Publication Date:
2008
abstract:
The audio treatment methods commonly used in many FM radios and in mastering processes for today's popular music tend to increase perceived loudness of the final audio product, producing loud but 'flat' sound. This is satisfactory for most commercial purposes, ut ultimately introduces unwanted effects such as reduced dynamic range, poor transients, and loss of fidelity. We have proposed a method for time-domain transient restoration of recorded music that restores clarity similar to that of a live performance. The algorithm can easily be implemented with low-cost digital signal processors, i.e. in stand-alone devices or included in (iPod-like) hardware or (Winamp-like) software audio-player applications. A method for using on-the-fly feature extraction from the songs played to automatically tune the effect is currently under study. This goal is being jointly pursued with a research group from the Technische Universität Wien working on music classification and retrieval, within the framework of the EU MUSCLE-NoE project
Iris type:
02.01 Contributo in volume (Capitolo o Saggio)
Keywords:
Music computing; Digital embedded system
List of contributors:
Bertini, Graziano; Magrini, Massimo
Book title:
DSP Application Day 2007