Introduction to Csound

1. Introduction

CSOUND is a computer software system designed to synthesize digital audio signals and facilitate computer music composition.

Csound achieves precise, sample-accurate control over sound generation at every instant through two complementary files:

Orchestra instruments: Programmatic routines that define the digital signal processing algorithms capable of generating and manipulating sound.
Score events: A structured data file specifying when and how instruments are invoked, containing pitch, duration, amplitude, and spatial coordinates encoded in a standard numerical score format.

Historical Background

Below is a brief introduction to the evolution of sound synthesis software leading up to CSOUND:

1960s — Music 4:
Written by Max Mathews at Bell Telephone Laboratories in the early 1960s. This program introduced wavetable lookup synthesis and established the terminology that allowed computer music researchers worldwide to communicate effectively. Godfrey Winham at Princeton made crucial contributions with Music 4B, and Barry Vercoe’s Music 360 (1968) built directly upon this foundation.
1973 — Music 11:
Developed by Barry Vercoe at MIT, Music 11 established the fundamental architectural separation between audio signal processing and control signal processing—a dual-rate architecture that remains a cornerstone of Csound today. Because Csound was subsequently rewritten entirely in portable C, it could be compiled and run across diverse platforms: Unix workstations, VAX/DEC under Ultrix, SunOS, SGI IRIX, DOS/Windows, and Macintosh.

By establishing a unified, hardware-independent language for digital signal processing along with support for standard audio formats (such as WAV and AIFF), Csound enabled musicians and researchers to transfer projects seamlessly across platforms.

1991:
A major update introduced the Phase Vocoder for time-scale and pitch-scale modification, FOF (Fonction d’onde formantique) synthesis for singing voice simulation, and spectral data types.
1992:
MIDI conversion and control opcodes were developed, enabling Csound to respond to external MIDI hardware controllers and standard MIDI files.
1994:
Audio analysis tools (LPC, PVOC) were directly integrated into the main executable, consolidating the entire workflow into a single command-line tool. Csound also added real-time iterative score generation.
1995:
Real-time MIDI capabilities were greatly enhanced, adding the linseg generator, Butterworth filter opcodes, granular synthesis, and real-time interactive score event handling.