As musicians, the raw material for our art is sound. Any sound that we hear is a complex relationship of three basic elements: pitch, timbre and loudness. Musicians have always thought in terms of these elements when writing and performing music, but for the most part have used subjective terms to describe them. When dealing with synthesizers or other technology tools however, we need to think of these elements in measurable ways. There is no “mellow” button that gets pushed to put the finishing touches on a jazz mix; no “phat” button to give Puff Daddy’s mixes a street savvy. There are, however, a number of tools that we commonly use that can enhance these aspects of a production, and those tools are designed to work with specific elements of sound in well-defined, predictable ways.
Pitch - What makes a sound appear "higher" or "lower."
Pitch is a subjective description of how we perceive a note as being either “low” or “high.” We use different pitches when we play or sing to define the shape of a melody, and it’s the element of sound that makes a melody memorable. Not all sounds we use are going to have a clearly defined pitch. Ever try and play “Happy Birthday” on a snare drum? Not likely since drum, percussion and a whole range of other glorious sounds that we’ll call noise have no clearly defined pitch. So when we think of the kinds of sounds that we’ll use in building a piece of music, we can separate any sound we can imagine into two separate categories: pitched and non-pitched musical sounds. Generally speaking we’ll used pitched musical sound for traditional melody and harmony parts while non-pitched sounds will handle purely rhythmic functions like drums and percussion.
All sound that has recognizable pitch is the result of periodic vibrations; that is, a vibration that repeats itself over time in cycles. We measure pitch in terms of these periodic cycles over time, usually in cycles per second. The frequency with which these vibrational cycles occur determines the perceived pitch. The more cycles of a periodic vibration, the higher the frequency with which they occur, the higher the pitch.
Timbre (pronounced: tam-ber) Tone color, or what enables us to distinguish between two sources producing sustained sound at the same pitch.
Every sound whether it’s pitched or non-pitched has a certain tonal character, a kind of sonic fingerprint if you will, called timbre. Strictly speaking timbre is the element of sound that cues us in on the difference between two instruments playing the same melody. In addition to the basic note we hear as the pitch of a musical sound, the note that we’d sing, there are a whole range of frequencies that we call partials related to that note that give it a unique tone color. This lets us tell the difference between a piano and a flute playing the note middle C. When a sound has a clearly defined tone color like that of most wind and string instruments we say it has a “harmonic” set of partials. Sounds that are bell-like in nature and would fall into the category of pitched percussion sounds such as vibraphones and marimbas have “non-harmonic’ partials that in some cases make the note that they’re playing harder to hear. Non-pitched sounds like drums will also have non-harmonic partials.
Any sound has a unique Spectrum, a set of overtones or partials that causes it to have a unique timbre.
Loudness - The volume level or intensity of a sound.
For most of us the element of loudness is pretty obvious. We walk into a club and it’s LOUD! This is a common reaction to the overall intensity of the sound we hear, but when we think about any sound that we’ll use in making music, how loud it sounds will change over time. When you play a note on a piano it starts out with a certain loudness, depending on how hard you hit the key, and over time it will eventually fade away. For some musical sounds, like those found in traditional melody instruments, we’ll have a great deal of control over how long a note will sustain, and we’ll call those sustained sounds. For some, like percussion sounds, we’ll have little or no control, and we’ll call those kinds of sounds transient.
Historically, the effectiveness of an audio system has been judged on the basis of how accurately the system reproduces the original event. Typically we might listen to a recording of a string quartet and compare it with our experience of what that ensemble sounds like in a live performance. The slogan "is it live or is it Memorex?" reinforces this traditional notion of fidelity in the consumer marketplace. Today's audio productions have seized on the vast creative resources in audio technology to create recordings which defy our sense of a recording faithfully reproducing a live performance.
Issues of Fidelity
In comparing the performance of audio systems these three factors should be taken into consideration:
Any type of audio recording system has 3 major components:
At the heart of hard-disk recording and editing is digital audio. When we record digitally, sound is converted to an electrical signal by a microphone. That signal is coded into numbers by an analog-to-digital converter (ADC). The numbers are stored in memory, then played back upon demand by sending the numbers to a digital-to-analog converter (DAC). The resulting signal is sent through an amplifier and speakers so we hear a reproduction of the original sound. This is illustrated by the animation below:
Devices used to capture, store and access sound will fall into some combination of the following categories:
Some examples:
In digital recording we start with an analog audio signal and convert it to digital data to be stored. Changes in electrical voltage are encoded as discreet samples. On playback we retrieve the digital data and convert it back to an analog signal. Here, fidelity is dependent on the quality and function of the Analog-to-Digital (A-to-D) and the Digital-to-Analog (D-to-A) converter. Once an audio signal is stored as digital data, the storage media has no effect on the quality of sound.
The following chart shows storage requirements for a one-minute stereo audio file at different sample rates and resolutions. What is the tradeoff here in terms of fidelity vs storage? What is the frequency response for the different sample rates.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Storage requirements for one minute of stereo digital audio.
|
|
||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96k
|
22 MB
|
33 MB
|
Storage requirements for one minute of stereo digital audio.
Properties of Sound
Amplitude
Aperiodic
Compression/rarefaction
Decibel
Frequency
Fundamental
Harmonic
Hertz
Loudness
Non-harmonic
Oscillator
Overtones
Partial
Periodic
Pitch
Resonance
Sine wave
Spectrum
SPL
Timbre
Digital Audio
ADC
Analog
Bit depth
DAC
Digital
Distortion
Frequency Response
Linear/Non-Linear
Nyquist Frequency
Random Access
Resolution
Sample Rate
Signal to Noise Ratio