|
PAPERS |
|
Errors in Real-Time Room Acoustics Dereverberation
(PDF-806K) |
|
|
Panagiotis D. Hatziantoniou and John N. Mourjopoulos |
883 |
|
While researchers have been studying methods for removing the effects of
room acoustics for over 20 years, results are often inconsistent between
computer simulations and real applications. This discrepancy has not
been formally studied. Errors in measurements of real spaces are
significantly higher than expected, even when they are carefully
performed. Additionally, when combined with long filters, perceptual
artifacts are often worse than the advantages of such systems. High-Q
poles used in such filters are very sensitive to listener location and
to time-varying acoustic noise present inside the room. By using complex
smoothing, which implies a reduction in the degree of room compensation,
unpleasant artifacts are greatly reduced. |
|
|
Comment on this paper
|
|
| |
|
Development and Validation of a Method for Predicting the Perceived Naturalness of Sounds Subjected to Spectral Distortion
(PDF-444K) |
|
|
Brian C. J. Moore and Chin-Tuan Tan |
900 |
|
Having partitioned distortion into linear and nonlinear distortion and
having explored the degree to which such distortions degrade the
perception of naturalness, the authors now propose a model to explain
the perceptual effects of linear distortion, namely, irregularities in
the frequency response. Using the data from 168 filtering conditions, a
perceptual model based on excitation patterns has been developed to
predict subjective judgments. The model is based on the difference
between the excitation patterns for undistorted and distorted pink
noise. Two validation experiments supported the model. |
|
|
Comment on this paper
|
|
| |
|
Interpositional Transfer Function for 3D-Sound Generation
(PDF-1.1MB) |
|
|
F. P. Freeland, L. W. P. Biscainho, and P. S. R. Diniz |
915 |
|
In order for head-related transfer functions to gain widespread use in
binaural audio applications, there needs to be efficient ways for
interpolating among a limited number of reference cases that are
quantized in elevation and azimuth without loss of perceived accuracy. A
triangular interpolation combined with an auxiliary function, called
interpositional transfer function, save more than 40% of the operations
required by the traditional bilinear methods. This method was employed
in a system that generated moving locations. |
|
|
Comment on this paper
|
|
| |
|
Measurement and Application of Equivalent Input Distortion
(PDF-1.2MB) |
|
|
Wolfgang Klippel |
931 |
|
Because a loudspeaker can be considered as a single-input,
multiple-output transducer, distortion should ideally be measured at
multiple spatial locations. This produces a massive collection of data,
much of which is redundant. Rather, the author shows that the dominant
nonlinearities add distortion to the input signal, which is transferred
by a linear system to each point in space. This model leads to a new
measurement technique that gives deeper insight into the distortion
mechanisms and the influence of the acoustical environment, noise, and
parasitic vibrations. |
|
|
Comment on this paper
|
|
| |
|
STANDARDS AND INFORMATION DOCUMENTS |
|
AES Standards Committee News
(PDF-114K) |
948 |
|
Digital audio measurement; storage and handling; digital library and
archive systems; acoustics-sound source modeling; loudspeaker
measurement; IEEE 1394; audio metadata
|
|
| |
|
FEATURES |
|
25th Conference Report, London
(PDF-2.7MB) |
952 |
|
Bass Handling in Spatial Reproduction
(PDF-481K) |
962 |
|
118th Convention, Barcelona, Call for Papers
(PDF-27K) |
991 |
| |
|
DEPARTMENTS |
|
News of the Sections
(PDF-365K) |
968 |
|
Sound Track
(PDF-61K) |
977 |
|
Upcoming Meetings
(PDF-41K) |
978 |
|
New Products and Developments
(PDF-37K) |
979 |
|
Available Literature
(PDF-32K) |
980 |
|
Membership Information
(PDF-400K) |
981 |
|
Advertiser Internet Directory
(PDF-63K) |
982 |
|
Sections Contacts Directory
(PDF-69K) |
993 |
|
AES Conventions and Conferences
(PDF-935K) |
1000 |
| |
|
EXTRAS |
|
Cover & Sustaining Members List
(PDF-280K) |
|
|
VIP List & Editorial Staff
(PDF-55K) |
|