Harmonics Separation
Hi,
I will be investigating ways to isolate individual harmonics from a mic'd trumpet player, who will be sustaining a note. The idea is to be able to capture this audio in max and somehow separate the individual harmonics present in the sound, and then hopefully map them individually to a midi controller pad, so we can demonstrate how the trumpet sound/note being played is actually composed of individual sounds.
Has anybody tried anything similar, or can see any obvious flaws with my idea?
Thanks in advance,
Benjamin
have a look at the zsa descriptors package- i think the zsa.freqpeak external might be of use
you might want to check out SPEAR first.. it's not realtime, but it's very suitable for such demonstration.
Yes, Spear is great for what you´re looking for, but not in realtime.
There is a real-time peak detector in my library called [sadam.peakExtractor~]. You can access it on the ZiMT website as part of the Klangpilot 3 environment (http://www.mdw.ac.at/zimt/downloads.html). Unfortunately it is not documented yet, but as the attributes can be seen in the object inspector, with some experimenting it should be quite straightforward.
Hi,
You wish to decompose the sound into its harmonics?
I was thinking something like using a pitch detector to recognize the fundamental pitch, then use this note (frequency) to calculate the harmonic series for this pitch, and use this to configure automatically a series of filters (for example reson~ inside polys).
Tell me what you think.
best wishes
also check out the sigmund~ object (which has a nice resynthesis option as well)