digital signal filter

The quadrature mirror filters (QMF) are two filters with frequency characteristics symmetric about $1/4$ of sampling frequency (i.e. $\pi /2$ ). They are used especially in process of orthogonal discrete wavelet transform design.

## Simple variant

In notation of Z-transform, we can create the quadrature mirror filter $H_{1}(z)$  to (original) filter $H_{0}(z)$  by substitution $z$  with $-z$  in the transfer function of $H_{0}(z)$ .

$H_{1}(z)=H_{0}(-z)\,$

By doing it, the transfer characteristic of $H_{1}(z)$  is shifted to $H_{0}(z)$  by $\pi$ .

$|H_{1}(e^{j\omega })|=|H_{0}(e^{j(\pi -\omega )})|\,$

Impulse characteristic is therefore

$h_{1}[n]=(-1)^{n}h_{0}[n]\,$  for $0\leq n , where $N$  is filter length.

According to the picture above, the signal split and passed into these filters can be downsampled by a factor of two. Nevertheless, original signal can be still reconstructed by using reconstruction filters $G_{0}(z)$  and $G_{1}(z)$ . Reconstruction filters are given by time reversal analysis filters.

$G_{0}(z)=H_{0}(z^{-1})\,$
$G_{1}(z)=H_{1}(z^{-1})\,$

## Orthogonal filter banks

For orthogonal discrete wavelet transform $H_{1}(z)$  is given by

$H_{1}(z)=z^{-N}H_{0}(-z^{-1})\,$ , where $N$  is filter length.

Impulse characteristic is

$h_{1}[n]=(-1)^{n}h_{0}[N-1-n]\,$  for $0\leq n .

Reconstruction filters are still given by same equations.

$G_{0}(z)=H_{0}(z^{-1})\,$
$G_{1}(z)=H_{1}(z^{-1})\,$