By June Kinoshita, Executive Director, New Harmony Line
It seems like every education and arts conference this year is talking about Artificial Intelligence. Can generative AI be creative? Will it enhance or destroy creative industries, not to mention all kinds of jobs? One popular meme going around on social media says “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” Amen to that.
In fact, Hyperscore is an example of a software that uses AI to do the more tedious chores, freeing people to explore the fun, creative, and essential parts of composing. Centering human expression and creativity in technology design is the core principle of Hyperscore’s inventors at the Opera of the Future Group at the M.I.T. Media Lab. We believe it is becoming ever more urgent to uphold these values as societal and economic pressures threaten to push AI in the opposite direction.
A deeper dive
One of Hyperscore’s advanced features is the “harmony line,” the dark gray line that is centered on the horizontal axis of the Sketch Window. This line can be manipulated to create regions of harmonic tension, release, and modulation. The Hyperscore composer can play with the harmony line and decide whether the results sound the way they want them to. A more advanced composer can do this more intentionally, and of course they can always specify the exact notes they want in the Melody Window. Hyperscore allows a user to be fully in control of every note, or to collaborate playfully with its algorithms to discover and shape the elements of their composition.
Hyperscore’s automated harmony algorithm was originally implemented using AI trained on classical music. You can read the technical details in this article (Farbood, M., Kaufman, H., Jennings., K. “Composing with Hyperscore: An Intuitive Interface for Visualizing Musical Structure” International Computer Music Conference, 2007.) To quote:
“In Hyperscore, users describe harmonic progressions by shaping the harmony line. It is parsed into sections which are then mapped to functional identifiers…which consist of four categories: Statement, Antecedent, Consequent, and Modulation. The harmony line, which runs through the center of each sketch window, can be modified by clicking and dragging. Colored bands appear to indicate the line’s parsing. Sections are classified as one of four visual types, each corresponding to a functional identifier:
• Statement – flat section, colored white. Musically defined as a statement or prolongation of the tonic.
• Antecedent – upward-sloping section, colored green. Musically defined as a combination of chords that need resolution (e.g. dominant chords or combinations of subdominant and dominant chords).
• Consequent – downward-sloping section, colored blue. Resolution of preceding Antecedent section. If not preceded by an Antecedent, then restates the tonic.
• Modulation – defined by a sharp pointed region or spike, colored yellow. Progression toward a new key.
After the line is parsed, chords are assigned to each section based on its functional identifier, how many beats it spans, and how textured or “bumpy” the section is. The instability of the chords assigned to a section is directly proportional to the amount of texture. The chords chosen are taken from a database that returns either single chords or small progressions based on the selection criteria. The database consists of chord progressions commonly found in Bach chorales.”
Explore Hyperscore
If you’re intrigued and want to play around with Hyperscore, you can sign up for a free version (with limited functionality) or a subscription version (starting at a very affordable $3.99 per month). You can also try out the full classroom version by signing up for the Hyperscore Challenge, which runs through June 21.
To help you get started, we have Quick Tips that walk you through all of the features of Hyperscore. We also invite you to attend our free Second Saturday Zoom workshop where you can participate in composing a new Hyperscore piece in real time.
Image credit: “Better Sharing With AI” by Creative Commons was generated by the DALL-E 2 AI platform with the text prompt “A surrealist painting in the style of Salvador Dali of a robot giving a gift to a person playing a cello.” CC dedicates any rights it holds to the image to the public domain via CC0.
