The Energy University
Department of Electrical & Electronics Engineering
College of Engineering
EEEB363 / EEEB3024 Digital Signal Processing
Semester 2, Academic Year 2022/2023
Group Project – What’s that sound?
Group Project: You must attempt this project in a group of 2-3 people.
Evaluation:
Marks are based on the report (20%) and demonstration (80%). Marks are divided as follows:
Report (20%) Demonstration (80%)
Criteria Marks (20%) Criteria Marks (80%)
Appearance and Formatting 5 Familiarity on tool(s)
Language 5 - Able to implement DSP theory 20
- Able to demonstrate important
Content & Data presentation:
10 MATLAB functions 20
i.e. Figures, Tables, etc.
- Code neatness
Operation of tools(s)
- Able to generate output graphs 20
- Able to explain output graphs 20
Report due date:
The group project report is to be submitted by 12 May 2023 via Brighten.
Demonstration schedule:
You will need to demonstrate your project works. The demonstration is scheduled on week,
2 – 11 May 2023.
The problem:
An annoying noise is found in all these audio recordings. For each recording, there is a specific
sound that appears somewhat periodically. I have isolated this sound for your use.
Correlation can be used to measure the similarity between two signals. You can find if a signal
appears inside another signal and find out where it appears. Use cross-correlation to discover
where the noise are inside the audio recording.
Filtering in the frequency domain is a useful method to reduce or remove unwanted noise from
audio signals. Use frequency domain filtering methods to improve an audio signal that is
provided by the lecturer
Your task:
1. Select only one audio file and its respective noise signal.
2. You are free to use any MATLAB technique to accomplish your task, as best you can.
3. Apply cross-correlation to detect when the noise occurs on the time-scale.
4. You must design at least one FIR or IIR filter to be applied, using the step-by-step
techniques learnt in class.
5. You may apply more than one filter.
6. Demonstrate your correlation and filtering system on your chosen audio file.
7. Marks are given on effort. Show your work. Do your best.
What to report on:
Generally: introduction, methodology/design, analysis, discussions, conclusion, references.
Specifically:
1. Show and explain time-domain cross-correlation lag plots.
2. Show and explain frequency domain plots of the audio before and after filtering.
3. Compare the differences between FIR and IIR filters. Then present and discuss in detail
your design for the FIR or IIR filter you used to accomplish the task above.
4. Explain your MATLAB approach and explain some important excerpts of your codes.
5. Include your complete m-file script as an appendix to the report.
6. Report format: IEEE manuscript template.
[Link]
Tips:
sound(x,Ft) Convert matrix of signal data to sound
clear sound Stops the audio from playing
Audio Files (choose only one set)
Set Description Files
1 She keeps her Moet et Chandon. In her pretty cabinet. • killer_queen.mp3
"Let them eat cake", she says, just like Marie • killer_queen_noise.mp3
Antoinette. A built-in remedy for Khrushchev and
Kennedy. At any time an invitation you can't decline.
Is that a Jojo reference?
Piano arrangement by OnlinePianist
2 The Mandalorian is a Star Wars story set after the fall • mando.mp3
of the Galactic Empire. We follow the travels of a lone • mando_noise.mp3
gunfighter across the galaxy, dragging along a 50 year
old green force sensitive baby.
Piano arrangement by Drew Collier
3 Great Fairy Fountains can be found throughout Hyrule • fairyfountain.mp3
Kingdom. After Link provides the Great Fairy with • fairyfountain_noise.mp3
enough Rupees, a large flower bud will open, revealing
a small pool of water in its centre which the Great Fairy
will emerge from, then Link’s armour receives an
upgrade.
Piano arrangement by Hari Sivan
All the best!