MIDI CC 120: How All Sound Off Works

MIDI Continuous Controller 120 (CC 120) functions as the "All Sound Off" command, a specialized Channel Mode Message designed to instantly silence all active audio generation on a designated MIDI channel. Unlike standard note-ending commands that permit sounds to naturally decay through their programmed envelope release stages, CC 120 commands the sound engine to bypass the release phase and mute all voice output immediately. This article explains the protocol mechanics of CC 120, its technical execution within sound engines, and how it differs from related panic commands like CC 123.

The Anatomy of the CC 120 Message

Under the standard MIDI 1.0 specification, CC 120 is classified as a Channel Mode Message rather than a continuous modulation parameter. It is transmitted as a three-byte packet over a single MIDI channel:

1. Status Byte: 0xBn (where B denotes a Control Change message, and n represents the zero-indexed channel number from 0 to 15).
2. Controller Number: 0x78 (the hexadecimal equivalent of 120).
3. Data Value: 0x00 (MIDI specifications dictate that the data byte for CC 120 must be sent as zero; values other than zero are typically ignored by compliant receivers).

Because it targets a specific channel, CC 120 only silences instruments and voices routed to that particular MIDI channel, leaving other channels untouched.

Immediate Voice Termination vs. Release Phase

To understand how CC 120 achieves immediate silence, it must be contrasted with normal Note Off commands and CC 123 (All Notes Off).

When a standard Note Off or CC 123 message is received, the synthesizer transitions its Amplitude Envelope Generator (AEG) into the "Release" stage. If a patch features a five-second reverb tail, a long release phase, or an active sustain pedal (CC 64), the sound continues to output audio until the envelope completes its trajectory or the sustain pedal is released.

In contrast, CC 120 instructs the receiving digital signal processor (DSP) or analog voice circuitry to instantly attenuate the gain of all active voices assigned to that channel to zero. It overrides the following parameters:

• The AEG Release Stage: Envelope times are entirely bypassed, stopping oscillator output on the next audio buffer cycle.
• Hold/Sustain Pedals (CC 64): The active state of the damper pedal is overridden for currently sounding voices.
• Voice Allocation: Polyphonic voices tied up by runaway or hanging notes are immediately unallocated and returned to the instrument's idle voice pool.

Handling Internal Effects and Resonance

The immediate nature of CC 120 specifically targets the voice generation block. In modern hardware and software instruments, voices are routed through filters and internal effect sends:

• Oscillators and Samplers: The raw sound source stops calculating or reading from memory tables instantly.
• Filter Resonance: Self-oscillating filters tied directly to the voice architecture are clamped to prevent lingering ringing.
• Effects Sends: While CC 120 immediately stops the voice engine itself, the handling of global time-based effects (such as external or master bus delays and reverbs) varies by manufacturer. Some synthesizers flush their internal delay lines and reverb buffers upon receiving CC 120, while others allow existing decay reflections to dissipate naturally once the primary voice signal has been silenced.

Implementation in DAW and Synthesizer Architecture

Most digital audio workstations (DAWs) and virtual instruments bind CC 120 to their software "Panic" functions. If a sequence experiences dropped MIDI Note Off packets—often caused by interrupted loops, unplugged MIDI interfaces, or buffer overruns—hanging notes can lock up audio polyphony.

When invoked, the DAW dispatches CC 120 across all 16 MIDI channels simultaneously, forcing a hard reset of every active audio voice across the entire arrangement in a single processing block.