Signal enhancement describes attempts at approximating original musical signals from non-ideal recordings, In a novel processing method, a digitized audio signal is split into phase-linear, complementary narrow-band signals for individual processing prior to being re-emerged to form the enhanced output signal. A simple model indicates which SNR improvement may be achieved in terms of the original signal quality. The behavior of some signal enhancement algorithms is demonstrated both graphically and via digital recordings of digitally processed analog originals.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.