Kenku FM is a desktop application for Windows, MacOS and Linux designed to be the easiest way to share music in a Discord voice call.
- Use the Kenku Player interface to share your local music and sound effects to your discord calls.
- Use the built in web browser to share audio from your favourite websites like YouTube and Spotify.
- Trigger and control sounds with the Kenku FM plugin for the Elgato Stream Deck.
- If you already have a virtual audio cable setup use Kenku FM as an easy way to connect to Discord with your existing software.
Prebuilt binaries can be found at kenku.fm or from the GitHub releases.
Docs on using Kenku FM can be found here.
- Kenku FM is an Electron application primarily written in Typescript and React.
- Electron browser views are used to display external web content or the built in audio player app.
- The user creates and provides their own Discord bot token to connect to Discord.
- The Electron media capture API is used to capture audio from each browser view. The audio is then mixed using a Web Audio Context and sent to a Discord voice call using Eris.
- An optional HTTP server allows users to trigger and control the playback of the built in player app. An example of this in action can be seen with our Stream Deck plugin.
- Enable external inputs to allow mixing in OS audio inputs.
- Enable multiple outputs to send your audio to multiple Discord servers at once.
- If you plan to use Kenku FM for streaming you can also output to your local machine for capture by a streaming app and a discord call for your players at the same time.
Kenku FM uses Yarn as a package manager and Electron Forge as an Electron builder.
To install all the dependencies run:
yarn
To run Kenku FM in a development mode run:
yarn start
To make a production build run:
yarn make
If you wish to add protected media playback support follow the build steps from the Electron for Content Security repo.
The stream deck plugin can be found here.
As we act as a web browser in order to play protected media we need to support Google's Widevine Content Decryption Module (CDM). To do this Kenku FM uses the Electron for Content Security version of Electron provided by Castlabs.
This allows us to support loading of DRM protected media on Windows and MacOS.
Unfortunately the story isn't as simple on Linux. While we are able to provide support for protected media on x64 Linux builds, Linux doesn't support the Widevine Verified Media Path (VMP) this means that sites that require the use of VMP will not work. Unfortunately at this time Spotify requires VMP. You can read more about VMP and Linux here.
Next Google doesn't provide a publicly available ARM Linux version of their Widevine Content Delivery Module so we are unable to provide any protected media support on ARM Linux.
If any of these change we'll be happy to update Kenku FM with full support for protected media on Linux.
All source files can be found in the src folder, our build scripts for CI/CD are in the publish folder.
Within the src folder the index.ts file and main folder contains the code for the main process of Electron. This includes things like managing the Discord connection, creating the HTTP server for the remote control and managing the browser views.
The renderer.ts file and renderer folder contains code for the renderer process of Electron. The renderer is written in React and uses Redux Toolkit for state management.
The preload.ts file and preload folder contains code for the preload script for the renderer. The preload script mainly acts as a bridge to expose functionality from the main process to the renderer.
The player folder contains the code for the built in audio player. This runs as a separate web view and is loaded as a separate entry point in the forge.config.js. The player app is written in React and also uses Redux Toolkit for state management.
Kenku FM is licensed under the GNU General Public Licence v3.0.
For our own wellbeing Kenku FM follows a similar contribution policy projects like Litestream.
This means we are open to pull requests for bug fixes only. Pull requests for new features will not be accepted due to the burden of maintaining these features into the future.