echonest

When you say the matching is far from perfect, is that at your end or on the part of the echoprint / echonest code?

Really nice performance too - for reference Echonest advises samples be 30s or longer for optimum recognition.

The echonest remix API has similar capabilities, and can quantize into granularities of measures, beats, tatums.

The actual identification of individual songs is done using the Echonest API (they have a huge database of song hashes).

I created this site using the Spotify API and EchoNest API for music information and song, data respectively.

I was thinking about this too but read before that the echonest fingerprinting doesn't handle pitch changes like Shazam.

The hashing algorithm is provided by Echonest in their Echoprint service and used to compare against their database (hopefully growing after their Spotify acquisition).

Which renders the usual recommendations engines hardly useful to me - it seems everyone's coming up with the same suggestions (or they all use EchoNest).

EchoNest used to do this[1] and is open source 1.

EchoNest, and other recommendation approaches, convolutional neural nets, etc.

I think I've read on the echonest board that the most common solution is to index multiple pitch variants of the same songs.

Echonest can work with voice but is optimized for music so you might encounter a lot of false positive with it.

By the way, in the article, they avoid using external data, but for my project, I am trying to use AcousticBrainz data, and also explored Echonest's.