Are there noises that do not echo?

How do we locate noises in closed rooms?

There are echoes in every natural environment - noises not only reach the ears directly, but are also reflected by objects and reach the listener from all sides with a short time delay. The fact that we can still locate the source of a noise is due to the "echo suppression" - the processing of acoustic information in the brain suppresses the directional information of the reverberation. Scientists working with Benedikt Grothe from the Ludwig Maximilians University in Munich have found out which neuronal circuits are the basis for this mechanism in mammals.

We can definitely locate the echo of calls in the mountains, which only reverberates after a long time. In closed rooms, on the other hand, where the echo is only delayed by up to 20 milliseconds, the brain suppresses the directional information of the reverberation. Scientists working with Grothe examined neurons in the "Dorsal Nucleus of the Lateral Lemniscus" (DNLL), a region of the brain that is involved in locating sounds. Sounds coming from the right side are louder in the right ear than in the left. In the DNLL of the left hemisphere there are neurons that are stimulated by signals from the right ear and inhibited by signals from the left ear. So they only react when a noise comes from the right, it is said that they are "direction-sensitive".

The scientists have now shown that this inhibition by sounds from the left ear lasts up to 20 milliseconds longer than the sound is present. This is extremely long, normally the duration of the inhibition of a neuron corresponds to the duration of the signal that causes it to the nearest millisecond. The researchers were able to show that this long-term inhibition suppresses the reaction of the direction-sensitive cells to the echo - they become “deaf” for 20 milliseconds. The source of this long-term inhibition is in the opposite DNLL. The signal from the left ear therefore takes a detour via the right hemisphere in order to then suppress the echo again in the left DNLL. Of course, this process also applies in reverse for direction-sensitive neurons in the right DNLL.

The fact that direction-sensitive cells in the DNLL are deaf to the echo only half explains the echo suppression phenomenon. After all, we do perceive the echo, only the directional information of the echo is missing. In a computer model of the interconnection of other acoustic brain regions, Grothe and his colleagues have shown that the echo in higher brain regions definitely causes a neural reaction and that the long-term inhibition in the DNLL only reduces the directional information of this perception. In psychophysical experiments with human subjects, the scientists were able to confirm the predictions from their model.

The results are used, among other things, in robotics. In order for a robot to be able to react to commands such as “come here”, it must also be able to locate noises in closed rooms. Understanding the echo cancellation circuitry in humans can make an important contribution to the development of machine hearing.


Contact Personlink

Prof. Dr. Benedict Grothe

Inhibiting the inhibition: a neural network for sound localization in reverberant environments

Chair of Neurobiology, Ludwig Maximilians University, Munich
Großhaderner Str. 2
82152 Planegg-Martinsried
Phone: 089 / 2180-74302
Fax: 089 / 2180-74304

Pecka M, Zahn TP, Saunier-Rebori B, Siveke I, Felmy F, Wiegrebe L, Klug A, Pollak GD, Grothe B. J Neurosci. 27 (7): 1782-90.

Homepage