Sound Design with Everyday Words

Musicians often describe the quality of musical sounds with words such as "bright" or "warm". Our project investigates the relationship between auditory perception and language in this context: we are interested in finding whether people use a common terminology to describe timbre or if their choice of words is linked to their musical or cultural background. We deployed an online survey in which over a 1000 participants were asked to find words to describe the sounds they heard. We are now analyzing whether the words they used correlate with timbral features. Our objective is to design an audio processing engine that can automatically tag sounds in a database for retrieval purposes. It could also synthesize or modify sounds according to descriptive words instead of technical parameters.

Project 1.1: Data Analysis and Machine Learning

Objectives:
- Structure and parse database records
- Extract audio features from sound files
- Apply statistical techniques and pattern recognition algorithms
- Analyze and discuss results
- Future work: evaluate appropriate sound synthesis techniques; design and develop a sound synthesizer based on the previous findings

Pre-requisites:
- PHP, mySQL (or ability to learn them quickly on the job)
- Familiarity with Sound Synthesis, Digital Signal Processing, and/or Machine Learning techniques are a plus

This work will result in the publication of a paper.

Real-Time Network Music Performance

Thanks to the Internet, musicians located in different countries can now aspire to play together in real-time as if they were in the same space. However, the time delays due to the inherent latency in computer networks are unsuitable for real-time musical applications. To address this, we propose an online musical collaboration system that combines machine listening and machine learning to represent, transmit, and generate music over computer networks. Trained for a particular instrument, here the Indian tabla drum, our system recognizes drum strokes at one end and sends hierarchical symbolic structures over the network. At the receiving end, our system predicts the next strokes by analyzing previous phrases, and synthesizes an estimated audio output.

Project 2.1: Visualizations

Objectives:
- Develop visualization algorithms and interfaces
- Design "aesthetically-pleasing" visualization techniques based on an incoming stream of audio data and network packets
- Include remote video feeds (like teleconferencing)

Pre-requisites:
- GUI toolkit (e.g. GTK), graphical programming language (e.g. OpenGL, Processing), or graphics package (e.g. Jitter)
- Familiarity with Digital Signal Processing and/or Networking are a plus

Project 2.2: System Integration

Objectives:
- Design a GUI
- Build testing capability into the software package
- Develop software modules
- Test and setup the hardware and software system
- Design and setup the audio and lighting system for remote performances

Pre-requisites:
- GUI package (e.g. FLTK, GTK, Cocoa)
- Good programming skills in a high-level language (e.g. C, C++, Python)
- Background in Digital Signal Processing is a plus
- Familiarity with the tabla is a plus

This work (positions 2.1 and 2.2) will result in a public demonstration of the system during an official Media Lab event.

Please include a resume (or a description of skills and interests), a paragraph describing your motivation and relevant experience (classes or projects), and the position you are interested in.
Examples of previous work are necessary for the visualization project.

All positions require a strong interest in music technology, a high-degree of initiative and self-motivation. Positions start in May or June, and may be continued over the Fall term (for pay or credit).

Note that UROP positions are open ONLY to current MIT or Wellesley undergraduates. For more information, please refer to the UROP website.