Using response time to evaluate noise fluctuations and Lombard speech in auralizations

This work presents developments on the usage of a viable measure of listening effort, that is the response time, in the framework of auralization. The approach is confirmed to be informative and suitable to evaluate the current means of acoustical renderings. In particular the manipulation of the noise signals and of the speakers’ spectrum is introduced in the auralization process. It is shown that, even in conditions where speech intelligibility is kept constant, the response time can disclose the listener effort according to the required processing load to achieve the given accuracy of reception. Applications to the acoustical design of rooms for speech communication are discussed.