A Vision for Sound – Failure’s Audio Part 1

Hi everyone,

This blog post is long overdue, both in terms of not having posted anything for the past two / three weeks (reasons below) and in terms of the topic. This is going to be a two part post, the first of which will be focused on the vision for Failure's Sound Design. The second part which will be released in a couple of months time will discuss the technical aspects of the audio including some of the techniques we used to achieved the vision.

Before I move onto this weeks topic, I wanted to mention the reason for not being more social the past two weeks: we've been incredibly busy putting together our application for the CoreLabs Accelerator - www.corelabsaccelerator.com

I wont go into too many details this week, we're saving that for next week's blog post when we find out if we were successful or not, but let me just say it was a hectic couple of weeks of crunch trying to prepare everything and we're pretty happy with what we managed to submit. As a little taster, here's an image of how the game is currently looking after the latest sprint:

WHY UNITY 5’S AUDIO SYSTEM?

With our previous project, The Tower, I decided that we should use something that allowed for much more control over my audio assets and the mix. I suggested the integration of FMod with the project and it really opened up some great possibilities for interactive music, and much more refined control over all the sounds within the game.

Moving onto Failure we continued to work with FMod but quickly ran into issues as we moved the project over to the Unity 5 Beta. Unfortunately the FMod integration wasn't being updated quickly enough and we were spending too much time trying to solve problems, we decided enough was enough and we would move to Unity's new audio features.

I'm so glad we did as they are really rather wonderful. One of the reasons I wanted to use FMod originally was the ability to mix the audio for game while the game was running as well as making use of snapshots to change mixes on the fly via triggers and events. Unity 5 now has both these features and more, and due to the tight integration I don't need to change a script and faff around to get it working properly, the live mixing just works and setting up snapshots is such a breeze. I do miss some of FMods features, but the integration just isn't as fluid as using Unity's built in tools and for Failure I don't have plans for interactive music (though there is quite an easy way to do this in Unity, though it eats up your voices).

A VISION FOR THE AUDIO IN FAILURE

I went to a great talk a couple of years ago at Develop Conference where Alistair Lindsay talked about his sound design work on Introversion's Prison Architect - he discussed several cool things in terms of the psychology of sound, but one of the things I took from the talk that I'm trying to put into practice here is the idea behind close proximity sounds and the changing soundscape based on zoom level - with prison architect it's a bit different, there are predominantly organic sounds, ie people and objects within the world that create a variety of sounds and when combined together create soundscapes that we're all familiar with. With Failure this is a bit harder as all of our sounds are unknown and unfamiliar, the world is based within a virtual reality, a cyber world - but the world is also alive is some ways, it's an ever expanding network with a core controlled by an ai - each part of the network is like this Ai's nervous system and the data travelling around is it's lifeblood. With this in mind I have some interesting ideas on how I can give the world a pseudo organic feeling using predominantly digital sounds. For now I'm working on layering sound in interesting ways and this starts with the basic ambiances.

Ambiances are very much at the forefront of the mix that I'm trying to accomplish. This includes creating several different ambiances for the actual spaces / levels based on where the player actually is in relation to the network - the outer rims being almost tranquil in it's overall qualities and the central parts of the network having a darker, mysterious and edgy sound. I plan to create layers of different sounds that will make up each ambiance and then control the intensity of elements via snapshots and events - as the player gets closer to completing a level certain elements will be mixed in and out to try and give the impression that the NeuroNet it self is becoming agitated by the player's ability to destroy it's defences. With our new territory system I can potentially link up the snapshot changes to how much territory the player has, though there might be other interesting ways we could do it.

In terms of other elements that make sound, I wanted to make sure that each of our constructs (buildings / towers) also have their own ambience / signature sound - these also have a number of states based on whether they have been captured, just built or upgraded and once again here I am using scripts to blend audio elements together based on their states.

I've set-up the distance falloff curves on most of the sounds within the game to be quite steep as I don't want the players to be overwhelmed by sound, this is especially important for the unit spawn, shoot and death sounds which trigger on a regular basis, they are also mixed in at a level that doesn't make them the focus of the sound scape.

The Audio is still in it's infancy and having to juggle creative direction, design and audio duties has been hard but I'm hoping to spend a bit more time in a couple of weeks really trying to get this audio vision working.

Next time I'll go into detail about some of the ways we've gone about setting up Unity's audio mixer and how we're controlling snapshot changes and triggering sounds.

Until then

Justin French