Properties of the system
Linearity : No
The sum of two audio signals(vectors with amplitude and frequency of vibration), that
is the air pressure created due to them is added linearly and hence the equivalent
pressure is a like the pressure created due to the ‘sum‘ of the two signals
An example: Two signals A*sin(2pa*t) and B*sin(2pb*t) are signals with
frequencies ‘a’ and ‘b’ and intensities ‘A’ and ’B’ respectively produce certain
variations in the air pressure with time Pa(t) and Pb(t) respectively then the
equivalent pressure sensed by the system will be a sum equal to Pa(t) + Pb(t) with the
net result being that the system gives a linear output since the part of the system is
just represents the analog voltage data in digital form keeping its value unchanged or
maybe scaling it by a factor at the most.
If the microphone output does not linearly varying with the air pressure then the
above won’t hold though the system excluding the microphone would be linear again
assuming that the sum of two input pressures would not be above the dynamic uppe
limit in which case clipping (a phenomenon in which all the values above the highest
are clipped to the highest value) takes place.
Causality : Yes
Clearly no input pressure variation till time t=0 the voltage waveform of the
microphone is zero for all t<0 since there is no excitation. Thus no nonzero digits are
given out at the output end since values of all the samples is zero. Also this may be
said to be so since all real systems are causal.
Time Invariance :Yes
Since the value of the output binary number depends only on the value of current
input the system is time-invariant.
Stability : Yes
Since the value of the output signal is just a scaled version of the input pressure of the
air which in turn is always bounded, the system is a stable one.
Memory : No
Also the sampler gives out only the output at only a given instant in time the system is
memoryless. However systems cascaded to this one use zero order hold, linear
approximation etcetera then the system(inclusive of the approximating system) no
longer remains memoryless.
Invertibility : No
Since the data of the values of the original signal between the sampling points is no
longer known, the original waveform cannot be reconstruced however good the
sampling mechanism be and hence the system cannot be invertible.
With respect to the other inputs the system, the sampling rate and the bit-depth is Causal,
Time Invariant and Memoryless. The other two properties have no significance with
respect to these inputs.
References:
Fundamentals of Digital audio : P. Jeffrey Bloom and Guy W. McNally
http://www-ccrma.stanford.edu/courses/192b/lectures/2/2.html
http://www.tc.umn.edu/~erick205/Papers/paper.html
http://www.teamcombooks.com/mp3handbook/MP3_Handbook.htm
|