Indicate steps in registering a new bevy::audio::Decodable type

How can Bevy's documentation be improved?

Documentation for creating a new Decodable type is non-existent.

Here are the steps in creating a new Decodable type:

Create a type (we'll call it CustomAudio) that can be registered as an Asset and as Decodable.
Create a Decodable::Decoder iterator type that generate samples. This must be Send, as this type will be sent to the audio thread. This also must implement rodio::Source.
Register the Decodable type (CustomAudio) as an asset in the app.
Initialize the Audio<CustomAudio> resource in App.
Initialize the AudioOutput<CustomAudio> non-send resource.
Add the play_queued_audio_system::<CustomAudio> function to the PostUpdate stage.

Congratulations, you now have a working audio type that can be played by bevy!

working on this.

Squashed commit of the following: