I’m starting to get a sense that the Spotify on Echo experience has some bugs. Specifically, the Echo will announce that the song is playing and then you’re greeted with nothing but silence. Asking for the song again seems to jostle loose whatever is going on and the song starts to play. This has happened enough times (maybe 20% of requests from Spotify) that I’m starting to feel it’s the norm.
I’m curious if some of the first gen Amazon Alexa devices would pass Amazon’s AVS certification on functional tests. These are the tests that Amazon requires third party hardware makers to pass to be able to market their devices as Alexa enabled. The issue with Spotify playback would likely be flagged as an issue.
On the topic of music playback, I’m curious how Alexa picks form multiple results. An example was a request for Y.M.C.A.. Alexa brought up a result for Y.M.C.A. by Old Skool Disco Masters. I had to specifically ask for the Village People to get the right song. How popular is the former version?
The same was for Macarena. The results was this:
Not bad, but definitely not what I was looking for.
Voice based music playback is hard. There are a lot of moving parts and many players. Many things can go wrong. When it works, it’s magical.
When we had Grooveshark and SoundCloud integration with the Ubi, it was a great experience but with SoundCloud, the results through its API were a mixed bag. You could get Thriller by Michael Jackson or get thrilled by some wacky version.
For Alexa and all voice-based music, there’s still a lot of work ahead.