Abstract
In this article, advantages of audio signal separation into a tonal and a noise part for the purpose of linear predictive (LP) modeling will be studied. The LP modeling of the whole original audio signal is difficult. Hence, the article deals with the intention of LP modeling of noise part only. It will be studied whether the orders of LP filter for the LP modeling significantly differ for noise/tonal parts and its original audio signal or not. Firstly, a method for determination of an optimal LP order will be shortly described. Secondly, LP optimal orders for tonal/noise and its original tested audio signals will be measured. And finally, statistical analysis over the measurement data will be performed.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
