Computer Audio Glossary


Bit rate- The amount of data per unit time necessary to store or transmit audio information for a particular audio encoding method.

A higher bit-rate usually translates to higher quality audio, for a given encoding scheme. Lower bit-rates have the advantage of lower cost of storage and lower cost of transmission. In computer audio, look for a three-way trade-off between bit-rates, audio quality, and computation requirements.

Sampling rate- For digital audio, the continuous audio waveform is measured(sampled) at discrete points in time. The number of sampling points in a length of time gives the sampling rate.

A high sampling rate allows higher frequency audio sounds. Higher sampling rates typically mean higher bit rates. For some applications, frequencies are already limited by the system, so it makes no sense to go to a higher sampling rate. Telephone systems, for instance, may be limited to maximum voice frequencies of around 4kHz, so a sampling rate greater than 8000 samples/sec is unnecessary.

Typical sampling rates range from 8000 samples/sec to 48000 samples/sec.

Resolution- When the continuous audio waveform is measured (sampled), the resolution indicates how much “precision” is used for each measurement. Think of two rulers of equal length. The first ruler has ten marks, and the second has 100 marks. The second ruler has higher resolution. In 8 bit resolution, 8bits are used to store the measurement, so the measurement can only take28 or 256 different values. In 16 bit resolution, 16 bits store the measurement, so the measurement can take on 216 or 64k different values.

Lower resolution translates to more noise. Lower resolution audio will usually have a louder back-ground hiss than higher resolution. Higher resolution usually means higher bit-rates. Very roughly, 8-bit resolution is telephone quality, and 16-bit resolution is CD quality. For some applications,resolution is already limited by the system, so it makes little sense to record at higher resolution. Recording telephone conversations at 8 bit resolution maybe fine– higher resolution may not be necessary.

For several of our programs, the audio resolution setting is a little more subtle: It gives the resolution of the audio before compression, and may or may not indicate the resolution of the compressed audio.

Audio Compression- Clever people look for ways to change the digital representation of audio to reduce bit-rate without greatly reducing quality.Some schemes try to take advantage of the repetitive nature of the audio waveform, and other methods take advantage of the fact that human ears perceive some sounds better than others, so we can through out the information used for sounds we wouldn’t miss.

Early audio compression schemes often didn’t reduce the bit rate very much.More recent compression (MPEG and others) give low bit-rates with good quality(at the cost of higher computation load, but Moore’s Law helps us there) and has spurred a revolution in digital audio.

Codec- From Coder-Decoder. (In this context, refers to a software codec.Hardware codecs have a somewhat different meaning.) The software which compresses the audio to a particular coding scheme, and then decompresses it again during playback.

In Windows, Microsoft defined a standard for software audio codecs, the ACM Audio Compression Manager. You have several audio codecs installed on your computer. Some came as part of the operating system, and some were installed as a by-product of installing or operating other software. (Some codecs, even from Microsoft, have their own interface, and do not conform to ACM. Several popularmp3 encoders do not support ACM.)

A good codec for many applications running under Windows is “MPEG Layer-3”. For a list of ACM codecs on your computer, look at Start button->Settings->Control Panel->Multimedia->Devices->Audio Compression Codecs. Mpeg layer-3 compresses to a small size (low bit rate),and the quality is usually acceptable. If you do not have this codec on your system yet, try installing the latest version of Windows Media Player, and see if that installs the codec.

Wave- An audio file format supported natively in Windows, with a .WAV file extension. Wave files can be uncompressed (PCM, pulse code modulation) or compressed, using one of the many ACM codecs.

Be aware, when you use compression in a Wave file that you intend to send to someone else, that not everyone has the same compression codecs installed on their computer.

Although Wave files are supported by many Windows programs, the audio does not “stream” well. That takes us to the next format:

Mp3- Based on MPEG layer 2 and layer 3 compression, mp3 is a streamable audio file format that has enabled many new digital audio applications, and stimulated the development of newer improved streaming formats.

Typical uncompressed audio bit rates range from 64kbits/s (kilo bits per second) to 1536 kbits/s. Even the lower rate won’t fit through a dial-up modem.Mp3 bit rates range from 8kbits/sec to 448kbits/s, with 128kbits/s being more typical for high end.

Quality- See fidelity.

Fidelity- See quality. ;-