Dynamics

Audio... these are the voyages of the radio station URY, on it's brave adventure to turn literally EVERYTHING into a sosig.

Audio dynamics is simply the difference in loudness between sampleable elements of a continuous piece of audio. Surely this is simple? Audio (being the vibrations of air molecules or an electronic representation of this) can be loud and then audio can be quiet, and considerable variations of this. Whilst it is all very simple, the reality is human ears pick up on these differences very easily and even small variations to how you handle audio signals can greatly affect how the audio is perceived.

For most audio dynamics processing, the changes made by a certain piece of audio equipment or software to the 'pure' input signal can be represented as a simple 2-axis graph, mapping a sample of an audio signal between 1 and -1 to an equivalent graph with a practically equivalent range like so:

These graphs are not perfect, as processing can be context sensitive.- i.e have a different response to a signal if it is part of a representation of a higher frequency -. However for most basic applications they are a reasonable compromise to keep things simple.