Module 5 - Digital audio


Digital recording and playback
Sampling rates
Sampling resolution
Recording - levels and editing
Popular File formats
A word about MIDI
Things to do


You will be able to:

  • understand the basic concepts of frequency
  • understand the basic concepts of digital recording or sampling
  • select and manipulate software options for optimum quality recording
  • recognize and understand various sound formats

You should already know that sound is produced when something vibrates and sends waves of pressure through the air.  The number of wave cycles per second that sound produces is known as frequency.

These wave cycles are measured in Hertz.  One cycle is equivalent to one Hertz (Hz). 

The higher the pitch of a sound, the higher the frequency.

Frequency is measured in Hertz.  1 Hertz is equal to one cycle.

Human beings can perceive frequencies of about 20Hz to 20,000Hz.  20Hz would be a very, very low sound that you would almost feel instead of hear.  20,000Hz would sound like a high hiss if your ears are even sensitive enough to hear it!

As the frequency goes up or increases, you hear the pitch getting higher.  When the frequency goes down, you hear sound going down in pitch.

Digital recording and playback

A good recording device should be able to produce frequencies accurately within the range of human hearing.  An analog recording converts sound waves into electrical signals which in a tape recorder, line up magnetic particles glued onto a plastic film… or tape.  Everytime you make a copy of an analog recording, there is a degradation of quality with each generation made.

Digital recording takes the information from the original sound waves, converts them into a series of numbers, and stores them on a digital storage medium (such as a hard drive or mini-disc).  This process is known as digitizing or sampling.  Theoretically, digital recordings do not loose any quality no matter how many "copies of copies" you make.  After all, you are only copying numbers!

Digital Recording

The original sound waves are converted into an electrical signal by a microphone.  In the computer, an analog-to-digital converter (ADC) changes this signal into digital information and stores it.

Digital Playback

When played back, a digital-to-analog converter (DAC) in the computer changes the digital information back into an electrical signal which is then amplified and sent to a speaker.  The sound waves produced are virtually identical to the original waves.

Sampling rates

The number of samples taken from a waveform in one second is known as the sampling rate.  Notice how taking more samples per second produces a more accurate depiction of the wave, and therefore, yields a higher-fidelity recording.  As you can guess, higher or faster sample rates also create larger files.

Normal hearing ranges from 20Hz  to 20,000Hz.  Always use a sampling rate that is at least twice that of the highest frequency you want to reproduce.  This is because something known as "anti-aliasing"  occurs when you sample at the same rate as the frequency.  This "half of" frequency is known as the nyquist limit or frequency.

Computers offer 44.1kHz, 24 kHz, 22kHz, 11kHz, and 6kHz sampling rates (or numbers very close to these).  11kHz is acceptable for speech only if you are desperate to save storage space.  It is also better to sample at a higher rate initially and then reduce or downsample it if you have to.

Sampling resolution

Sampling resolution is another factor that affects digital sound quality.  This refers to how many bits will store information about the amplitude or loudness of a sample.  There are two resolutions that we will deal with: 8-bit sampling resolution means that 256 volume or loudness levels can be created.  16-bit resolution enables thousands of distinct loudness levels.  Using the highest sampling rates and resolution produces the best quality recordings, and amounts to a tremendous amount of data.  A stereo recording with 2 tracks at 44,100 samples/second at 16 bits/sample requires a data rate of 1,411,200 bits/second.

You should use the 44kHz, 16-bit settings when recording audio.




Click on the audio link icons to hear the difference in quality between a 5.5KHz/8-bit, 11.25KHz/8-bit and 44.1KHz/16-bit sampled  file.  The files have been modified to play quickly in Quicktime, but the original 44.1KHz file was almost 10 times larger than the 5.5KHz file!

Setting levels

As in any type of audio recording, the proper adjustment of recording levels is crucial to quality.  The record level should register as high as possible without going into the red.  Levels that are too high will sound distorted (clipped), and levels that are too low will have a different type of digital distortion, as well as pick up background noise.

The above is an actual waveform from Macromedia's SoundEdit 16

The top meter (L) is going into the red which indicates that the recording level is too high.  This will result in distortion.  The bottom meter indicates the proper recording level.

After your sound has been recorded it may be edited.  You can apply various filters and effects including reverb, echo, phase shifting, equalization and more.  You can cut and paste sound clips as you do words in a word processor.  The instructor will demonstrate this in class.

Popular File Formats

AIFF (.aif)  - Audio Interchange File Format is a popular high-resolution format. The format was designed with cross-platform file exchange in mind. It's your best choice when working on the Mac.

MIDI - Musical Instrument Digital Interface files are used mostly be musicians.  The file size is very small so they are quite popular on the Internet.

MP3 - a compressed audio file that uses a form of compression which achieves nearly a 12:1 compression ratio without much loss of sound quality. Very popular on the Internet for streaming music!

RealAudio (.ra) - another compressed format for streaming audio over the Internet.

SYSTEM SOUNDS - are Mac standard file formats.  They can easily be played from the desktop or integrated into most programs.

WAVE (.wav) - files are a Windows sound file format.

Quicktime (.mov) - are a popular format for streaming or quick downloading audio or video on the WWW or for use in other applications.

A word on MIDI

MIDI files are very small because they do not contain actual sound, but rather a set of numerical instructions.  For example, there are special number sequences for which note was played, how long, how hard and by which instrument.  This information is sent to a sound card, sound module or even software which applies these instructions to the specific instrument sounds that they themselves contain.

To listen to some MIDI files click here.  If you have a Mac it will convert the MIDI file to a QT movie and play it using software generated instruments.  If you have a PC, your sound card will have the actual instrument sounds built-in.  If you have an expensive sound card or an external sound module, the music can sound very realistic.

Things to do
  • Begin this assignment on the audio segment of your present project. 
  • Browse through some audio sites in the resources section.  Try downloading a few wave and midi files to your desktop.
  • If you would like to learn more about MIDI, click here for a tutorial.

Michael Shaw, 1997