Many of you may have heard of, or seen, Israeli musician Kutiman’s Thru-you video mash-up from 2009. What Kutiman – also known as Ophir Kutiel – did was go through thousands of Youtube videos of people singing or playing instruments and find those with the same rhythm, chords and melodies, to eventually “mash” them up. Since then he has continued with his Youtube creations and has even had a show at the Guggenheim and partnered with popular U.S. band Maroon 5.
Now a group of brainy students from the Technion, Israel’s Institute of Technology, based in Haifa, decided there must be an easier, or more mathematical way to mash songs up.
That’s how they thought of Absolute Pitch- an algorithm that detects different aspects of a music piece, in mathematical-digital ways, and identifies which songs will combine a good mash-up. The algorithm scans the data from a music file and can determine in what pitch the songs is heard, what chords are used and on which music instruments. Kutiman put a couple of months’ work into his piece- and now it can be done in seconds.
“Currently, the data available on any song in the web is the data that users wrote about it and not the data of the music itself (chords, pitch, instruments etc). We wanted to create a search engine that can detect rhythms, musical harmonies, etc by itself, without the assistance of human tagging” says Gidi Naveh, the project’s guide and founder who worked alongside Kutiman. He adds that the new algorithm can “provide musicians a new way of finding relevant music pieces by entering a musical piece like a whistle, a vocal chorus or a violin.”
In order to check the algorithm’s efficiency and accuracy, the group took Kutiman’s mash-up and entered it into the program and found a 60% match between Kutiman’s choices and the algorithm’s choices. Excited about the new findings, Naveh, along with Ofir Lindlebaum and Shay Maskit, who came up with the original idea, created their own mash-up . “We asked Kutiman to review the result, and he approved that the piece musically. The innovation in this project is using mathematical tools to analyze music” said Naveh.
The algorithm is still considered an academic research project, and will take a while to become publicly available. Meanwhile, Kutiman had the algorithm installed on his computer. “He was very happy with the results. We’re curious to hear the work he’ll make with it” says Nave.