Dj Jeroen Breebaart

Parametric coding of stereo audio

Parametric-stereo coding is a technique to efficiently code a stereo audio signal as a monaural signal plus a small amount of parametric overhead to describe the stereo image. The stereo properties are analyzed, encoded, and reinstated in a decoder according to spatial psychoacoustical principles. The monaural signal can be encoded using any (conventional) audio coder. Experiments show that the parameterized description of spatial properties enables a highly efficient, high-quality stereo audio representation.

Binaural processing model based on contralateral inhibition. I. Model structure

This article presents a quantitative binaural signal detection model which extends the monaural model described by Dau et al. [J. Acoust. Soc. Am. 99, 3615-3622 (1996)]. The model is divided into three stages. The first stage comprises peripheral preprocessing in the right and left monaural channels. The second stage is a binaural processor which produces a time-dependent internal representation of the binaurally presented stimuli. This stage is based on the Jeffress delay line extended with tapped attenuator lines. Through this extension, the internal representation codes both interaural time and intensity differences. In contrast to most present-day models, which are based on excitatory-excitatory interaction, the binaural interaction in the present model is based on contralateral inhibition of ipsilateral signals. The last stage, a central processor, extracts a decision variable that can be used to detect the presence of a signal in a detection task, but could also derive information about the position and the compactness of a sound source. In two accompanying articles, the model predictions are compared with data obtained with human observers in a great variety of experimental conditions.

Binaural processing model based on contralateral inhibition. III. Dependence on temporal parameters

This paper and two accompanying papers [Breebaart et al., J. Acoust. Soc. Am. 110, 1074-1088 (2001); 110, 1089-1104 (2001)] describe a computational model for the signal processing of the binaural auditory system. The model consists of several stages of monaural and binaural preprocessing combined with an optimal detector. Simulations of binaural masking experiments were performed as a function of temporal stimulus parameters and compared to psychophysical data adapted from literature. For this purpose, the model was used as an artificial observer in a three-interval, forced-choice procedure. All model parameters were kept constant for all simulations. Model predictions were obtained as a function of the interaural correlation of a masking noise and as a function of both masker and signal duration. Furthermore, maskers with a time-varying interaural correlation were used. Predictions were also obtained for stimuli with time-varying interaural time or intensity differences. Finally, binaural forward-masking conditions were simulated. The results show that the combination of a temporal integrator followed by an optimal detector in the time domain can account for all conditions that were tested, except for those using periodically varying interaural time differences (ITDs) and those measuring interaural correlation just-noticeable differences (jnds) as a function of bandwidth.

The influence of interaural stimulus uncertainty on binaural signal detection

This paper investigated the influence of stimulus uncertainty in binaural detection experiments and the predictions of several binaural models for such conditions. Masked thresholds of a 500-Hz sinusoid were measured in an NrhoSpi condition for both running and frozen-noise maskers using a three interval, forced-choice (3IFC) procedure. The nominal masker correlation varied between 0.64 and 1, and the bandwidth of the masker was either 10, 100, or 1,000 Hz. The running-noise thresholds were expected to be higher than the frozen-noise thresholds because of stimulus uncertainty in the running-noise conditions. For an interaural correlation close to +1, no difference between frozen-noise and running-noise thresholds was expected for all values of the masker bandwidth. These expectations were supported by the experimental data: for interaural correlations less than 1.0, substantial differences between frozen and running-noise conditions were observed for bandwidths of 10 and 100 Hz. Two additional conditions were tested to further investigate the influence of stimulus uncertainty. In the first condition a different masker sample was chosen on each trial, but the correlation of the masker was forced to a fixed value. In the second condition one of two independent frozen-noise maskers was randomly chosen on each trial. Results from these experiments emphasized the influence of stimulus uncertainty in binaural detection tasks: if the degree of uncertainty in binaural cues was reduced, thresholds decreased towards thresholds in the conditions without any stimulus uncertainty. In the analysis of the data, stimulus uncertainty was expressed in terms of three theories of binaural processing: the interaural correlation, the EC theory, and a model based on the processing of interaural intensity differences (IIDs) and interaural time differences (ITDs). This analysis revealed that none of the theories tested could quantitatively account for the observed thresholds. In addition, it was found that, in conditions with stimulus uncertainty, predictions based on correlation differ from those based on the EC theory.

Binaural signal detection with phase-shifted and time-delayed noise maskers

Detection thresholds were measured for interaurally in-phase sinusoids added to a narrow-band dichotic noise masker which was either interaurally phase shifted (NπSo condition) or time delayed (NτSo condition). The signals were spectrally centered in the noise bands and the delay τ equaled half the signal period. Both conditions were tested at 125 and 500 Hz for noise bandwidths of 10, 25, 50, and 100 Hz. In addition, NoSπ and NoSo thresholds were obtained. In contrast to expectations based on the EC theory, no differences in detection thresholds were observed between thresholds in phase-shifted and time-shifted maskers. The results also cannot be explained on the basis of more recent binaural models. The stimuli presented here might therefore serve as a useful validation tool in the development of new binaural theories and models.

Measuring the role of masker‐correlation uncertainty in binaural masking experiments

Masked thresholds of a 500‐Hz sinusoid were measured in an NρSπ condition for both running and frozen‐noise maskers using a 3IFC procedure. The nominal masker correlation varied between 0.64 and 1, and the bandwidth of the masker was either 10, 100, or 1000 Hz. The running‐noise thresholds were expected to be higher than the frozen‐noise thresholds because of interaural correlation uncertainty of the masker intervals for running‐noise conditions. Since this correlation uncertainty decreases with increasing masker bandwidth, differences between running‐ and frozen‐noise conditions should decrease with increasing bandwidth for interaural correlations smaller than +1. For an interaural correlation close to +1, no difference between frozen‐noise and running‐noise thresholds is expected for all values of the masker bandwidth. These expectations are supported by the experimental data. For the 10‐Hz running‐noise condition, the thresholds can be accounted for in terms of interaural correlation uncertainty. For t...

A new model for binaural signal detection

Most present‐day binaural models make use of coincidence detectors following an internal delay, so‐called crosscorrelation models. We present a new modeling approach based on a subtractive mechanism as found in excitatory–inhibitory type neurons in the auditory pathway. This approach transforms, for each auditory filter, time‐domain waveforms into a time‐varying two‐dimensional activity pattern, where each unit within that pattern has a characteristic interaural time delay and intensity difference. Limits of resolution are modeled by adding the same amount of internal noise to each unit. From this activity pattern, several features of the stimulus can be extracted, such as the apparent lateralization and the presence of a test signal within a masker. For modeling binaural detection it is assumed that a test signal must induce a significant change within the internal representation in order to be detectable. Depending on the condition, detection is limited by internal noise (NoSπ) or by external fluctuatio...