Digital Signal Processing Reference
In-Depth Information
ADC Input00
Da ta Window
Time domain
FFT
Cu t Out
Freq. do mai n
30
22
15
08
00
Hz
Illustration 17:
Frequency-time landscape of a sequence of tones
Frequency-time-landscapes are used among other things for the analysis of voice or in the investigation of
certain transient processes. The way in which a signal changes over time is represented. The frequency
axis is horizontal and the time axis goes back obliquely. The vertical axis gives the level.
Only a few years ago measuring instruments of this kind were very expensive (up to $ 25.000). In this case
this measuring instrument was put together by means of DASYLab using only 6 modules and of these two
are superfluous (modules "time domain" and "cut out") in this application.
It takes weeks or months to program a complex measuring instrument of this kind in the traditional
fashion. Using DASYLab it takes a few minutes including all settings and trial attempts. Is it necessary to
say any more about the benefits and advantages of the graphic programming of technical signal systems?
You only need to plug a microphone into the sound card and click on to the picture in the electronic PDF-
document). Have fun experimenting.
• An important special field of computer-aided signal analysis is data analysis. Here the
essential thing is to present a large amount of (stored) data - for instance, a large
number of measurements - in an ordered and structured fashion which makes it
possible to interpret them. In Illustration 16:, for example, this is the graphic, three-
dimensional representation of the contours of a deep-sea rift which results from
millions of echolocation measurements. These "measurements" can, for example, also
be share prices at the stock exchange and the aim of the analysis of data an
improved estimate or prediction of stock exchange trends. In this context completely
new technologies are used which can "learn" or which can be optimised by training :
fuzzy logic and neuronal networks or a combination of the two - neuro-fuzzy.
