1 of 12

Procedural Audio�The Story So Far

PRESENTED BY: KEVIN DORAN

2 of 12

What is Procedural Audio?

​

​

“Procedural audio is non-linear, often synthetic sound, created in real time according to a set of programmatic rules and live input.” (Farnell, 2007)

​

3 of 12

We’re Already Engaged in it

  • Scripting dynamic behaviour (RTPC’s)
  • Programming synthesisers for sound design

​

4 of 12

Beyond Good & Evil (context required)

POTENTIAL

  • Loads of good research
  • Scales well
  • Better interactive systems (VR)
  • Separation & spatialisation
  • Simpler asset management

CONSIDERATIONS

  • Variable cost & stability
  • Choosing the right model
  • Aesthetic
  • Time
  • Limiting parameters
  • Legal issues

​

5 of 12

Rationale

NECESSARY IF

  • Many Control Parameters
  • Uncertain Outcome (procedurally or user-created content)

​

HELPFUL FOR

  • Addressing Storage Limitations
  • Avoiding Repetition

Gameplay Narrative

6 of 12

Approaches

TOP-DOWN

  • Apparent Synthesis
  • Bigger Picture
  • Longer Operational Learning Time

​

BOTTOM-UP

  • Physically Inspired Modelling
  • Concerned With Details
  • Longer Development Time

Examine physical process

Disseminate process

Write a system

Gather examples

Analyse defining characteristics

Find suitable synthesis method

7 of 12

Synthesis Methods

Taxonomy of analysis/synthesis methods by the

types of sounds for which they work well. (Misra and Cook 2009)

8 of 12

My Granular Footstep Generator

  • Footsteps have loads of control parameters (muscular, behavioural, physical interactions)

​

  • Good candidate for procedural audio?
  • Granular synthesis is great for textured sounds, not so great for contact sounds
  • Created Wwise plugin using Heavy Compiler

​

9 of 12

Conclusions

  • Everybody can benefit
  • Worst case scenario: You learn more about synthesis
  • Even in the patch isn’t viable in real-time, you might have a new tool

10 of 12

Resources

​

11 of 12

Ready-made Packages

​

12 of 12

References