resona

README

🔊 Resona

Resona is the friendly, fast, and flexible audio and DSP toolkit for Go. Whether you're building a synth, sequencer, effect, player, or a full-on singing voice synthesizer, Resona provides the tools you need, without getting in your way.

Resona comes from the Latin word "resono", meaning "to resonate".

✨ Features

• Super lightweight: no bloat, just clean Go code (with a few optional deps)
• Modular, Go stdlib-style API for audio/DSP
• Supports loading WAV, MP3, FLAC, and much more!
• Core DSP math: windows, filters, oscillators, etc...
• Basic audio effects: gain, filter, etc...
• Basic generators: noise, oscillator, etc...

🚀 Getting Started

Install using this command:

go get github.com/MatusOllah/resona

🧩 Conventions & Audio Data Model

Resona represents all audio data interleaved 32-bit float samples in the range [-1.0, 1.0]. For more details, see the documentation.

📚 Documentation

All documentation is available at pkg.go.dev/github.com/MatusOllah/resona.

⚖️ License

Copyright © 2025 Matúš Ollah

Licensed under the MIT License (see LICENSE) - free to use, fork, remix, and share!

Documentation

Overview

Package resona is the audio and DSP toolkit for Go.

Interleaved Audio Format

All sample slices in Resona represent interleaved, multi-channel audio data. In an interleaved layout, samples for each channel are stored in sequence for each frame. For example, in stereo (2-channel) audio, the slice:

[]float32{L0, R0, L1, R1, L2, R2, ...}

contains successive frames, where each frame consists of one sample per channel (left and right). This layout is common in many audio APIs and is efficient for streaming and hardware buffers. However, this may be less convenient for certain processing tasks. If channel-separated (planar) access is required, callers may convert interleaved slices using the audio package.

The number of channels is not specified by the aio package interfaces themselves; it is an implicit contract between the caller and the implementation. Implementations should clearly document the expected or provided number of channels and sample rate (e.g. using the afmt package).

Sample Format

All samples are represented as 32-bit floating-point numbers (float32). The value range is typically normalized between -1.0 and +1.0, where:

• 0.0 represents silence
• -1.0 to +1.0 represents full-scale audio signal
• Values outside this range may be clipped or distorted depending on the backend

Seeking

All seekable streams implement the io.Seeker interface. Seek offset is measured in frames.

Index

• func Version() (version, sum string)

Functions

func Version

func Version() (version, sum string)

Version returns the version of Resona and its checksum. The returned values are only valid in binaries built with module support.

If a replace directive exists in the Resona go.mod, the replace will be reported in the version in the following format:

"version=>[replace-path] [replace-version]"

and the replace sum will be returned in place of the original sum.