2004 March, Volume 52 Number 3

CONTENT

INTRODUCTION

High-Resolution Audio
Rhonda J. Wilson   116

PAPERS

Coding for High-Resolution Audio Systems
J. Robert Stuart   117
To achieve the highest audio quality, one must consider not only each component of the chain but also the entire chain as a system. Furthermore, quality acquires a meaning that depends on the goals, application, and cost tradeoff. Choices for each element-such as sampling rate, encoding format, word size, filtering, and noise floors-may or may not influence the auditory quality of the total system. A review of various channel-coding methods within the context of auditory perception illustrates the principle of a "coding space," which is the amount of useful information preserved or destroyed. For example, in an archival application the encoding noise floor should be at least two bits lower than the self-noise of the best audio signal.

Audio Analog-to-Digital Converters
Mike Story   117
By considering the requirements for audio conversion within the wider context of conversion applications, we can better appreciate the implication of choosing a particular approach. Choices include: one bit versus multibit, use of feedback or multistage sequential processing, and switched capacitor versus continuous time. A review of the resulting performance shows the delicate balance among such parameters as accuracy, frequency, and burdens on implementation.

Future Design Challenges for Audio Converter Products
Julian Hayes, John Pennock, and Anthony Magrath   159
Extensive discussion about the theoretical limits of various digital conversion techniques, while interesting,ignore the practical implications of implementation difficulties. Physical devices, under economic pressure from the requirement of high yield and low expense, degrade performance to a significant degree. Chip designers must contend with timing errors, substrate noise, internal crosstalk, mismatched components, temperature gradients, nonlinear slew rates, wiring inductance, nonuniform doping, impurities, and numerous other issues. Special processes and computer modeling help reduce these corrupting influences, but problems still remain when attempting to achieve performance to the theoretical limits.

One-Bit Audio: An Overview
Derk Reefman and Erwin Janssen   166
While the CD format solidified full digital words of 16 to 24 bit as an audio coding standard, digital conversion technology moved toward oversampling using a few bits or only one bit. By using noise shaping within a sigma-delta modulator, one-bit conversion can produce very high-quality audio. A review of the various implications leads to the conclusion that one-bit coding is an attractive approach rather than converting to the traditional pulse-code modulation.

Lossless Compression of One-Bit Audio
Eric Knapen, Derk Reefman, Erwin Janssen, and Fons Bruekers   190
Because the goal of the Super Audio CD is to produce the highest quality, lossless compression offers a way to reduce the storage capacity without sacrificing quality. A proposed design produces compression by using a prediction filter operating on a one-bit audio stream and then uses a probability lookup table to achieve a compression rate greater than 2.5:1 on typical music samples. However, the benefits produce a small uncertainty in playing time. Instantaneous compression rate varies dramatically over the extremes of peak transients or noise and intervals of silence.

Pulse-Code Modulation-An Overview
Stanley P. Lipshitz and John Vanderkooy   200
The authors provide an overview of pulse-code modulation. They graphically demonstrate the properties of sampling and reconstruction, establishing that PCM allows band-limited signals to be time accurate to infinite precision. Any bandwidth can be accommodated with proper choice of the sampling frequency. The correct use of dither is described. It renders a multibit quantization distortionless and perfect in the sense that it adds only a benign signal-independent noise. Any SNR can be accommodated with appropriate choice of wordlength. The use of noise-shaping allows a tradeoff between in-band and out-of-band noise levels, and between wordlength and sample rate. The authors conclude that PCM forms the logical way for a digital audio system to best encompass high-resolution audio.

Antialias Filters and System Transient Response at High Sample Rates
Peter G. Craven   216
With the use of very high sampling rates, a designer has additional options for balancing the conflicting requirements in both the time and frequency domains. Lower sampling rates require brick-wall filters, which produce time smear. By using a class of gentle frequency filters, called apodizing, pre- and postringing can be reduced or removed. It is argued that these temporal artifacts justify the use of higher sampling rates. While there is no attempt to prove which combination of parameters is perceptually optimum, there are clearly a wide range of choices and consequences.

The MLP Lossless Compression System for PCM Audio
M. A. Gerzon, P. G. Craven, J. R. Stuart, M. J. Law, R. J. Wilson   243
As an alternative to the standard types of compression, a lossless architecture does not need to consider perceptual issues because the recovered audio is identical to the original. However in exchange for this property, the compression rate depends on the signal details at any given moment. By using a novel four-level approach that incorporates matrices, a high degree of compression is readily obtained. Error checking and repair makes such an approach very robust.

LETTERS TO THE EDITOR

Comments on "Differences in Performance and Preference of Trained versus Untrained Listeners in Loudspeaker Tests: A Case Study"
Kenneth Gundry   261

Author's Reply
Sean E. Olive   261

CORRECTIONS

Correction to "Comments on 'Analysis of Traditional and Reverberation-Reducing Methods of Room Equalization'"
John N. Mourjopoulos   262

STANDARDS AND INFORMATION DOCUMENTS

AES Standards Committee News   263
Audio metadata for libraries and archives; loudspeaker measurements; shielding and EMI

FEATURES

116th Convention Preview, Berlin   266
    Exhibitors   268
    Exhibit Previews   272
Audio Gets Smart: A Workshop on Semantic Audio Analysis   288
Audio for Games: Let the Games Continue   292
117th Convention, San Francisco, Call for Papers   319
26th Conference, Baarn, Call for Papers   320

DEPARTMENTS

News of the Sections   297
Sound Track   302
Upcoming Meetings   303
Available Literature   304
Membership Information   305
Advertiser Internet Directory   307
In Memoriam   316
AES Annual Report   321
Sections Contacts Directory   322
AES Conventions and Conferences   328

FPO For the cover:

2004 March, Volume 52 Number 3

spine: 2004 March, Volume 52 Number 3