432 vs 440 Hz
Henric Mattsson, one of the technicians at Salford Acoustics asked me about a set of YouTube videos which have been re-tuned so the music is at a lower frequency. Here an example for Sadie’s Cherish the Day, one is the original, the other has been flattened.
Normally musicians tune to 440Hz, but the second video has been altered (presumably using a pitch shifter) so the frequency of the tuning A is lowered by 8 Hz to 432Hz. The YouTube description for these flattened videos make some curious claims:
“Listen to the how the opening guitar part comes at you compared to the original recording … The sound feels very centered to me and it has a relaxed, spacious, organic “real” quality … Every single note of the C-major scale becomes a whole frequency, unlike the fractions of frequencies it is with A-440.”
Has anyone tested this scientifically? Well there is one acoustician, Hugo Fastl, who has compared classical music tuned at 432 Hz to music at 440 Hz using proper psychoacoustic methods. He presented the work at the recent ICA conference in Montreal. [1]
Much classical music was composed with a lower tuning frequency of 432 Hz, so maybe old music sounds better at a lower frequency? This is what Fastl tested. He used a Welte-Steinway grand piano driven by punched paper which had captured the performances of famous pianists such as Claude Debussey. The piano was tuned to 440 Hz and recordings made. And then re-tuned to 432 Hz and new recordings made. Listening tests then got subjects to compare and rate the recordings. No preference was found for 432 Hz, if anything there was a slight preference for 440 Hz.
This could just be a learnt preference for 440 Hz, after all that is what we’re used to hearing nowadays. If it a learnt preference, then the original Sadie video should sound better than the flattened on. Which video did you prefer, the original or 432 Hz? Please comment below
Notes
[1] Basics and applications of psychoacoustics POMA Volume 19, pp. 032002 (June 2013); (23 pages) Hugo Fastl
The Sound Blog 