Matteo Cerquone

Basically we would have the sound for the sandstorm itself come from the actual visual effects of the sandstorm. But obviously it will be moving on the map. So if we have metal rattling or, if you're inside the Javelin, obviously all the sounds will need to come through to a specific area in the level. So we would have the sound for the sand breeze, or mostly for the wind noises, coming directly from the sandstorm itself. But as it moves and approaches different areas, that's where we will place different silent audio, let's call it like that, silent audio trigger spots that will start resonating as the sandstorm approaches.

So depending on where you are in relation to the sandstorm and what's between you and the sandstorm, it will actually alter the audio?

Yeah, that's correct. So basically you can be in different areas on the map and you will have different results audio-wise on what will start resonating around you.

It's a little bit of both. Obviously we need to create some audio for that. Sometimes we use libraries, but it never comes straight out of the library. We always need to retweak things. So once we've got all those layers, as we call them, then we start populating the level. And we will let Wwise handle what sounds need to play and how they will have to play. So in the case of the sandstorm, we basically had, it's called a meter. So we will meter how loud the wind will be, and based on that, it will drive the audio from all the audio triggers that are attached in a specific area in the level. So that means that if you have a wind gust that happens, we will know exactly what the volume parameter is at that specific moment, and that will drive the audio of everything surrounding it. It's a bit like side-chaining, but it's actually the opposite, because side-chaining will basically duck down a sound based on another sound. In this case, we bring up the level of some sounds based on how loud the wind is.

That's one of the things that Wwise allows us to do. It's one of the tools that are inside Wwise, and then we can drive real-time parameters, we call it RTPC, again based on volumes. Mostly you can use it for mixing, but it can be used in more creative ways, just like we did with the sandstorm.

Just recently we got some more help from the audio coders, because one of the main problems that we had with the sounds of the spaceships is that we didn't have many real-time parameters to describe certain things. So, for example, maneuvering, like rolling or pitching your ship. We couldn't really work much on that one because we didn't have any numbers coming out from the engine itself. So we got all these new parameters from the audio coders, and that's when we really need to go back and add much more sound and refactor all of that. Because we want to give more feedback to the player. You want to know when your ship is rolling, pitching, or you want to hear your engine braking. We have engine strain as well. So that means if you're going at a certain speed and you're braking, all of a sudden you can have a ramp-up of the engine so that you can understand how much the engine is working at that moment.

So if your HUD is damaged or broken, or if you're a more instinctual pilot, you can sort of tell how fast you're going and how your ship is maneuvering based on the sounds that the different thrusters for pitch, yaw and roll are making.

Yes, that's right. That's the goal anyway.

It's mostly playing with the different instruments that you have. I like creating glitchy sounds, and that's really what makes things sound not realistic. So you can play with synthesizers, you can play with different tools. One of the ways I created some sounds for the Xi'an Scout was, since they wanted it to be organic, glitchy kinds of sounds, I just recorded dog food, for example, and vegetables and things like that. It makes really squishy sounds.

So obviously dog food doesn't make a sound of its own. You're just pouring it out of a can, sloshing it back and forth.

Yes, exactly. And then from there, it was just a matter of playing with the different tools that we have. And again, it's a matter of creating an entire library of sounds that you recorded and manipulated. And after that, it's a matter of doing a cut and paste and making sure that it syncs with the actual animation.

Yeah, of course it will. We don't have all the tech ready for that right now. But the idea is to have, again, all the different positional sounds in the level. Each different audio trigger spot, each different sound, we can have a blend between different sounds, and they can change according to the weather, according to the wind, according to whether it's daytime or nighttime. We're trying to make things as dynamic as possible, because usually you would have a bespoke event in a game and you would know exactly what's going to happen and when it's going to happen. In these cases it's a bit different. It's more like combining different ideas for different solutions. There's never one solution that addresses them all.

You mentioned that a sound might sound different between day and night?

Yes, that's correct. So one of the things would be the ambiences, for example. If you're in the desert, you would have a specific ambience. If the temperature is quite high, you would have a different ambience. If the temperature drops down and it's nighttime, you would have a different ambient sound.