This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: ↑ L ↓ M estimat e of B , the number of bits per sample sampling rat e reduction sampling rat e rest oration estimat e of B , the number of bits per sample delay Spectral Analysis and Sample Rate Selection Overview This project is a follow up on the Filter Design for Sample Rate Conversion project. For this the project, we focus less on classes of filters and more on the selection of an appropriate sampling rate. Our goal is to design a very simple audio compression system. Audio that is originally sampled at 44.1 kHz will be processed to reduce the sampling rate to R Hz. Then, the samples at the lower sampling rate will be quantized to B bits per sample for a compressed bit rate of RB bits per second. Since many intricacies of quantization are beyond the scope of this course, we’ll use some simple rules to give a correspondence between the step size of a uniform quantizer ∆ and the number of bits per sample B . The overall compression system is then as depicted below: Part A – Spectral Analysis Sampling rate reduction usually does not preserve the information in a signal. In the case of using ideal filters, we would lose the information in the signal above R /2 Hz. In this section of the project, you must use spectral analysis techniques to estimate the frequency spectrum of six files which are available on the course website. (Use the function wavread to input the files into M ATLAB .) For each spectrum estimation technique, produce a single spectrum from the file. Keep in mind that the purpose of the spectral analysis is to be able to predict the power of the signal x[n] that will later be lost in the sampling rate reduction Ch. 7 s2005c process. Another important aspect of spectral estimation is that computational overhead can be reduced (at the expense of making the estimate less accurate). Try to use parameters which you find make the computation practical....
View
Full Document
 Fall '05
 HEMAMI
 Digital Signal Processing, Signal Processing, sampling rate, Spectral Analysis, fil ters

Click to edit the document details