Comparing performances of Beethoven symphonies—with machine learning

“A Duke University team has developed a machine learning algorithm that ‘listens’ to multiple performances of the same piece and can tell the difference between, say, the Berlin Philharmonic and the London Symphony Orchestra,” writes Robin Smith in Monday’s (3/22) Duke Today, published by the university. “In a study published in a recent issue of the journal Annals of Applied Statistics, the team set the algorithm loose on all nine Beethoven symphonies as performed by 10 different orchestras over nearly eight decades, from a 1939 recording of the NBC Symphony Orchestra conducted by Arturo Toscanini, to Simon Rattle’s version with the Berlin Philharmonic in 2016…. Ph.D. student Anna Yanchenko and statistical science professor Peter Hoff converted each audio file into plots, called spectrograms and chromatograms…. After aligning the plots, they calculated the timbre, tempo and volume changes for each movement, using new statistical methods they developed to look for consistent differences and similarities among orchestras in their playing.… Yanchenko [is] a longtime concert-goer at the Boston Symphony Orchestra in her home state of Massachusetts. But she says her work helps her compare performance styles on a much larger scale than would be possible by ear alone.”