Data processing: generic control systems or specific application – Specific application – apparatus or process – Digital audio data processing system
Reexamination Certificate
1998-04-20
2004-01-06
Harvey, Minsun Oh (Department: 2747)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Digital audio data processing system
Reexamination Certificate
active
06675054
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to the field of digital audio, and, more specifically, to digital audio applications in a network environment.
Sun, Sun Microsystems, the Sun logo, Sparc, Java and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States and other countries.
2. Background Art
Computers and computer networks are used to exchange information in many fields such as media, commerce, and telecommunications, for example. One form of information that is commonly exchanged is audio data, i.e., data representing a digitized sound or sequence of sounds. Voice telephone transmissions and video conferencing feeds are examples of telecommunication information which include audio data. Other examples of audio data include audio streams or files associated with digitized music, radio and television performances, or portions thereof, though audio data may be associated with any type of sound waveform. It is also possible to synthesize sound waveforms by artificially generating audio data having desired magnitude and frequency characteristics.
For the purposes of this discussion, the exchange of information between computers on a network occurs between a computer acting as a “transmitter” and a computer acting as a “receiver.” In audio applications, the information contains audio data, and the services provided by the transmitter are associated with the processing and transmission of the audio data. A problem with current network systems is that multiple services, provided by one or more computers acting as transmitters, may provide audio data using different audio protocols. The complexity of the receiver is necessarily increased by the need to accommodate each of the different audio protocols. Further problems associated with the transmission of audio data over a network include errors in the audio signal caused by packet loss, as well as undesirable latency in real-time, or time-critical, audio-related applications such as video conferencing. The following description of audio technology and an example network scheme are given below to provide a better understanding of the problems involved in transmitting audio data over a network.
General Audio Technology
Audio data technology allows for the capture, storage, transmission and reproduction of sound. To understand how sound can be represented electronically as audio data, it is useful to understand the general nature of sound. Sound refers to a pressure wave propagated through a medium, such as the air. A pressure wave of this sort may be generated, for example, by the vibration of the vocal chords in a human throat, as when speaking or singing, or by a collision of two objects, where a portion of the energy of the collision is dissipated as a pressure wave. The medium through which the pressure wave is propagated attenuates the pressure wave over time in accordance with the physical characteristics, or “acoustic properties,” of the medium.
When pressure waves meet the eardrum of a human ear, the eardrum flexes and vibrates in response. The vibration, or modulation, in the eardrum is interpreted by the brain as a sound. An electronic capture mechanism, such as a microphone, has a similar mechanism for detecting pressure waves and generating an electronic signal containing corresponding audio data. A sensor mechanism in the microphone is physically modulated by a pressure wave, and the modulation is electro-mechanically transformed into an electronic signal. The electronic signal may be transmitted or stored directly, or, as is now typically done, the electronic signal may first be digitized (i.e., sampled and quantized). A sound is reproduced from audio data by transforming the electronic signal back into a pressure wave, for example, by electro-mechanically modulating a membrane to create the appropriate pressure wave.
Sound Waveforms and Data Sampling
The electronic signal corresponding to a captured sound may be graphically represented by a sound waveform, such as sound waveform
100
illustrated in FIG.
1
A. The vertical axis of
FIG. 1A
, as well as that of
FIGS. 1B and 1C
, represents the amplitude of the sound waveform, with the horizontal axis representing time over a period of one millisecond. Sound waveform
100
is a continuous waveform.
FIGS. 1B and 1C
illustrate discrete sampled waveforms generated by sampling sound waveform
100
at sampling rates of twenty-four kilohertz and eight kilohertz, respectively.
A sampling rate is expressed in hertz or samples per second. A sampling rate of twenty-four kilohertz implies that twenty-four thousand samples are taken per second, or one sample is taken approximately every forty-two microseconds. As one would expect, the sampled waveform of
FIG. 1C
, with a sampling rate of eight kilohertz, has one-third as many samples as the sampled waveform of FIG.
1
B.
Higher sample rates generally entail correspondingly greater resource costs in terms of storage and transmission bandwidth requirements to accommodate the data associated with the larger number of samples. However, a higher sampling rate generally provides a more precise reproduction of a sound waveform. The ability to reproduce an original waveform from a set of sampled data is determined by the frequency characteristics of the original waveform and the Nyquist limit of the sample rate. Every signal or waveform has frequency characteristics. A relatively fast changing signal level is associated with higher frequency behavior, whereas a signal level that changes slowly is associated with lower frequency behavior. Most signals have frequency contributions across a broad spectrum of frequencies. The frequencies associated with audible signals, and hence sound waveforms, reside generally within the range of 20-20,000 kilohertz.
According to Nyquist theory, a sampled signal can reconstruct an original waveform from sampled data if the original waveform does not contain frequencies in excess of one-half of the sampling rate. That is, if an original waveform is bandlimited below ten kilohertz, a sampling rate of twenty kilohertz or higher would be sufficient to reproduce the original waveform without distortion. When relatively low sampling rates are used, it is common to pre-filter waveforms to bandlimit frequency behavior and prevent or diminish distortion caused by the sampling process. However, filtering of a sound waveform may result in lower sound quality because higher frequency components of the waveform are attenuated.
Different audio protocols may use different sample rates for audio data. A receiver that is generating sound output from audio data needs to be able to handle the different possible sample rates of the different audio protocols to maintain correct timing intervals between samples during reconstruction of the sound waveform from the audio data samples.
Data Resolution and Quantization Schemes
Another aspect of audio data that differs between audio protocols is the quantization-scheme used to quantize or digitize the amplitude of the sampled audio data into digital values that can be represented by a fixed number of bits. The number of bits used to represent each sample of audio data is the resolution of the given audio protocol. Typically, for M bits of resolution, 2
M
possible digital values or quantization levels exist for sample quantization. For example, eight bits of resolution provide 2
8
, or 256, quantization levels. Higher resolution typically provides for better sound reproduction as sound samples are more precisely represented. Higher resolution also entails higher costs in storage resources and transmission bandwidth to support the larger number of bits.
Just as there are different possible resolutions for audio data, there are also different quantization schemes for distributing the quantization levels across an amplitude range.
FIGS. 2A and 2B
illustrate examples of linear and non-linear quantization functions, respectively. The horizontal axis of each of
FIGS. 2A and 2B
Harvey Minsun Oh
O'Melveny & Myers LLP
Sun Microsystems Inc.
LandOfFree
Method and apparatus of supporting an audio protocol in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus of supporting an audio protocol in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus of supporting an audio protocol in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3230017