In Pro Tools, voices are unique, discrete audio streams that can be routed to and from Pro Tools audio tracks, and physical audio outputs and inputs on your Pro Tools audio interfaces. The audio paths to and from Pro Tools software routed to Pro Tools hardware use voices. If you exceed the number of available voices in your system, you have effectively exceeded the available number of audio paths.
Typically, each audio channel for each track in your Pro Tools session uses a single voice. So, for a mono audio track, a single voice is used; for a stereo audio track, two voices are used. When using Punch Recording, two voices are needed for every single audio channel (one for playback and one for recording on punch in and out). In some situations, with Avid HDX and Pro Tools|HD systems, more than one voice may be necessary for a single channel (such as when using host processing on a DSP or TDM bus).
Voices can be allocated dynamically (with all Pro Tools systems, any available voices are automatically assigned as they are needed) or explicitly (with Pro Tools|HD systems only, you can explicitly assign audio channels to specific available voices to ensure that audio on those channels is always voiced).
On Avid HDX and Pro Tools|HD systems, the number of available voices is dependent on the amount of dedicated DSP processing in your system (see Playback, Recording, and Voice Limits with Pro Tools HD).
On all other Pro Tools systems, the full number of available voices for your system may be limited by the host processing power of your computer (see Pro Tools Capabilities with Different Hardware Configurations).
