Hacker News Reader

I wonder, does it know how to transpose? People may end up humming tunes in different keys that their voices can more easily reach.

Also, why the hell do I need an app for this? Is it actually running a local model? Because my guess is that it's just hitting an API.

View on HN

I wondered whether pitch might play a role. EDIT: singing in a higher key got me English results, so I guess so?

There's no way it would do much locally, but maybe they just wanted to make sure the audio passed to the API has a certain sample rate and encoding?

Very unlikely that they are considering the actual sung or hummed pitch as very few people, including professional musicians, would start singing at the correct pitch without accompaniment.

Most likely they are mapping the interval between the sung notes and using that as part of the ‘melodic fingerprint’ for matching.

Reply on news.ycombinator.com