MUS 570 / CSC 570
DIGITAL AUDIO TECHNIQUES

Terms

Return to Home Page

Return to MUS 570 / CSC 570 Menu

AIFF--Audio Interchange File Format developed by Apple for the Macintosh OS.  Compare to the Windows .wav format, but these files contain more information in a header section which allows the possibility of more advanced audio processing.

ALIASING--Unwanted noise that occurs from sampling a frequency at a sample rate that is less than what is required to accurately represent that frequency. (See Nyquist Theorem.) 

AMPLITUDE--Amplitude represents the loudness of a signal, measured by its distance from the center-line (=  0 amplitude value). Amplitude is most often measured in decibels (dB).

ANALOG--Represents events that are recorded as continuous, as opposed to digital, where events are recorded as discrete steps or numbers.  Example:  a wristwatch with hands would be analog, whereas an LED/LCD watch with only number readings would be digital.  In terms of recording, an analog recording uses voltages to represent acoustical vibrations that are normally recorded as a corresponding series of magnetic fluctuations on tape.

AU--A file format native to Sun workstations.  It remains in use due in part to the fact that most Internet servers are running Unix.  

AUDIO FILE FORMAT--The method used for writing audio data to disk for storage is known as File Format. Not all file formats are compatible.  Windows PCM (.wav) is found most often in the Windows environment. Typically, a variety of file formats can be chosen under the menu heading File when saving.

BAUD RATE--The number of bits per second being transmitted.

BIOS--Basic Input Output System that controls a computer and provides its low-level routines. It is the BIOS which handles the instructions for starting a computer, initializing the hardware, and launching the operating system.

BIT--Part of the digital numbering system used in electronic equipment. A bit makes up a larger number, called a word, which is used to represent the voltage level (amplitude) of an analog signal. The number of bits used to make up a word determines the bit resolution, which in turn determines the number of voltage levels possible in representing amplitude. Therefore, greater bit resolutions make for higher signal-to-noise ratios.

BIT RESOLUTION--The number of bits used in measuring amplitude for a sample. Choosing 8-bit resolution will provide 256 possible unique "volumes", while choosing a 16-bit resolution will provide 65,536 possible unique "volumes," which results in a 96 dB signal-to-noise ratio. Obviously, a much greater dynamic range can be reproduced at 16-bit resolution than at 8-bit resolution, which only has a 48 dB signal-to-noise ratio. Compact disk players have a 16-bit resolution. Some sound cards support resolutions greater than 16-bit. Audition supports up to 32-bit sample rates. It is best to remain at the 32-bit level while working in Audition, and convert down for output, in order to best represent the audio after transforms have been preformed.  (The more bits are used to describe each sample, the more possible amplitudes it has. Eight-bit sound allows for each sample to have 256 possible amplitudes, 16-bit allows for 65,536, and 24-bit allows for almost 16.8 million amplitudes.)

BROWN NOISE--See Noise, Brown.

BURN--Writing to a CD-R or CD-RW.

CD--Compact disc (audio only).  A digital storage medium that holds 700 MB of data, or 80 minutes of digital audio; uses a 44.1kHz sample rate, Stereo, 16-bit linear PCM.

.CDA--Compact Disc Audio (CD-DA, Compact Disc-Digital Audio), a format which is the basis for most commercial CDs.

.CEL--Cool Edit Loop file format, compressed but does not extend file time lengths.

CD-R--A CD that may be recorded one time--available in two capacities: 650 MB (74-minute) and 700 MB (80-minute). Mini CD-R--available in a 180 MB (21-minute) capacity only.

CD-RW--A CD that may be re-recorded--available in a 650 MB (74-minute) capacity only..  You cannot play a music CD made on a CD-RW disc in most home or car stereo CD players. 

CHORUS EFFECT--Combining the main signal with several delayed versions of the signal. It replicates the effect of a chorus singing a single note providing a fuller sound.

CLICK TRACK--An audio track comprised of clicks that keep time, like a metronome. They are silenced for the finished product.

CLIPPING--When the amplitude of a signal exceeds the maximum level that is available for current conditions (i.e. 256 in 8-bit audio), a process known as clipping occurs. This causes the signal to distort, and visually appears in the display as a "chopped-off" waveform at the top. Lower recording, or source output, levels when recording if you experience clipping.

CROSSFADE--To fade from one audio track to another.  One track fades as another increases in volume.

DAC--Digital-to-Analog Converter.

DAT--Digital Audio Tape. DAT tapes are sampled at 16-bit and 32,000, 44,100, and 48,000 (which is considered "DAT quality") samples per second.

DECIBEL (dB)--A standard unit of loudness; 1/10th of a Bel (named after Alexander Graham Bell). Example:  threshold of hearing = 0 dB; normal conversation = 60 dB; threshold of pain = 130 dB.

DIGITAL--Events are recorded as discrete steps or numbers, as opposed to analog, where events are recorded as continuous.  Using numbers (digital) to represent acoustical vibrations which are measured at equal intervals of time.

DISC--Storage medium for music or video information.

DISK--Computer storage medium, usually 3.5".

DITHERING--Adding a low level of noise to an audio signal so that fainter audio can be heard that otherwise would be cut off when converting from a higher bit rate to a lower bit rate (e.g., converting 24-bit to 16-bit).

DVD--Digital Video (or Versatile) Disc. A storage medium similar to Compact Disc (CD), but with much higher bandwidth and storage capabilities. Audio stored on DVD movies is generally 96 kHz/24-bit.

EFFECTS--Echo, Chorus, Digital Delay, Reverb and other modifications produced by passing a sound or audio signal through an effects processor.

EQUALIZATION--Changing the frequency balance of the original waveform by boosting or eliminating particular frequency bands.  An equalizer has a series of sliders, each with a particular frequency listed beneath.  One may increase or decrease the volume of particular frequency ranges by adjusting the sliders.  (See Equalizer graphic under SoundStream on Roxio Easy CD Creator page.)

FILE SIZE:  Remember that changing any of the three sample categories (sampling rate, mono/stereo, bit resolution) will significantly change the file size, as the chart below illustrates.  Simply experiment to see what settings will best meet your needs.  (Red cell represents CD-quality rate ... 10.3MB per minute.)

 

Approximate file size for ONE MINUTE of sound
(1 floppy = 1.44 MB)
(1 CD-R = 700 MB)
Channels MONO MONO STEREO STEREO
Resolution 8-BIT 16-BIT 8-BIT 16-BIT
Sampling Rate        
44,100 Hz 2.5 MB 5.1 MB 5.1 MB 10.3 MB
22,050 Hz 1.2 MB 2.5 MB 2.5 MB 5.1 MB
11,025 Hz 646 kb 1.2 MB 1.2 MB 2.5 MB
8,000 Hz 469 kb 938 kb 938 kb 1.8 MB

FREQUENCY--Measured in Hertz (Hz), vibrations per second that determine pitch.  For example, the orchestral tuning note, A440 = 400 vibrations per second, sounding the pitch of "A."

FREQUENCY RANGE (of human hearing)--A sound must have a frequency of roughly 20 Hz or more on the low end to be heard.  On the high end, young ears can hear frequencies of up to 17,000 Hz (17 kHz) or more.  Older folks are doing well to hear 12 kHz frequencies.

HERTZ (Hz)--The unit of measurement of frequency equal to one cycle of a periodic waveform per second (in English ... vibrations per second ;-).  (K = thousand.  3 kHz = 3,000 Hz)

IMA--Interactive Multimedia Association file format which has a 4-to-1 compression ratio.

LEVEL METERS / VU (Volume Unit) METERS--Visual meters that indicate the volume levels of incoming and outgoing audio signals.

LIMITER--Limiters are compressors with a compression ratio of 10:1. They reduce or "limit" input signals that exceed a specified threshold level, so that the output does not increase in gain beyond the specified point.

MIDI--Musical Instrument Digital Interface, a standardized protocol for communication between electronic devices and computers.  File extension is .mid

MIDI TIME CODE (MTC)--A method of sending timing information between MIDI devices. For instance, SMPTE, which is an audio signal, can be converted to MTC to sync to and control Cool Edit Pro's transport from a device such as a VCR or tape deck.

MINIJACK PORT COLORS--The 1/8" phone plug ports are typically color coded.  Red = microphone.  Green = speakers.  Blue = Line In.  Black = Line Out.  Hot Pink = Man, where'd you buy that computer??

MINIPLUG--A 1/8" phone plug (stereo or mono).

MIXDOWN--The process of combining the output of all enabled tracks (or selected tracks) into a new, single stereo waveform. When doing a Mixdown, track properties such as Volume and Pan are reflected in the resulting waveform. Mixdown is generally the final operation, done when you are happy with the way your session sounds, or can be used to create submixes, or mixes of selected tracks (such as all of the drums) which is useful for uncluttering your workspace, or for creating alternate mixes.

MOV--Apple's QuickTime file format, a common multimedia player.

.MP3--technically, "MP3-1 Audio Layer 3," as so-dubbed by the Motion Picture Experts Group (MPEG).  The short answer is that MP3 is a file compression technique that reduces sound files to about one-tenth of their original size.  All the sounds are stored as numbers.  MP3 removes any numbers representing sounds beyond the range of human hearing.  While MP3 files aren't "exactly" the same quality as a CD audio file, they come pretty close.

MPEG--Audio compression schemes created by the Motion Picture Experts Group.

MULTITRACK RECORDING--Recording sounds on individual tracks that all play back at the same time (unless muted).

MUTE--A switch or control that zeroes out or turns off the audio from a specified track.

NOISE, BROWN--Brown noise has a spectral frequency of 1/f^2. This means that there is much more low-end, and there are many more low-frequency components to the noise. This results in thunder- and waterfall-like sounds. Brown noise is so called because, when viewed, the wave follows a Brownian motion curve. That is, the next sample in the waveform is equal to the previous sample, plus a small random amount. When graphed, this waveform looks like a mountain range. The wave pattern is very predictable.

NOISE, PINK--Pink noise has a spectral frequency of 1/f and is found mostly in nature. It is the most natural sounding of the noises. By equalizing the sounds, you can generate rainfall, waterfalls, wind, rushing river, and other natural sounds. Pink noise is exactly between brown and white noise (which is why some people used to call it tan noise, but pink was more appealing). It is neither random nor predictable. It has a fractal-like nature when viewed. When zoomed in, the pattern looks identical to when zoomed out, except at a lower amplitude.

NOISE, WHITE--White noise has a spectral frequency of 1. In other words, equal proportions of all frequencies are present. Because the human ear is more susceptible to high frequencies, white noise sounds very "hissy." Cool Edit generates white noise by choosing random values for each sample.

NYQUIST THEOREM--Signals with frequencies up to a little less than half the sample rate can be sampled without distortion. That is, the sampling rate must be at least twice the number of cycles of vibration per second of the highest recorded frequency in order to accurately represent the sound without distortion.  The Nyquist Frequency is that magic sampling frequency where one can sample without distortion.

PAN--The placement of sound left-to-right in the stereo field.

PCM--Pulse Code Modulation is the standard method of digitally encoding audio. It is the basic uncompressed data format used in file types such as Windows' .wav and Apple's AIFF.

PINK NOISE--See Noise, Pink.

PUNCH IN--Selecting a specified area in which to record or re-record.

RAM--Random Access Memory, the memory required to load and run programs. This is what you refer to when you say that your computer has 64 megabytes of memory.

REAL AUDIO (*.ra)--Real Audio is a compressed format used for streaming audio over the Internet in real-time. You can save to this format from 8- or 16-bit, mono or stereo, with a number of valid sampling rates.

REVERB--An audio effect that adds a spacious, echo-like character to a sound.

RM or RAM--Real Audio Media file format used by RealNetworks, one of the most commonly found streaming technologies on the web.  

RMF--Rich Music Format developed by Beatnik; similar to MIDI files. 

ROM--Read Only Memory Memory that cannot be written to after it is constructed.

SAMPLE--To digitize a sound, a number of snapshots, or samples, are taken every second (sample rate) and are changed into numeric representations. A sample is a single "snapshot" of the sound. The term sample can also refer to the digitally recorded sound itself, i.e., the word can be a verb or a noun.

SAMPLE RATE--The sample rate determines the number of times per second the snapshot of the audio is taken. Higher sample rates produce higher quality audio, but at the expense of requiring more disk space. Frequencies of up to 1/2 the sample rate can be produced effectively, so to reproduce a frequency of 10Khz, a sample rate of at least 20Khz must be chosen. (See Nyquist Theorem.) 
8,000 Hz = telephone quality
11,025 Hz = poor AM radio quality
16,000 Hz = reasonable compromise between 11 KHz and 22 KHz
22,050 Hz = near FM radio quality
32,075 Hz = better than FM radio quality (Some boards support 32,000 instead)
44,100 Hz = CD quality

48,000 Hz = DAT quality
96,000 Hz = DVD quality.
  (Also, see File Size chart above.)

SPECTRAL VIEW--The alternate view of waveform data, as compared to waveform view.  This mode is helpful for spotting prevalent frequencies via color.  Brighter colors represent greater amplitudes.  Dark blue represents virtually no amplitude components within a frequency range.  Bright yellow represents higher-amplitude frequencies.  Lower frequencies appear near the bottom of the window, and higher frequencies are displayed from the middle-to-top part of the scale.

SMPTE TIME CODE--Developed by the Society of Motion Picture & Television Engineers (SMPTE), this is the standard digital code used to synchronize sight and sound.

S/N RATIO--The Signal-to-Noise Ratio is the ratio between the noise floor (present noise) and signal. Higher ratios are more desirable, as they mask more background noise.

STRIPE--To copy SMPTE time code to a single track of a multitrack tape. The SMPTE is then used to reference time on the other tracks for purposes of syncing those tracks to other devices.

SURROUND SOUND--A feature of Audition, although there are a few requirements your computer will have to meet first.  Audition offers six-channel surround sound whereby six speakers are placed in a quasi-circular pattern around a central point ... the listener.  

TIME CODE--An audio or digital signal designed to synchronize time between multiple devices, such as audio and video. The most common forms are SMPTE and MIDI time code (MTC).

TRACK--A channel on a tape recorder, sequencer or other recording device; one of a set of concentric circles on a floppy disk.

.WAV--The audio format for the Windows operating system, which has become the standard for sampled audio on the web.  .WAV encodes sound using Pulse Code Modulation (PCM), a process that digitizes the analog sound wave by sampling it a certain number of times per second and storing each sample as an 8-bit or 16-bit "word" (bit depth).  The higher the sample rate and bit depth, the more accurate is the sound reproduction.  On the down side, .wav files tend to be large.  Audio CD resolution is 16-bit, stereo, 44.1 kHz.  Web audio can be as low as 8-bit, 8 kHz.

WHITE NOISE--See Noise, White.

.WMA--Windows Media Audio file, developed by Microsoft.  Like MP3 files, WMA files compress information.  WMA files also have the ability to program the file so it cannot be copied.  WMA file sizes can be significantly smaller than even the compressed MP3 files.

Return to Home Page

Return to MUS 570 / CSC 570 Menu