First time here? Check out the FAQ!
x

Dictionary based methods

0 votes
303 views
I've been enjoying Curtis Roads' book "Composing Electronic Music" recently and been reading about DBM for matching input sounds to grains from a dictionary based on one or more factors (e.g. frequency, spectral centroid etc.) and it's obviously got me thinking of how to do it in Kyma.

Having asked in the Kata group the consensus was that any real-time process would need the ability to write a value to a particular step in a memory writer/wave table at sample rate (the reverse of a wave shaper). This is not possible in Kyma. Can this be made possible and/or any tips for how a DBM might be created?

i was thinking about something like a live spectral analysis where the timeindex values driving the sample/input are captured somehow to keep a record of when conditions are met so they can be looked up later. But I'm 100% sure that would fall down when I come to actually try.

 Thanks
asked Jan 22, 2020 in Using Kyma by ghood (Master) (3,060 points)
"i was thinking about something like a live spectral analysis where the timeindex values driving the sample/input are captured somehow to keep a record of when conditions are met so they can be looked up later."
Perhaps you could try some initial experiments by capturing presets at the time a condition is met. Then use an InterpolatePresets where the Preset is your index (dictionary key). This may not be what you want as the ultimate solution, but it would give you a sense for what input (key) values should correspond to which outputs (values).
Thanks. I’ll give it a go and report back.

Please log in or register to answer this question.

...