Audio Encoder for ICECAST server

Simple cross-platform, command-line based audio encoder for icecast streams.

This project is sponsored by PLAY.CZ

supported OS:

• MacOS (intel / apple silicone)
• Linux (x86 / arm)
• Windows

Configuration

Create config.json in same directory as executable

{
  "ffmpegBinaryPath": "ffmpeg.exe",
  "ffmpegCaptureMode": "dshow",
  "encoders": [
    {
      "id": "test-128mp3",
      "bitrate": 128,
      "chanels": 2,
      "samplerate": 44100,
      "format": "mp3",
      "captureAudioCard": ":0",
      "icecast": {
        "mount": "aaa.mp3",
        "username": "source",
        "password": "password",
        "server": "icecast.play.cz:8000"
      }
    },
  ]
}

property	type	description
ffmpegBinaryPath	string	Path to ffmpeg binary you want to use
ffmpegCaptureMode	dshow, avfoundation, alsa	dshow - Windows avfoundation - MacOS alsa - Linux
encoders	Object[]	Array of individual encoders config

Encoder settings

property	type	description	example
id	string	Inique ID	test-stream
bitrate	int	Bitrate in kb	128
channels	int	Number of audio channels (1 = mono / 2 = stereo )	2
samplerate	int	Audio sample rate in Hz	44100
format	mp3, aac, aac+	Audio format. Please note, that aac+ requires custom ffmpeg build with libfdk_aac build into it.	mp3
captureAudioCard	string	Audio cart ID for ffmpeg * see later in this docs	":0" for MacOS "audio=\"Your Audio Device Name\"" for Windows "default" for Linux
icecast->mount	string	Icecast mountpoint name	test.mp3
icecast->username	string	Username for this mountpoint (if you do not know, the source is default)	source
icecast->password	string	Password for this mountpoint	password
icecast->server	string	Icecast server with optional port	icecast-server.example.com:8080

captureAudioCard & ffmpegCaptureMode

Identifing soundcard input is not so easy, but if you follow this guide, it is simple to understand.

First, you need to use correct capture mode for OS:

• dshow - Windows
• avfoundation - MacOS
• alsa - Linux

Then, you need to list your available capture devices using ffmpeg to find ID or name (depending on platform)

Windows

ffmpeg.exe -f dshow -list_devices true -i dummy

MacOS

ffmpeg -f avfoundation -list_devices true -i dummy

Linux

ffmpeg -f alsa -list_devices true -i dummy

Example output:

AVFoundation video devices:
[0] FaceTime HD Camera
[1] LadaSoukup Camera
[2] LadaSoukup Desk View Camera
[3] Capture screen 0
[4] Capture screen 1
AVFoundation audio devices:
[0] MacBook Pro Microphone
[1] LadaSoukup Microphone

Last step is to configure the audio input

Windows On Windows, you have to use the device name and set it using audio="" syntax.

"audio=\"Your Audio Device Name\""

MacOS On MacOS, you have to use system index and notation video:audio. We need only audio, so we will skip the video.

":0"

Linux

"default"