Authors:Tylka, Joseph G.; Choueiri, Edgar Y.
Affiliation:Princeton University, Princeton, New Jersey
Virtual navigation within a three-dimensional ambisonics-encoded sound field (sound field that has been decomposed into spherical harmonics) enables a listener to explore with 6 degrees of freedom an acoustic space. This allows for experiencing a spatially- and tonally-accurate perception of the sound field. The authors propose and characterize through numerical simulations an interpolation-based method for virtual navigation, wherein a subset of microphones is parametrically determined to ensure that the region of validity restriction is not violated. An existing alternative method, in which navigation is performed by computing a weighted average of the higher-order ambisonics (HOA) signals from each microphone, was shown to incur spectral distortions due to comb-filtering and localization errors The proposed method employs knowledge of the locations of any near-field sources in order to determine which HOA microphones are valid for use in the navigation calculation as a function of the desired listening position. Additionally, at low frequencies, the proposed method applies a matrix of regularized least-squares inverse filters to estimate the ambisonics signals at the listening position, while at high frequencies, the weighted average method is employed. The numerical simulations were validated against experimental measurements, which showed that the observed discrepancies, and therefore the fidelity of the simulations, do not depend significantly on the navigational method, microphone spacing, or source position.
Download: PDF (HIGH Res) (1.4MB)
Download: PDF (LOW Res) (1.1MB)
Authors:Tylka, Joseph G.; Choueiri, Edgar Y.
Affiliation:3D Audio and Applied Acoustics Laboratory, Princeton University, Princeton, New Jersey 08544, USA
Performance errors are characterized for two representative linear extrapolation methods for virtual navigation of higher-order ambisonics sound fields. For such methods, navigation is theoretically restricted to within the so-called region of validity, which extends spherically from the recording ambisonics microphone to its nearest source, but the precise consequences of violating that restriction has not been previously established. To that end, the errors introduced by each method are objectively evaluated in terms of metrics for sound level, spectral coloration, source localization, and diffuseness, through numerical simulations over a range of valid and invalid conditions. Under valid conditions, results show that the first method, based on translating along plane-waves, accurately reproduces both the level and localization of a source, whereas the second method, based on ambisonics translation coefficients, incurs significant errors in both level and spectral content that increase steadily with translation distance. Under invalid conditions, two common features of the performance of both methods are identified: significant localization errors are introduced and the reproduced level is too low. It is argued that these penalties are inherent to all methods that are bound by the region of validity restriction, including all linear extrapolation methods.
Download: PDF (HIGH Res) (1.9MB)
Download: PDF (LOW Res) (1.6MB)
Authors:Pereira, Igor; Distante, Cosimo; Silveira, Luiz F.; Gonçalves, Luiz
Affiliation:Institute of Applied Sciences and Intelligent Systems, Lecce, Italy; Federal University of Rio Grande do Norte, Natal, Brazil
This research proposes an approach for computing the time offsets between audio sequences that contain musical sounds from different instruments produced in a distributed way and which have a set of weak features that are not useful as alignment points. It is therefore necessary to apply transformations in order to find a set of distinctive features to compute the offset values in a suitable way. The main issue that occurs with such a system is nonlinearity that does not allow the delay to be predicted by using a linear function. To solve this problem, the authors propose a set of long short-term memory (LSTM) layers to create a neural network model capable of learning such features transformations in a supervised approach, using a gradient-descent optimizer. This demonstrates the use of a recurrence matrix to extract timing information from a set of transformed features given by the neural network output. With this approach, the algorithm can classify up to 60% of a specific combination from the MedleyDB data set, and reduce the search space to five possibilities with accuracy up to 90% while keeping the precision of 10 ms. This performance is equal or better than state-of-the-art methods.
Download: PDF (HIGH Res) (1.9MB)
Download: PDF (LOW Res) (246KB)
Authors:Dong, Hao; Shen, Yong
Affiliation:Key Laboratory of Modern Acoustics ( MOE ), Institute of Acoustics, Nanjing University, Nanjing, 210093, China
Band-pass loudspeaker systems are widely used as subwoofers to reproduce sound in the bass range. This paper presents design procedures and theoretical analysis of an active eighth-order band-pass loudspeaker system that consists of an active low-pass filter and a first- or second-type sixth-order band-pass loudspeaker system. Based on the reactance transformation method, the overall system response can be shaped into a symmetric filter function with a Butterworth or Chebyshev alignment. Four classes of alignments are derived when aligning the two types of systems, which results in a total of eight groups of system implementations. Although the derivation of eighth-order alignments is relatively complex, analytical expressions for required loudspeaker system parameters are successfully obtained, facilitating an easily executed design. Based on the alignment parameters, a thorough analysis of system small- to large-signal performance, including sensitivity, efficiency, displacement, transient behavior, and acoustic power rating, is presented, enabling comparisons of the different system implementations. Therefore, designers can choose an optimal alignment class and system type to satisfy various design requirements.
Download: PDF (HIGH Res) (1.5MB)
Download: PDF (LOW Res) (1.3MB)
Authors:Novak, Antonin; Lihoreau, Bertrand; Brasseur, Emmanuel; Lotton, Pierrick; Simon, Laurent
Affiliation:Laboratoire d'Acoustique de l'Université du Mans, LAUM-UMR 6613 CNRS, Le Mans Université Avenue Olivier Messiaen, CEDEX 9, France
This paper addresses the question of why some guitar pickups distort more than others. The electromagnetic pickup of an electric guitar is a nonlinear device that can provide a pleasant distortion. Although the pickup is a simple device consisting of a coil and a few magnets or pole pieces, the measurement of its nonlinear function is a difficult task. This research shows a measurement technique that can estimate the nonlinear function of a pickup in both y and z directions of the vibrating string. The experimental results are provided for three different types of pickups: a single-coil pickup with six staggered pole-pieces, a humbucker pickup with six equal-height pole-pieces, and a humbucker rail pickup. The measured nonlinear functions of the three pick-ups are very different from each other, leading to different distortions. These experimental results confirm that the pickup geometry plays an important role in distortion.
Download: PDF (HIGH Res) (3.1MB)
Download: PDF (LOW Res) (1016KB)
Authors:Allan, Jon; Berg, Jan
Affiliation:Luleå University of Technology, Luleå Sweden
Different ballistic definitions for the momentary time scale used in live loudness measurement were evaluated. Definitions from the ITU and EBU were compared as well as a faster version of the ITU version, two asymmetric time scales and the deprecated ballistics, defined in EBU Tech 3205-E, for peak program meters. The goal was to identify the ballistics definition that would function as the best complementary tool to a short-term time scale. Engineers within the broadcast industry and students in audio technology performed an audio alignment task in a simulated live broadcast environment using one ballistics definition per trial. Fader movements and output levels were recorded. After each trial, a set of assessment scales were rated by the subjects. Some results were: a decay time constant of 250 ms yielded better representation of the low-level parts of the dynamics in the signal compared to a 400-ms time constant; the present ITU version of the momentary time scale yielded an estimated less eye fatigue; effects on the resulting output levels, related the gate in ITU-R BS.1770 in conjunction with live compensation of unadjusted audio material were shown.
Download: PDF (HIGH Res) (1.6MB)
Download: PDF (LOW Res) (970KB)
There is a thriving network of enthusiastic synthesizer and effects builders, using a combination of analog modules and embedded computing devices, as well as some digital signal processing, working to create just about any electronic music system that you can think of. Low-cost off-the-shelf interfaces and embedded computing devices enable people to build interactive instruments and novel controllers with relative ease, without needing to understand everything about microelectronics. Two workshop sessions on electronic instrument design and applications from the recent New York convention are mined for the latest developments.
Download: PDF (220KB)