ECE 403
Audio Engineering

Section Type Times Days Location Instructor
E DIS 1230 - 1350 T R   260 Everitt Lab  Paris Smaragdis
Web Page http://hear.beckman.illinois.edu/wiki/Courses/ECE403-2012AudioEngineering
Official Description Resonance and wave phenomena; acoustics of rooms and auditoriums; artificial reverberation and sound localization-spatialization; loudspeakers, enclosures, and microphones; topics in digital audio. Course Information: Prerequisite: ECE 290, ECE 310, and ECE 473.
Subject Area Biomedical Imaging, Bioengineering, and Acoustics
Course Prerequisites Credit in ECE 310
Credit in ECE 473
Credit in ECE 290
Course Directors Jont Allen
Detailed Description and Outline

To obtain an understanding of acoustics and signal processing fundamentals as they apply to the audio field. To facilitate an ability to read articles at the level of the Journal of the Audio Engineering Society. To contribute to a background useful for a position in the audio industry.

Topics:

  • Acoustics review and signal processing review
  • Acoustic transducers
  • Noise, intensity, and time-varying signals
  • Auditory psychophysics
  • Digital audio coding
  • Room/auditorium acoustics
  • Sound localization/spatialization
  • Speech and audio recognition
Topical Prerequisities
  • Passive circuit steady-state and transient analysis
  • Acoustic plane and spherical wave phenomena
  • Fourier and Z transform
  • IIR and FIR filter design
Texts
L. E. Kinsler et al., Fundamentals of Acoustics, 4th ed., John Wiley & Sons, 2000.
Course notes and various published articles.
ABET Category
Engineering Science: 3 credits or 100%
Course Goals

Goals

  • Audio Engineering is the practical integration of Electroacoustics, Psychoacoustics, and Signal Processing for the purpose of creating really cool stuff.
  • Electroacoustics: The study of speakers, mics, and other things that make electrical signals acoustic and vice versa.
  • Psychoacoustics: The study of pitch,loudness, timbre, masking, and other perceptual qualities of a sound.
  • Signal Processing: The study of filters and spectral representations used for signal description, interpre­tation, or compression.
  • After learning about the core theoretical areas, we will study one or more of the following application areas. Usually the class votes to see which areas we will study.
  • Room Acoustics: The art of measuring, modeling, and simulating the presence, early echoes and rever­beration of a room.
  • Speech/Music/FX Recognition and Unmixing: The art of analyzing a signal to figureout(1) how many audio sources are present, and(2) what they were(which word was spoken, which speaker spoke, what was the tempo or melody of a musical piece, whether a recording was made outdoors or indoors, etcetera.)
  • 3D Audio Spatialization: The art of reproducing sounds at specified apparent locations around the listener.
  • Digital Audio Coding: The art of coding a signal with the smallest possible number of bits compatible with perceptually excellent audio quality.
  • Film Audio: The art of choosing microphones,recording audio for filmandtelevision,mixing,and creating sound effects.
  • Pre-requisites are digital signal processing(ECE310) and acoustics(ECE373). The level is advanced undergraduate or introductory graduate.
Instructional Objectives

By the time of the midterm exam, students should be able to do the following:

1. Write the acoustic wave equation and its solutions, in both time domain and phasor representation, for a plane wave and for a spherically symmetric wave (a,m,n,e).

2. Compute the frequency response and impulse response of a duct, or of a rectangular room (a,k).

3. Given a matrix frequency response representing a two-port system, calculate the one-port frequency response that results from constraints on any two of the inputs (a,b,e,m).

4. Compute the magnitude and phase response of a speaker/microphone pair given tonepip measurements at different frequencies (a,b,c).

5. Analyze an equivalent circuit to calculate the magnitude and phase response of a moving armature loudspeaker, electrostatic loudspeaker, electrodynamic microphone, or condenser microphone (a,b,d,g,m).

6. Create a desired microphone directivity pattern by scaling and adding omnidirectional and figure-eight patterns (a,b,e).

7. Calculate the perceptual loudness of a sound made up of spectrally distinct tones and noise bands (b,c,k,m).

8. Determine whether or not one simple sound (tone or noiseband) will perceptually mask another (b,e,h,m).

9. Estimate the pitch of a speech or musical signal using autocorrelation (a,b,e).

10. Use the correlogram method to estimate the pitches of two bandpass-filtered harmonic signals played simultaneously (a).


By the time of the final exam, students should be able to do all of the above, plus the following:

1. If we study room acoustics: use the image source method to calculate the impulse response and reverberation time of a rectangular room. Students should also be able to calculate reverberation time using Sabine’s formula,given the geometry of the room, a description of the wall coverings, and a table of material absorptivities.

2. If we study auditory scene analysis: use the principles of common pitch, common onset, and common offset to cluster energy bands into distinct auditory objects.

3. If we study recognition features: compute the linear-time-warped mel-frequency cepstral coefficient representations of three sounds. Determine which two sounds are similar and which one is different.

4. If we study 3D audio: use a simple anatomical model to compute the head-related transfer function from any given azimuth and elevation.

5. If we study digital audio coding: analyze an FFT to compute the signal-to-mask ratio in each band of a sub-band coder, and allocate bits in order to minimize noise audibility.

6. If we study pitch/time-scale modification, use sinusoidal transform coding(STC) to speed up or slow down an input signal without changing the pitch.

ECE 403 includes a small final project, intended to be roughly equivalent to a normal computer assignment. The final project should demonstrate that a student can (b,g,i,k):


1. Understand an article from the professional literature well enough to implement the algorithm it describes (a,e,g,i).


2. Compose a written report that demonstrates understanding of the theoretical motivation of the algo­rithm and the meaning of results (a,e,g,i).

Last updated: 5/23/2013