A gesture-controlled audio mixer can translate hand movements into practical control signals without relying on conventional faders or knobs. The Theremixer design applies a theremin-inspired approach to audio mixing: two antennas sense a user’s proximity, oscillator circuits turn changes in proximity into electrical signals, and a microcontroller uses those signals to adjust music playback. The concept offers a useful example of how sensing, analog electronics and embedded digital signal processing can work together in a single interface.
From theremin gestures to mixer controls
The traditional theremin uses antennas and hand position to influence sound. The Theremixer adapts that interaction for a different purpose: controlling selected properties of a song rather than directly playing notes. Its intended controls are bass, midrange, treble and volume. A compact housing would contain the electronics and user interface, with two antennas extending from the enclosure.
One antenna is assigned to function selection, while the other detects magnitude. In the proposed design, function selection determines which audio property is being controlled, and magnitude represents the degree of adjustment. This separates the user’s choice of parameter from the amount of change, giving the gesture interface two distinct sensing roles.
How theremin antenna sensing works
The antenna circuits are designed around the fact that a nearby hand changes the antenna’s effective capacitance. That change influences the behavior of a connected oscillator, producing a frequency shift that can be measured electronically. In this way, the antenna does not need to make physical contact with the user: proximity changes the electrical conditions of the sensing circuit.
The magnitude-detection channel uses a variable-frequency oscillator intended to respond to hand position near its antenna. The function-selection channel uses a variable oscillator alongside a fixed-frequency oscillator. A detector compares the two signals through heterodyning, producing an output related to their frequency difference. The design proposes using that output as an input for interpreting hand movement in the function-selection channel.
These circuits require calibration because antenna behavior depends on effective capacitance and on the surrounding environment. The design calls for measuring antenna capacitance with a vector network analyzer and checking oscillator outputs with an oscilloscope. Tuning circuits are included to adjust the oscillator response toward the specified operating ranges. Such measurements are important to verify the intended behavior rather than assuming that component values alone will produce it.
Converting sensor signals into audio adjustments
After the antenna circuits generate analog signals, the proposed signal chain sends them to an analog-to-digital converter (ADC). The ADC communicates with a Teensy 3.6 microcontroller over SPI, allowing the controller to process digitized sensor inputs alongside audio read from a WAV file on an SD card. The concept uses a 40-millisecond buffer to apply processing before sending the mixed audio to a speaker.
The planned audio adjustments use frequency-selective finite impulse response (FIR) filters. The design describes transforming audio into the frequency domain with a fast Fourier transform (FFT), multiplying the result by a filter transfer function, and producing an output with the selected range boosted or attenuated. The intended bands are:
- Bass: 0–250 Hz.
- Midrange: 250–2,000 Hz.
- Treble: 2,000–6,000 Hz.
- Volume: overall output-level adjustment, with a proposed range of 50% to 100%.
For the bass and midrange controls, the design aims to leave frequencies outside the selected band substantially unchanged. The treble filter is intended to affect the 2,000–6,000 Hz range while avoiding changes below or above it. Separating the audio into defined bands makes the relationship between a gesture and its effect more specific than simply changing the entire signal level.
Verification, tolerance and integration considerations
The design sets measurable targets for its subsystems, including oscillator frequencies, antenna capacitance changes, ADC sampling rate and filter behavior. Proposed checks include measuring supply rails with a multimeter, inspecting waveforms with an oscilloscope, and using spectrogram analysis to assess whether filters affect the intended frequency bands. These tests connect system-level goals—responsive control and useful audio adjustment—to observable electrical and signal-processing behavior.
Environmental sensitivity is a central design consideration. Nearby materials can affect antenna capacitance, so testing is intended to take place in an area where other metal objects are kept away from the antennas. The document also identifies calibration and output latency as risks: the sensing circuits need to produce stable, usable signals, and the processing path must respond quickly enough for gestures to feel connected to the resulting audio change. Tuning, measurement and system-level testing therefore matter as much as selecting the individual components.
Power and construction also require attention. The proposed supply uses an AC transformer, a rectifier and positive and negative voltage regulators to provide rails for the circuit. The housing must provide space for the electronics, controls, antennas and power inlet. Safe lab practice, sound electrical connections and care around line voltage are part of the design and test process.
Practical implications for interface designers
The Theremixer illustrates a broader engineering pattern: a non-contact sensor can feed an analog measurement circuit, an ADC and embedded software to create an alternative human-machine interface. Its specific application is music mixing, but the design process highlights considerations relevant to other gesture-driven systems: define what each sensor channel controls, establish measurable signal targets, calibrate for the installation environment, and verify the complete path from input to output.
For the Theremixer concept, the practical challenge is coordinating antenna sensitivity, oscillator stability, digital sampling and audio processing so that a hand movement produces the intended, timely adjustment. The design’s staged verification approach provides a way to examine those links individually and then as an integrated system. More broadly, it shows how established sensing principles can be combined with embedded processing to create a distinctive control interface.