Go to Wesleyan Homepage Go to Navigation Menu Go to Directories Go to Events Calendar Go to Search Wesleyan Go to Portfolio Sign-in
Multimedia Classroom
Title: Technology
graphic
Button Home
Button About Us
Button For Alumni
Button For Everyone
Button For Faculty & Staff
Button For Students
ACS Home Page
Academic (Technology) Roundtable
New Media Lab
Blackboard
Classroom Recording
Classroom Technology
Communications Tools
Faculty Support
Media Resources
Research Computing & UNIX
 

Digital audio basics

Following are some basic terms and concepts associated with digital audio.

Amplitude
Amplitude indicates the relative strength or loudness of a sound. When you see a waveform representation of a sound, the taller the wave, the louder the sound.

Bit depth
A different number of bits can be assigned to each digital sample. The more bits assigned, the greater the dynamic range, or difference between the softest and loudest sound. The most common bit depths are 8 and 16-bits. An 8-bit sample has a dynamic range of 48 decibels (dB) -- comparable to AM radio. A 16-bit sample has a dynamic range of 96 dB -- comparable to CD audio. The human ear can hear differences in sound up to approximately 120 dB.

Clipping occurs when the recording levels are set too high. In a graphical representation of the sound wave, the tops and bottom of the waveform appear cut off, and in those areas where the sound is clipped.

Compression and file size
A CD-quality audio file (44/16/stereo) is about 10 Mb per minute in size. If the same file is saved for the web at (22/16, mono), the file will be about 5 Mb per minute. Doubling either the bit rate or the sample size, doubles the file size, as does moving from mono to stereo. In addition to reducing the file size by controlling sample rate and bit depth, it's possible to apply compression. MP3 is currently the most popular compressor for audio.

File format

Uncompressed
AIFF (.aiff) - Audio Interchange File Format. Supported on Mac and Windows. Archival format. 
WAVE (.wav) - A Windows standard, also supported on Macs.
 
Compressed
MP3 - MPEG-1 Layer 3 (.mp3) excellent compression/quality. ISO standard. Any player.
QuickTime (.mov) - proprietary, supported on Mac and Windows. Needs QuickTime player.
Real Audio (.ra) - proprietary, needs Real Player.  
Windows Media (.asf )  proprietary, needs Windows Media Player.

Filters
Audio editing software generally include a number of filters to correct or enhance sound. Following are some standard filter:

Dynamic Range is the difference between the loudest and quietest part of the signal. Orchestral pieces generally have wide dynamic range, rock a narrow range. This filter can boost quiet parts while limiting loud parts. 
Normalize to boost volume to maximum possible without causing clipping or distortion. Normalize at about 85% for web audio.
Noise Removal removes random noise: hiss, static, ventilation hum, fan, etc.
Notch Filter removes specific frequency (eg. 60 kHz for electrical line noise)
Noise Gate silences sections of audio that are quiet but still have some noise. Set "threshold" and "duration." Use Noise Removal and/or Notch Filter first.
Low Pass Filter allows low frequencies to pass through, eliminates high frequencies above selected frequency
High Pass Filter removes frequencies below value given in kHz
Reverb can be added to create a richer sound, although this will not square with audio purists.

Frequency or pitch
Refers to the highness or lowness of a sound. The closer together the peaks in a sound wave, the higher the sound. Frequency is measured in Hz or kHz. Human hearing ranges from approximately 20-20,000 Hz. In a waveform representation of sound, the closer together the waves, the higher the frequency.

Sample rate
Is the rate at which sound is captured when it is converted from analog to digital. There is a direct relation between sample rate, sound quality and file size. The human voice can be effectively reproduced at a sample rate of 11 kHz (11,000 samples a second) . Audio CD quality requires a sample rate of 44 kHz. (Note that the sample rate of digitized sound is two times the playback rate -- a 44 kHz sample plays back at 22 kHz.)

Streaming audio
Sound files are either downloaded to the listener's computer from the web and then played with a player or  application or they are "streamed", which means they begin to play while they continue to download. There are two types of streaming, HTTP streaming, sometimes called  'fast start' or 'progressive download,' and RTSP streaming, which requires a streaming server. The campus network is fast enough that streaming is rarely needed for performance reasons.

 

DS 6/04

-->

graphic
graphic
graphic