A computer can only handle discrete pieces of data. A signal like
the one depicted above will therefore be transformed into a set of
discrete values. Each of these values is called a sample and
the number of values per second used to represent the signal is called
its sampling rate. Therefore, if a signal has a sampling rate of 44:1kHz
(the rate used in audio CDs), then the signal is measured 44100 times
per second. The individual values making up the signal are of course
discrete as well. If the signal is stored in the computer by using
2N different values, it is said to have a sampling
depth of N. Sound on an audio CD has a sampling depth of 16,
meaning that it is stored using 216 = 65,536 different values.
The characteristics (sampling rate and sampling depth) of a sound are
displayed in the bottom left corner of the sound window. For practical
purposes, it is usually assumed that samples take values between
–1 and 1. In the previous example of a 16bit sound, this means
that every sample can take 65,536 values that are equally spaced
through the interval [–1,1]. If a sound containing samples that
are close to these limits is amplified, a phenomenon known as
clipping
occurs. This is when, as a result of a sound effect
being applied to it, some samples would like
to take values
outside of the [–1,1] range and are forced back into it by the
computer. If the sound signal represented above was to be amplified
too much, it could therefore end up looking like this:
When clipping is severe, it introduces very noticeable artifacts that are quite difficult to get rid of, one should therefore take care to avoid it.
Humans perceive sounds with two ears. If one wishes to reconstruct accurately an auditory experience, it is common to use two microphones, one for each ear. This is called a stereo sound. When only one microphone is used, it is called a mono sound. The following screenshot shows how a stereo sound is displayed in Amadeus Pro:
The upper signal coresponds to the left ear and the bottom signal corresponds to the right ear.
Note that the structure of the signal (also commonly referred to
as the waveform
) only becomes apparent when one zooms into the
signal. For example, in the example above, the width of the whole
display only corresponds to about 50 milliseconds (one twentieth of a
second). Zooming into the display can be achieved with the
Horizontal zoom slider:
The number appearing next to the slider when you move it represents the number of samples of sound that fit into one pixel of your screen.
When a larger portion of the sound is shown, a typical display would rather look like this:
Here, you can see that the width of the display is about one minute. At this scale, the structure of the waveform is not apparent anymore. What you can see essentially represents the volume of the sound as a function of time (the fatter the dark blue area, the louder the sound).
