On the use of the Dual Resonance Non Linear filter for speech processing in a hearing aid algorithm