Next

The Time and Frequency Domain Models

 

We have seen that all periodic functions may be represented in the form . This is the time domain model for the signal as it expresses the amplitude as a function of time. Using the identity sin(A+B) = sin(A)cos(B) + sin(B)cos(A), we may express

Thus we may express all periodic functions in the form.

Using eiθ = cos(θ) + i sin(θ) and e-iθ = cos(-θ) + i sin(-θ) = cos(θ) - i sin(θ), we may derive cos(θ) = ½(eiθ + e-iθ) and sin(θ) = ½(eiθ - e-iθ). By substitution: , where c0=a0, for k>0 ck=½(ak-ibk), and for k<0 ck=½(ak+ibk).

Positive frequencies correspond to anticlockwise rotating vectors and negative frequencies to clockwise rotating vectors. The coefficients ck can be obtained from the equation: , where T is the fundamental period of the signal; T = 2π/ω.

General (non-periodic) signals may be represented in the form where in angular frequency; or , in frequency. H(ω) (or sometimes G(f) ) is called the Fourier transform of h(t).

We note that this complex function model for audio signals is only valid as a vector space; the linear combinations of signals perserve the model but products do not.

The fourier transform H(ω) is the angular frequency domain model for the signal h(t) as the domain is the angular frequency and value is the energy present at that frequency. The function G(f) is the frequency domain model for the signal h(t).

 

 

Graphical rule
Please send comments and suggestions to webmaster@theSongEditor.com