Aleksandar Pantić’s scientific contributions

Noise represents one of the most significant disturbances in measured room impulse responses (RIRs), and it has a potentially large impact on evaluation of the decay parameters. In order to reduce noise effects, various methods have been applied, including truncation of an RIR. In this paper, a procedure for the response truncation based on a model of RIR (nonlinear decay model) is presented. The model is represented by an exponential decay plus stationary noise. Unknown parameters of the model are calculated by an optimization that minimizes the difference between the curve generated by the model and the target one of the response to be truncated. Different curves can be applied in the optimization—absolute value of the RIR, logarithmic decay curve, and Schroeder curve obtained by the backward integration of the RIR. The proposed procedure is tested on various synthesized and measured impulse responses. It is compared with the procedure taken from the literature, often applied in practice.

Swept sine technique is one of the most used techniques for measurement of the impulse response of acoustical systems including rooms. During its application, some common disturbances, such as noise and distortion, could reduce the quality of the measurement results. Two different noise categories can be distinguished - background noise always present in room acoustic measurements and transient noise that appears suddenly and has a short duration. Distortion typically occurs in the transducer. The effects of these disturbances (especially transient noise and distortion) are investigated here with an emphasis on reduction of negative effects of transient noise. For that purpose, several procedures for noise effects reduction are proposed and analysed. The results show that noise and distortion cause certain artefacts disturbing the measured response. Some of these artefacts can be reduced by applying adequate procedures.

In the room impulse response measurements, one of the common disturbances affecting the measured results is nonlinearity. When maximum length sequences are applied for the measurements, the distortion products are distributed along the estimated impulse response. While this property of maximum length sequences has been widely used in the auditory system applications for characterization of a nonlinear system, it has not been exploited in room acoustics. Due to this reason, the possibility of separating the distortion products in the measurements of the room impulse responses with maximum length sequences is analyzed in this paper. The results show that this measurement method is sensitive to nonlinearities, but the separability of the distortion products could reduce the nonlinearity effects.

The absorption of sound in air represents one of the main problems of the scale model measurements. This absorption, especially at higher frequencies, is considerably greater than the value determined by the law of acoustical similarity between the full scale and the scale model. Different alternatives are applied for compensation of the excess air absorption including a numerical compensation. In this paper, a modified approach to numerical compensation is proposed. It is based on compensation of the sound decay only, and not background noise. As a consequence, there is no an increase of background noise in the compensated impulse response. The results obtained by the proposed procedure are compared to the corresponding ones obtained by the other procedures.

Determination of some important room acoustic quantities is based on backward (Schroeder) integration of a room impulse response. In order to increase the dynamic range of the obtained reverberation decay curve, certain common methods have been adopted. One of the most widely used methods is related to truncation of a room impulse response before the integration. It has been shown that the optimal truncation point is located at the intersection of the main decay slope and the noise floor level. Various procedures have been proposed for estimation of this truncation point. The response truncation based on the cumulative distribution function is proposed and analyzed in this paper. The parameters of the proposed procedure are optimized in order to obtain minimal deviation of the determined truncation point from the correct value. The proposed procedure is compared to the corresponding one already published in the literature.

One of common disturbances in the room impulse response measurements is nonlinearity. When maximum length sequence (MLS) method is used, the distortion products are distributed along the impulse response. This property has been widely used in the auditory system applications for characterization of nonlinear systems. However, it has not been exploited in room acoustics. Due to that, possibility of extracting the distortion products in the room impulse response measurements by MLS method is analyzed here.

Swept sine technique is one of the most used techniques for measurement of the impulse responses of the acoustical systems including rooms. It uses a swept sine as an excitation signal. During its application, some disturbance, such as noise, could reduce the quality of the measurement results. Two different categories of noise can be distinguished - background and transient noise. In this paper, the attention is focused to the transient noise, its influence on the room impulse response measurement, and reducing its negative effects. For that purpose, several procedures for noise effects reduction are analysed.