Idris or something. And you can communicate directly with engineering and get your engineer to do something. That's the dream. That'd be amazing.
There's some other system bits that have come recently that allow a lot more of that stuff, and you can see how it will all join together, and I think we'll hit some sort of tipping point with it, and suddenly a lot of that will roll out at once, and it'll be almost like a sport, the stuff you can do with it all.