When a speaker is talking in a room, the listener perceives the direct sound and the sound that is has reverberated off all the walls. In fact, what the listener hears is a signal characterised by its initial sound level, in addition to the same signal generated by the sound being reflected off the walls.

This second signal is almost the same as the direct sound, but involves a time delay.

All these reflections can reinforce the sound level received or, to the contrary, can give rise to a loss of information when the "tailing off" of previous syllables masks the direct sound.

Characterisation of the transmission channel between the speaker and listener is directly linked to the study of acoustic characteristics. Two of them are essential for understanding spoken messages :

• The "signal-to-noise" ratio : This specifies the extent to which the spoken words stand out from the noise. This noise may be brought about by sources within or outside the room. The higher the signal-to-noise ratio, the greater the intelligibility of the spoken words.
  Determination of this ratio is directly linked to the sound's decay within the space.

• The signal's time "deformation" : This characteristic is set by the reverberation time. If it is sufficiently long, the reverberation time, which indicates the ability of the room to maintain the sound after the emitting source has stopped, causes a partial covering of the words.
  The longer the reverberation time, the greater this acoustic superimposition is. It leads to a decline in the quality of the spoken words heard.

The reverberation time is determined by knowledge of the sound decay over time.

Thus, for any room, controlling spatial and temporal decays is vital. These values depend on the shape and dimensional characteristics of the rooms, as well as on the acoustic absorption coefficient of the various walls, and that of the furniture.