Realistic sounds can be generated by simulating the physics of vibrating materials — no recordings needed. A six-part progression from simple string vibration to complex environmental sounds.
Every sound in the physical world follows the same three-stage process: something excites a material, the material vibrates according to its physical properties, and those vibrations radiate into the air as sound. Model these stages in code and you can synthesise any sound from first principles.
A plucked string as a traveling wave in a delay loop. The simplest physical model that produces pitch, timbre, and natural decay from a handful of parameters.
Wave · Delay · Karplus-StrongThe wave equation as running code. Step through the finite difference stencil node by node, watch the simulation run at audio rate, and see how Karplus-Strong folds it into a single ring buffer.
Finite Difference · Stencil · Karplus-StrongA struck object as a set of independent vibrating modes. Different materials — bell, wood, glass — share the same model but with different mode frequencies and decay rates.
Modes · Resonance · MaterialThe same resonating object sounds completely different depending on how it is excited. Hammer, mallet, bow, and scrape each produce distinct sounds from identical material properties.
Excitation · Resonator · InteractionReal instruments are chains of coupled vibrating bodies. A guitar string drives a soundboard; a clarinet reed drives an air column. Energy transfers between systems shape the final sound.
Coupling · Feedback · ResonanceFriction, turbulence, and collisions introduce nonlinearity. The bowed string's stick-slip mechanism is a self-sustaining oscillation that produces natural variation and attack character.
Friction · Stick-Slip · ChaosFootsteps, scrapes, and impacts synthesised from physical parameters — surface material, contact force, motion speed. Every footstep sounds slightly different, like the real thing.
Footsteps · Scrape · ImpactBased on the work of Perry R. Cook — Real Sound Synthesis for Interactive Applications (A K Peters, 2002) and the PhISM line of research in physically-based sound synthesis.