ROS Package bob_synth 

A high-fidelity, low-latency ROS 2 synthesizer node designed for AI-driven audio generation and autonomous bot assistants.

AI Logic & Integration

This package is specifically optimized for AI interaction. It provides a clean ROS 2 API for real-time parameter manipulation and audio streaming.

Objective: Provide a stable, predictable sound engine for AI agents (like Eva).
Communication: Uses standard ROS 2 topics and services for parameter synchronization.
Payload: High-quality 16-bit PCM audio (Int16MultiArray) at 44.1kHz.

Architecture

The system follows a modular, decoupled architecture for maximum stability and flexibility:

Synth Node (C++): The real-time engine. Handles wave generation, State Variable Filtering (SVF), ADSR envelopes, and LFO modulation.
Synth GUI (Python): The human control interface (Moog-style rack).
Audio Sink (Python): The playback engine, available as a Headless Sink for server/docker use or integrated into the GUI. Uses sounddevice (ALSA) for jitter-free playback on hardware (HDMI, Speakers).

ROS 2 API

Topics

Topic	Type	Description
`audio_out`	`std_msgs/Int16MultiArray`	Raw stereo PCM data (44.1kHz).
`/synth_config`	`std_msgs/String`	AI Interface: Accepts JSON strings to update parameters.

Parameters

Name	Type	Range	Description
`frequency`	double	20 - 2000	Base frequency of the oscillator.
`waveform`	string	sine, square, triangle, saw	Oscillator shape.
`mod_frequency`	double	0 - 20	LFO frequency for vibrato.
`mod_depth`	double	0 - 100	LFO intensity.
`attack`	double	0.0 - 2.0	Envelope attack time (seconds).
`decay`	double	0.0 - 2.0	Envelope decay time.
`sustain`	double	0.0 - 1.0	Envelope sustain level.
`release`	double	0.0 - 2.0	Envelope release time.
`note_on`	bool	true/false	Triggers the ADS phase or starts Release.

Getting Started

Installation

Ensure you have the required Python dependencies:

pip3 install numpy sounddevice

Running

Start the Synthesis Engine:

ros2 run bob_synth synth_node

Start the GUI & Player (Desktop)

ros2 run bob_synth synth_gui.py

Headless Player (CLI/Docker)

For environments without a graphical environment (X11) or when only audio playback is needed:

ros2 run bob_synth audio_sink.py

CLI (One-Shot)

ros2 topic pub /synth_config std_msgs/msg/String "data: '{\"frequency\": 220.0, \"waveform\": \"sawtooth\", \"filter_cutoff\": 2500.0, \"filter_resonance\": 0.6}'"

AI Control Interface (JSON)

The synthesizer is primarily designed to provide AI agents with an additional “expressive output layer” (similar to emoticons/icons). Control is handled via the /synth_config topic using JSON strings.

JSON Schema

Key	Type	Range	Description
`note_on`	boolean	`true`/`false`	Triggers the ADSR envelope
`frequency`	float	20.0 - 2000.0	Base oscillator frequency in Hz
`waveform`	string	`sine`, `square`, `sawtooth`	Oscillator type
`amplitude`	float	0.0 - 1.0	Master gain (volume)
`filter_cutoff`	float	20.0 - 16000.0	Cutoff frequency of the SV filter
`filter_resonance`	float	0.0 - 0.95	Filter resonance (Q factor)
`mod_frequency`	float	0.0 - 20.0	LFO frequency (vibrato speed)
`mod_depth`	float	0.0 - 100.0	LFO intensity (modulation depth)
`attack`/`decay`	float	0.0 - 2.0	ADSR time values (seconds)

GUI & User Interface

The GUI is provided for visualization and manual parameter adjustments during development and testing.

Control Panels

OSCILLATOR: Waveform selection and base frequency adjustment.
MODULATION: LFO for frequency modulation (vibrato/experimental FX).
FILTER: 12dB State Variable Filter (low-pass) with resonance.
ENVELOPE: Standard ADSR envelope for volume shaping.
MASTER/TRIGGER: Final output volume and manual note trigger button.

ROS Package bob_synth