Procedural Sound design

Luca Petrillo
4 min readOct 20, 2020

--

The post will explain what it is this technique and the reasons why it’s a common technique used in game audio.

-Definition-

Procedural sound design is the technique to manipulate sounds not as already designed wav files but as algorithm. It is used in Video Game audio development to recreate real sounds. Dividing each sound in its own parts. Just think for example at footsteps in general. Footsteps has different part of it: the thump when the foot hit the ground, the tail when you release the foot and even an “attack time” when a person is shuffling. This sound can differ if the footstep is on a different material such as wooden floor or on dead leaves.

When we interact with objects in the real world, the physical processes that produce sound are never repeated in exactly the same way. (Stevens and Raybould, 2016).

-Why using it?-

Dividing sounds in different parts permits to recreate non repetitive sounds and noises inside the game engine, enhancing the immersion of the player.
This is one the three main reason a player would play video games. Immersion players desire to escape real life problems, engage in role-play and “be part of the story (Yee 2005; in press).

It is also useful because instead using wav files as a pre-recorded whole static sounds we can have only few short samples. The sound designer doesn’t have the necessity to record 30 or more variation of the same object hitting different surfaces. This lead to decrease the impact on the memory use of the system. (Stevens and Raybould, 2016)

-Practical Implementation-

It follows a simple algorithm creation to demonstrate the capabilities of this technique. As a DAW is used Nuendo 10. In this example as it can be seen from the Figure 1 there are three Sail samples each with a different spectrum and envelope. Using them as a loop for different objects would be too much repetitive and it could break the immersion inside the game.

Figure 1: Daw Window

In order to avoid this it is taken 4 different samples each for three different movements of the sail through the wind. In Figure 2 it is shown that each sample has his own envelope with different attack and release time. Also they are processed with some filters to emulate different wind forces that move and push through the Sail.

Figure 2: Chunks of Sail samples

After editing and exporting the samples, they are implemented in Unreal as shown in Figure 3.

Figure 3: First Implementation of Sail

This system allows to play the sample in a basic order taking from the heavy “thump” of the wind that hit the Sail to the medium and the Soft movement. Even though it’s a good algorithm it is still repetitive and it doesn’t represent the real phenomenon that is formed by different parts of the Sail that are moving in different times and strengths producing different parts of the sound as a whole system.

So in order to create a more complex and realistic process, it is added another “Concatenate” node that triggers a second stream of samples. It pick up one sample of the twelve stored in a randomised way. Also to create realism it is added a Delay and a modulator to the second stream. At the end, we have 2 streams that are always different in Volume, Pitch and Spectrum.

Figure 4: Sound Cue of the Sail with two stream of randomised samples

-Conclusion-

It is quite clear that procedural sound design is an important tool that sound designers can use in order to populate games with non- repetitive sounds that will increase the immersion and the awareness of the player.

-References-

--

--