Chapter 24. Audio on the Web
files found their way onto the Web in its
earliest days when they could be linked to and downloaded like any
other file. The drawback to this technique is that traditional audio
files are generally quite large and may take a prohibitively long
time to download. As the Web evolved, we've seen some major
breakthroughs in web audio. First, streaming audio (files that play
as they download) made long-playing audio and even live broadcasts
possible. Then the MP3 format exploded into popularity around 1999.
MP3's ability to crunch audio files to one-tenth their original
size while maintaining very good quality made it a perfect solution
for sharing music over the Internet.
Obviously, audio, even specialized for the Web, is a rich and complex
topic that cannot be thoroughly treated in a single chapter of a
Nutshell reference book. If you are interested in learning about all
the ins and outs of creating professional-quality audio for a web
site, I recommend starting with Designing Web
Audio by Josh Beggs and Dylan Thede (O'Reilly,
2001). It contains information on recording, editing, and optimizing
audio content, as well as in-depth discussions of popular web audio
This chapter introduces general audio concepts and a number of
popular web audio file formats, including WAV, AIFF, MP3, QuickTime,
MIDI, RealAudio, Windows Media, Liquid Audio, Flash audio, and
Beatnik's Rich Media Format. It also discusses the many options
for adding audio to a web site. It begins with an introduction to
basic audio terminology that will be useful to know when it comes
time to create and optimize sound files.
24.1. Basic Digital Audio Concepts
In order to distribute recorded speech or music over the Internet, an
analog signal must be converted to digital information (described by
bits and bytes). This process is called
It is analogous to scanning a photograph to a digital bitmap format,
and many of the same concepts regarding quality and file size apply.
Some audio file formats (such as MPEG) are compressed in size during
encoding using a specialized audio compression algorithm to save disk
space. In the encoding process, you may be asked to provide settings
for the following aspects of the audio file.
- Sampling rate
To convert an analog sound wave into a
digital description of that wave, samples of the wave are taken at
timed intervals (see Figure 24-1). The number of
samples taken per second is called the sampling
rate. The more samples taken per second, the more
accurately the digital description can recreate the original shape of
the sound wave, and therefore the better the quality of the digital
audio. In this respect, sampling rate is similar to image resolution
for digital images.
Sample rates are typically measured in kilohertz (KHz). On the high
end, CD-quality audio has a sampling rate of 44.1 KHz (or 44,100
samples per second). On the low end, 8 KHz produces a thin sound
quality that is equivalent to a transistor radio. Standard sampling
rates include 8 KHz, 11.025 KHz, 11.127 KHz, 22.05 KHz, 44.1 KHz, and
48 KHz. The new emerging high-end standard is 96K, which may be seen
in DVD audio but is not applicable to the Web. The higher the
sampling rate, the more information is contained in the file, and
therefore the larger the file size.
Figure 24-1. Audio wave after lowering sample rate and bit depth
- Bit depth
Like images, audio files are also
measured in terms of their bit depth (also called sampling resolution
or word length). The bit depth corresponds to the resolution of the
amplitude (or volume) of the sound file. The more bits, the better
the quality of the audio, and of course, the larger the resulting
audio file. This is similar to bit depth in images -- the more
bits, the more colors the image can contain.
Some common bit depths are 8-bit (which sounds thin or tinny, like a
telephone signal) and 16-bit, which is required to describe music of
CD quality. High end digital audio is now capable of 20-, 24-, 32-,
and 48-bit depths.
can support from one to six separate channels of audio information.
The most familiar of these are mono (one channel) and stereo (two
channels), but some formats can support three-, four-, or six-channel
(also referred to as 5.1-channel) surround sound. Most file formats
support only mono and stereo, but we will be seeing a lot more
support for multichannel surround formats in the coming years.
- Bit rate
Audio files are also measured in terms of bit rate, the number of
bits per second devoted to storing audio data. Bit rate is a function
of the file's bit depth and sampling rate, so you reduce the
bit rate by reducing a combination of those settings. Bit rate is
measured in kilobits per second (Kbps) and can be calculated by
dividing the file size by the length of the audio clip in seconds. In
general, it is advisable for the bit rate of streaming audio files to
be lower than the bit rate of the user's connection to the
Internet to ensure smooth playback.
It stands to reason that before you can put your own audio files on
the Web, you first need to create them. Your options are to find
existing audio resources (such as from a royalty-free CD) or to
record them yourself.
Copyright © 2002 O'Reilly & Associates. All rights reserved.