Bowen Dong is a recent graduate of the Master of Science in Music Technology program. His studies examine how synthesized voices are used in comparison to real voices in pop music.

"Vocaloid is voice synthesis software for electronic music production," Dong said. "Since 2007, an online subculture community has developed around Vocaloid music, which is heavily influenced by J-Pop but not identical. I've found that synthesized voices use similar pitches and note density to human voices, but tend to use larger intervals at higher beats-per-minute than humans."

"One of the most interesting things is because of the long period of the study, I can quantify how a singer's or band's style has changed over time."

Dong's undergraduate degree is in computer science. "But I have many years experience of music, and I wanted to study music after I graduated. I came to music technology and this program as a bridge from computer science to the music industry."

Yiwei Ding, a second-year master's student in the Music Technology program, is working with the Music Informatics Group using machine learning algorithms to analyze audio, especially music. "What I'm doing as music classification, for example, given audio of maybe 30 seconds long identify the genre and also automatically assigning the music tags, such as what instruments are playing."

"Machine learning is based on a huge amount of data, and with a huge amount of data we need very complex models," Ding said. "But these are very slow, especially if you want to run something like this on a phone. I'm trying to do model compression to make the model smaller, but perform as well as large models."

"We can use large models that have already been trained to guide the learning of small models, that can be used in more scenarios."

"I came from a machine learning background, but I'm really very interested in music," Ding said. "You can do machine learning at many computer science programs, but one thing special about Georgia Tech is that you can meet many musicians here and learn about music in the program, not just engineering and technology."

Levi Waterhouse is a second-year Bachelor of Science in Music Technology student minoring in Industrial Design. He is also a drummer for the Georgia Tech Pep Band, the outreach officer for the GT Table Tennis Association, and a past member of the marching band.

Waterhouse's interest in Music Technology is adaptability and accessibility in music. "Whether it be performance or overall experience, I have found limited areas of adjustment for people with physical disabilities in this industry," he said. "I hope to find ways through my major and minor to counteract this with advocacy and invention, although I don't yet know where that will take me."

"Georgia Tech was the perfect choice for this focus, as the various STEM and industry concepts standard in musical environments could be accessed with ease at GT (in comparison to more music-centered universities)."

Ash Vinay is teaching machines to hear music the way humans do. Vinay's primary goal has always been to build a better synthesizer of sounds. "But the numbers our algorithms produce don't necessarily align to human perception," he said. "That disconnect is where it's interesting."

"If we want our machines to behave more human, they need to hear and produce sound in a way humans find interesting and not annoying."

Music technology allows musicians to explore music through a vast network of technical fields, according to Vinay. "In this program, we have cognitive psychology, digital signal processing, programming, machine learning, musicology, biofeedback, and there's a lot of science that goes into all these things," he said.

Jiaying Li is figuring out how to inject new blood into pop music trends. A Master's student, Li identified trends in how Chinese pop music reuses chord progressions, making songs sound increasingly similar. She is currently working on a project to convert one singer's voice within a song to another's voice.

"I want to make it easier for everyone to create good songs," Li said. "I played piano as a child until my hands were injured and I couldn't play any more. But music isn't just performance, it's also production. I learned about Music Technology, connecting music with science, techniques, and technology."

"I love this program because it's more technical. A lot of our goal is to find a more technical way to set up music. When we talk about Japanese music and Chinese music and Western music, we find all the styles are different. But if we talk about music technology, there is no difference across the nations in the engineering and scientific principles."

Rhythm Jain, who is a musician and a scientist, comes from a musical family that gave her the perfect name for the work she is doing now, which is all about music.

“I became fascinated by this notion that each country expressed its patriotism in different ways melodically, and how the people from different countries perceived the music differently,” Jain said. Her interest in perception and cognition of music brought her to Georgia Tech, where some of her research while pursuing a Master of Music Technology was focused on studying ragas, the melodic framework in Hindustani classical music.

Then she shifted gears to music transcription, developing algorithms to transcribe and translate the music of oral traditions. Now Jain, who has taught herself the guitar and is dabbling with electronic music and synthesizers and she is using the knowledge and skills she developed at the Georgia Tech School of Music to build music recommendation algorithms for Sirius XM/Pandora.