The word binaural (literally "two ears") describes the natural human way of perceiving sound and the solutions implemented to emulate this .

Binaural recording means capturing the information which naturally reaches the eardrums.
Binaural sound reproduction provides each ear with its specific part of the sound field, preserving inter aural cues.

With just two ears (i.e. two sensors), we are able to localize sound in a 3 dimensional space (above-below, front-rear and left-right) : binaural recording necessarily uses stereo techniques.
The ears and the torso, interacting with the surrounding space, help our brain estimate the location of a source thanks to difference cues such as time of arrival, intensity and frequency variations. These characteristics are known as the Head-Related Transfer Function (HRTF).

Binaural production is stereophonic and the specifications are those of natural listening. Playback calls for a strict right and left channel isolation and therefore requires the use of headphones.

Natural listening :

• We very seldom perceive the noise-floor of our auditory system
• The dynamic range is approximately 120dB
• The frequency response covers the standard range of 20Hz-20KHz