The temperament police

Keyboard temperaments have been extensively discussed in historical treatises as well as more recent scholarly texts. The work presented in this article aims to provide a unique perspective on the subject, by applying signal processing techniques for estimating the temperaments used in harpsichord solo recordings, to show what is done in practice, thus facilitating a descriptive approach to tuning and temperament. Motivations, previous work, current methods and results obtained from a dataset which currently includes over 2,000 tracks from over 90 harpsichord solo CDs are discussed, together with some of the implications of these results, and the directions this work may take in the future.This work is supported by the AHRC ‘Digital Music Lab—Analysing Big Music Data’ project, grant no. AH/L01016X/1. Emmanouil Benetos is supported by a City University London Research Fellowship. The authors are very grateful to all those who helped collect the set of recordings used for this work