Cocktail party effect

The cocktail party effect, even intelligent or selective hearing, refers to the ability of the human auditory sense to extract the sound components of a particular sound source from the mixture of the noise in the presence of multiple sound sources. For example, the hearing is at a cocktail party, where many people speak at the same time to act on the situation, only the words of a speaker and suppress that of others.

Description

The hearing in this case reaches a noise suppression 9-15 dB, that is, the sound source to which a person concentrates, is held two to three times louder than the ( disturbing ) ambient noise. This means that in a störgeräuschbehafteten environment a person is able to pick out the sound of an acoustic source ( hörmäßig ). A unit placed in such an environment microphone seems, however, mainly reflect the noise. Furthermore, the cocktail party effect means that the sound sources are perceived in rooms from the ear with little spatial impression: the signals listen to " dry" and hardly to die away. On the other hand -picked in such surroundings microphone are strongly spatially - sounding signals again.

In addition, the ear without movement of the head can focus on any sound source and scored for lateral acoustic sources similar good results.

The signal processing strategies of human hearing are not yet fully deciphered. All technical attempts to extract with only two acoustic sensors from a sound mixture, only the signals of a source, achieve substantially worse results. In the sound recording equipment you are with direction-specific recordings remain dependent on microphone arrays or microphone arrays with a larger number of microphones and shotgun microphones (see also localization ).

Influences

The cocktail party effect is a binaural effect, it only occurs with beidohrigem hearing. People who have only one functioning ear or rely on a hearing aid, are therefore significantly more impaired than noise from people who hear equally with both ears.

The cocktail party effect is closely associated with the ability to localize sound sources. If the hearing established the direction of a sound source, it is also able to distinguish the sound components of this direction of parts from other directions.

It is assumed that the auditory system performs a kind of cross-correlation between the sound components of both ears. Cross-correlation functions formed from the signal components to an axis which describes the time difference between the two signals. If a sound from the side with an interaural time (English Interaural Time Difference, ITD), for example, ITD = 0.3 ms, one finds this signal to the 0.3 -ms point of the correlation axis again. In the presence of multiple sound sources, complex correlation patterns in which statistical parameters such as mean and variance of the directions and signal components of the sound sources involved depend arise. The hearing is obviously in a position to evaluate these patterns and to determine from the amount of signal from a desired direction. In addition to correlation method for interaural time differences also exist for such interaural level differences (English Interaural Level Difference, ILD).

In addition to the direction-specific signal processing, the listener, also other effects for noise reduction, such as knowledge of the signal characteristics; for example, can be in language non-verbal sounds and noises ignore as irrelevant as expectation of certain sounds with certain mouth movements.

History

The term was also used in conversation analysis, as researchers began to enter into discussions and discovered the inadequacy of monophonic microphones.