g-roma / floop Goto Github PK

If I understand correctly, when a new bpm is selected, floop makes 1 new request to the Freesound API for each of the sounds previously calculated to be in that bpm. This can mean too many queries :) These queries are to get sound name, preview URL and a couple more metadata items. I'd suggest that:

• Either all this data is downloaded and stored locally in the app (just like descriptors for the graph)
• Number of queries is optimized by requesting info for multiple sounds at once. This can be done using the filter parameter in the Freesound API. You can do something like "search for sounds with id 1, id 3, id 1003, and id 54". The limit is basically the length of the query, you can probably safely request info for 50 sounds at once. Doing this, the number of queries would be reduced by 50... and floop load bpms much faster! :)