Unformatted text preview: type of filter algorithm you would use for improved computational efficiency. L / M = 960 / 441 = 320 /147 L = 320, M = 147 G = L = 320 ± c (ideal) = min( ² / L , / M ) = / 320 (interpolating filter dominates) p (practical) = 2 ³ 20 kHz L ³ 44.1 kHz = 0.00283 = .0089 s (practical) = 2 ³ 24.1 kHz M ³ 44.1 kHz = 0.00342 = .0107 For efficency, implement the LPF as a polyphase decimating filter to reduce computational load by 147! You could also implement it as a polyphase upsampling filter, with a computational savings of 320. For super efficiency the commutator can skip M samples at a time, and you only compute the interpolator polyphase filter outputs when needed. This is a computational savings of (147)(320)!!!! ± L ² M H ( z ). LPF corner at c x [ n ] y [ k ]...
View Full Document
- Winter '12
- Digital Signal Processing, Sony corporation, sample rate, LPF, ideal filter corner, practical filter implementation