Abstract:

Speech is a method of communication between two or more minds. The actual forming and understanding of the message are in the physiological domain. The transmission of the speech by sound waves is in the physical domain. Only the latter can be controlled by means of acoustical engineering. The relevant processes in the physiological domain can be contained in the term ``proficiency of speaker and listener.'' The relevant processes in the physical domain can be contained in the term ``information loss.'' This paper will deal with: (a) the underlying factors of information loss and assessing them by measurement and calculation as accurately as possible; (b) the means and limitations of quantifying the information loss: AL[inf cons], STI, etc.; (c) for the calculation of the information loss, an adaption of the ``classic'' Peutz equation, suitable for data generated by contemporary computer programs, will be given; (d) the accuracy of the data, with some examples like comparison of results from word lists, STI measurement and AL[inf cons] calculation; (e) results in extreme situations like a large, closed sports stadium with a 9-reverberation time and a tunnel with a length of 1 k; (f) design considerations for sound systems regarding the proficiency of speaker and listener. [See NOISE-CON Proceedings for full paper.]