MuSA_RT MuSA_RT animates a visual representation of tonal patterns - pitches, chords, key - in music as it is being performed.

MuSA_RT applies music analysis algorithms rooted in Elaine Chew’s Spiral Array model of tonality, which also provides the 3D geometry for the visualization space. The app analyzes the audio signal received from a microphone to determine pitch names, maintain short term and longterm tonal context trackers, each a Center of Effect (CE), and compute the closest triads (3-note chords) and keys as the music unfolds in performance.
The 3D view presents a graphical representations of these tonal entities in context, smoothly rotating the virtual camera to provide an unobstructed view of the current information. These graphics can also be visualized in an experimental Augmented Reality on compatible devices.

User Guide (v. 3.0)

MuSA_RT menu

  1. Start the app and give the app access permission to the microphone when prompted
  2. Tap on “3D” to navigate to the live 3D visualization screen.
  3. Play some music (or sing?): MuSA_RT should start reacting.

To try out the Augmented Reality mode, tap on “AR” (available only if the device supports AR).

Note: MuSA_RT requires a microphone or some audio input device in order to function in Audio mode, or a MIDI input device to function in MIDI mode. MuSA_RT does not generate sound.

Visualization: Spiral Array

The Spiral Array Model is a geometric model of tonality invented by Elaine Chew.

Pitch, triads and keys correspond to locations in the 3D space contained inside the outer spiral (pitches). Pitches along the pitch spiral are separated by a perfect 5th; a given pitch is separated from the pitch directly above by a major 3rd. Major triads are therefore compact upright triangles. Similarly, minor triads are compact triangles. Centers of Effect track the long term (key and spelling) and short term (triad) tonal context.

Spiral Array model of tonality

Settings: Visuals

Pitch / Triad / Key

Select and configure the objects presented in the 3D visualization: pitches, triads, key and associated Center of Effects (CEs).


Face labels inward: when enabled, the pitch labels face towards the axis of the spirals and if orient is on, the spiral array rotates to show the active triad on the far side (from the viewpoint adopted in the 3D view), like in the original MuSA_RT. If this switch is off, the labels face outward and the active triad would be displayed in front, a setting more suitable for the AR scenario.

Orient: when enabled, the model smoothly rotates so that the currently active triad (if any) faces the “front” of the array in AR (corresponds to the user’s viewpoint in 3D mode)

Adjust height: when enabled, the Spiral Array smoothly moves vertically so the active triad remains centered at a reference height, and remains vertically centered in the 3D view.

The movement time constant determines how fast the motion occurs (the higher the time constant, the slower the motion)


Select dark/light/auto background mode.

Settings: Spiral Array

Adjust the parameters of the tonal analysis model.

Activation threshold: ignore activation values that are below the threshold (noise)

Time constants for CE computations: lower values make the object more responsive, higher values make the objects more stable. Adapt affects how the object reacts to signal, decay affects how quickly the object fades away in the absence of signal.

For example, to make active triad tracking more responsive, decrease the time constant value for Triad CE Adapt.

Settings: Input

FFT (audio) input MIDI input

Select and configure FFT (audio) or MIDI input, visualize activation levels for pitch classes


Select an audio input device from the dropdown list.


View connected MIDI devices. To connect to a Bluetooth MIDI device, tap the antenna button. Set the Bluetooth device in pairing mode. When the device appears in the list, tap Connect.

Tap the antenna button to initiate Bluetooth scan Searching for Bluetooth LE devices Device not connected Device connected Bluetooth MIDI input

For input activation modeling, choose whether to use sustain pedal signal, and whether to apply an activation decay. The time constant determines how fast the activation decays (the higher the time constant, the slower the decay).

Privacy Policy

The MuSA_RT app does not collect or share your personal information.


MuSA_RT Copyright 2012-2024 Alexandre R. J. François.

MuSA_RT 1.0, released in 2012, was originally commissioned by Elaine Chew’s MuCoaCo Lab as part of the Music on the Spiral Array . Real Time (MuSA.RT) project, and as companion software to her Springer ORMS Series book Mathematical and Computational Modeling of Tonality: Theory and Applications (2014).

Watch Elaine Chew talk about using mathematical and software tools to analyze music.

Audio input

MuSA_RT gratefully uses AudioKit v5 for audio and MIDI input and for low level audio processing (FFT).