This is only experimental to see the maximum amount of data I could get for a chunk of audio and how long it takes. I really need strategies to then find the most valueable features and stats. Before in JAVA I only had a mean stat for every feature and it was already working quite good. But I want to explore the usefulness of the other statistic methods.
Btw: @ blocking 1 mode is a bit faster but still can’t keep up with the js.
I’ll check fluid.dataset~ but I don’t think that it fits.
Thanks