Open
Access
This research focuses on sound localization experiments in which subjects report the position of an active sound source by turning toward it. A statistical framework for the analysis of the data is presented together with a case study from a large-scale listening experiment. The statistical framework is based on a model that is robust to the presence of front/back confusions and random errors. Closed-form natural estimators are derived, and one-sample and two-sample statistical tests are described. The framework is used to analyze the data of an auralized experiment undertaken by nearly nine hundred subjects. The objective was to explore localization performance in the horizontal plane in an informal setting and with little training, which are conditions that are similar to those typically encountered in consumer applications of binaural audio. Results show that responses had a rightward bias and that speech was harder to localize than percussion sounds, which are results consistent with the literature. Results also show that it was harder to localize sound in a simulated room with a high ceiling despite having a higher direct-to-reverberant ratio than other simulated rooms.
Download: PDF (HIGH Res) (1.8MB)
Download: PDF (LOW Res) (750KB)
When simulating a virtual reality, surfaces near the subject location can play a perceptual role. The authors studied the acoustic effect of nearby objects with both binaural response measurements and subjective listening tests. With a large, smooth, and flat surface, significant acoustical effects exist when the distance is less than about 50 cm. The largest effects can be attributed to comb filtering caused by the interaction between direct sound and reflected sound. At very short distances, some frequency-dependent resonances and shadowing effects that were evoked by the sound field between the subject and the surface could also be observed. Measurements showed some effects that could not be attributed to either simple comb filtering or resonance effects. In a listening test the subjects were asked to sort three binaurally-rendered samples in growing order of perceived distance. With relatively small distance triples, i.e., 1-11-21 cm, 3-13-23 cm, or 7-17-27 cm, the subjects performed better than guessing. However, even with shortest tested distances the proportion of correct answers was of the order of 50%. Two of 12 subjects consistently reported the distances in reverse order, which shows that the acoustic effect was significant, but it did not lead to the correct perception of distance of the surface. The results suggest that the acoustic effect of objects being close to an avatar’s ear may in some cases improve realism and sound quality in an acoustic virtual reality both in anechoic and in reverberant conditions.
Download: PDF (HIGH Res) (2.8MB)
Download: PDF (LOW Res) (258KB)
This report presents a methodology to measure transducers using a dual-added-mass technique in order to extract the motional impedance ZM(w) and force factor Bl from the total impedance. The method is more accurate than the classical single-added-mass approach insofar as the blocked electrical impedance can be completely filtered out of the total impedance. The methodology is suitable for determining moving mass and compliance in motional impedance models that include viscoelasticity and frequency-dependent damping. It is applicable to transducers for which adding mass to the moving parts is possible without introducing significant artifacts. A mass-consistency test is described to provide an internal validity check for the impedance fitting process. This test is quite stringent and provides an objective measure of the quality of the fit to the measured impedance. The quality of the proposed measurement technique was verified with an ANOVA Gage R&R measurement system analysis.
Download: PDF (HIGH Res) (7.5MB)
Download: PDF (LOW Res) (1.0MB)
The most efficient acoustic modeling reverberators use a two-stage model, producing detailed early reflections with generic late reflections. Most methods of this type do not accurately model energy flux in the late reverb module, hence its Clarity Index (C80) is inaccurate. C80 is a unit of measurement that quantifies the ratio of early to late reverb energy on a log scale. The authors propose an efficient method to model late reverb energy flux showing that it models C80 and related metrics, Definition (D50) and Centre Time (TS), with more than twice the accuracy of the baseline method. It was observed that the standard deviation for C80 was more than eight decibels across various listener positions in large rooms, indicating that it varies audibly with respect to location. The proposed method for computing energy modeling coefficients can be applied to existing hybrid acoustic modeling reverberators. This approach allows the listener and sound source to change locations in real time without using precomputed impulse responses.
Download: PDF (HIGH Res) (1.5MB)
Download: PDF (LOW Res) (302KB)
Increasing sophistication is evident in the methods that are used to construct and control the elements of array or cluster loudspeakers, and this is critical to enabling a more “designed” sound field, with control over the coverage of a large arena. Compared with the rather distant past when roadies often just put up a huge pile of loudspeakers and hoped for the best, today’s open-air live sound systems are calculated and designed to a much higher degree. Integral to that design is the idea that one might be able to control where low-frequency sound is directed, in order to avoid polluting the surrounding area too much.
Download: PDF (396KB)
Page: 1059
