Audio and Speech Coding Systems

Lectures
1. Introduction
2. Signal Processing Essentials
3. Rate Distortion Theory
     3.1 Uniform Quantization
     3.2 Non-Uniform Quantization
     3.3 Vector Quantization
4. Entropy Coding
    4.1 Huffman Coding
    4.2 Arithmetic Coding
5. Filterbanks I
   5.1 Perfect Reconstruction and Critical Sampling
   5.2 Quadrature Mirror Filterbanks
   5.3 Conjugate Mirror Filterbanks
   5.4 Polyphase Quadrature Mirror Fitlerbanks
   5.5 The MPEG filterbank
6. Filterbanks II
   6.1 DFT and FFT 
   6.2  Overlap and Add
   6.3 DCT
   6.4 MDCT
7. Psychoacoustic Models for Audio Coding
8. Speech Coding

Labs
1. Quantization and Dither
2. Entropy Coding
3. Filterbanks
4. Psychoacoustic Models
5. Evaluation of the Audio Coding Challenge

Seminars
1. Rate Distoriton Theory
2. Quantization
3. Entropy Coding
4. Psychoacoustics and Bit Allocation
5. Filterbanks and Bit Allocation
6. Quality of Audio Coders
7. Speech coding


Project: Audio Coding Challenge
      5-6 students create a group
      The group decides a name for their Perceptual Audio Coder
      Each group works together to create a modified Perceptual Audio 
      Coder
            The goal is to achieve minimum BitRate and maximum Quality
            The reference code is the MTG perceptual Audio Coder
             Possibilities (For example):
                        Use Vector Quantization instead of Scalar Quantization
                        Modify the number of Filters
                        Use an adaptive number of filters
                        Use a different Psychoacoustic Model
                        In Lab5 we will perform a evaluation where each Audio Coder of each
                          group will be compared against the coder of the other groups
      Objective evaluation: For a set of audio files the rate and the distortion has to be given
      Subjective Evaluation: Listening tests will be done by the students