Sound Modelling - NoiseSaw

© Ing. Radovan Konečný - radkon, 2012

About the project

The goal of the project was to create sounds similar to those of Supersaw, but by using a completely different algorithm with better features than Supersaw. The reasons are two:

• to achieve a fuller frequency spectrum (Supersaw would need an extremely large number of oscillators),
• to create perfect loops of a given length that will not be great.

In the case of Supersaw, the length of perfect loops is proportional to the density of the oscillators, so a fuller spectrum would mean longer loops - they would be extremely long. So I was looking for another solution to create such sounds, and I invented an algorithm that creates them by using a filtered white noise composition algorithm. The signal of one oscillator is made of white noise, which is filtered by a resonator digital filter with its high quality (Q). This produces a near-harmonic signal, but with the sidebands in the frequency spectrum. The resulting sound is generated by summing a number of such oscillators, exactly for each harmonic component. In addition to the shape of the sidebands, it is possible to determine the overall color of the sound (i.e., the amplitude dependence of the harmonic component from its index). The sound is similar to the sound of string ensemble.

The algorithm is designed to calculate sounds that are perfectly cyclical, so that perfect loops are created. This means that there is not an audible seam-point, because that point does not even exist. For calculation, it can be entered the desired loop length (as opposed to Supersaw), for example 10 seconds.

Sounds of this manner (but not cyclic) can also create my Struck instrument simulator.

Sound samples

(Note: Some sound samples may not work in Internet Explorer.)

Examples of the (raw) NoiseSaw sound (not entire loops):

- NoiseSaw (Q=100)

- NoiseSaw (Q=50)

- NoiseSaw (Q=30)

Usage

The sounds created in this way are put in my remade KORG M1 synthesizer and I use them in musical creation. I have put their stereo version in the Sampler and also use them in musical creation.

This algorithm has been improved with the option to specify the spectral composition of the sound. This modified algorithm is used in the Choir-sound modelling.