Hello everyone,
I am quite sure that I while ago I found a library/plugin for csound that had similar functionality to flucoma in terms of machine learning and audio analysis. At the time I did not have time to dig into it and now I can’t seem to find it again. I remember that I found it while reading about the CSound 7 release. I scoured the mailing list and the ICSC 2024 proceedings, but couldn’t find it again.
Does anyone here know what I am talking about?
I’m afraid not. At the last Csound conference in Vienna, Pierre Alexandre Tremblay did wonder if it was time for a Csound port, but I heard nothing about any work done on it. I know there are bindings for Pure Data, you could use those and the csound~ object perhaps?
Are there any plugins focused on machine learing in CSound? If there are, it might have been one of those, even if it didn’t have all the funcionality of Flucoma.
I don’t recall anything being spoken about this. If I was looking to investigate this area I would probably use an established ML library like tensorflow and tap into Csound via the API.
What are some of the ML applications for audio processing that are interesting?
I’ve no idea but I assume tensorflow can be used for musical applications too.
…a quick search later:
I’m mostly interested in realtime regression based on audio input (the mlpregressor~ in flucoma) and realtime timbre transfer (for example rave by ircam and ddsp by google). The realtime application of this is also what makes it hard to just use python for this.
I’m afraid I don’t know enough about it, but would be nice to have a Csound interface to flucoma from what I’ve seen.
