Testing Environments

For the purposes of testing streaming audio, you can do a variety of things to emulate a production environment. In my own test lab I have set up several machines in the following manner to test a live audio stream:

On a desktop machine or Machine A, here named sequitur, I play audio selections of MP3 files using XMMS or mpg321. (Winamp is a viable alternative on Windows machines.) For the purposes of this setup, any system will do, as long as it can decode or play MP3 audio files. The output is channeled through a set of speakers on my desk or simply through the output port on the sound card itself. I then have a stereo audio patch cable with two male ends plugged into the headphone outlet on the speakers while the other end is connected to the Line-In port on the sound card of my test encoding box or Server B named gratis. This is also considered the master audio server for relaying purposes.

Another option is to hook up a small CD or MP3 player and connect the player's Line-Out feed to the Line-In port on the test server's sound card. The former method using speakers does not make for high quality sound, while the latter option with a small MP3 player works well for keeping the sound quality above normal. I currently use this latter method with a small player that plays both CDs and MP3 files.

Once an audio feed is encoding on the primary icecast server or Server B, set up additional servers to act as slave or relay machines. I configured Server C or quasi to pull audio data off gratis. (I discuss relaying later in this book.) From here I can use any Linux machine or Windows laptop (Machine B) to connect to either the master streaming box or the relay server and thus test the audio quality of the audio stream.

Later in this book, I discuss how to set up multiple streams on a slave relay server. This is useful for configuring multiple differing audio streams.