Digital Signal Processing Reference
In-Depth Information
DASYLab and digital signal processing
We have worked with DASY
Lab
from the first chapter onwards, that is in the final
analysis we have carried out computer-based signal processing. In the chapters dealing
with basic principles this was not specially mentioned. For educational reasons the
impression was given that we were dealing with analog signals. Almost all the
Illustrations accordingly appear to show analog signals.
Now, however, the moment of truth has come. Computer-based signal processing is
always digital signal processing (DSP). Here is a brief reminder of what this means.
Illustration 188 once again shows the situation by means of an example. A digitalised
audio signal is read from a disk and shown on a screen. Optically it appears to be a con-
tinuous analog signal. If however, using the lens function a tiny segment is zoomed out,
individual measuring points can be see which are joined up by straight lines. These
straight lines are produced by the program to make it easier to see the course (of this
originally analog signal). They are not present for signal processing.
These measurement points can be brought out more clearly, if so desired. Thus, each point
can be represented additionally by a small cross or triangle. The representation of the
points as vertical bars is very graphic.The height of a bar is equivalent to a measurement.
We should emphasise once again:
A digital signal consists pictorially in the time domain of a
discrete equidistant sequence of measurements which reproduce
the curve of the (original) continuous analog signal more or less
accurately.
In a less pictorially graphic way Illustration 189 represents this digital signal much more
accurately. For this purpose the module “list” is used. Here it can be seen clearly
• at what time intervals measurements (samples) of the analog signal were taken and
• with what degree of accuracy the measurements were registered (quantization - see
Illustration 149)
As a result of quantization the measurements are value discrete, i.e. only a finite number
of different “graduated” measurements can occur (see also Illustration 148).
Digital signals are time and value discrete in contrast to the time
and value continuous analog signals. This results in particular
requirements which must be taken into account in the processing
of digital signals. Without the necessary background knowledge
mistakes are unavoidable.
