Hi, I found out that Gen10 has an issue with dealing with odd harmonics (not fundamental). It's easy to repro simply taking the example code and using the following coefficients instead: g10.coefs( [0., 0., 0.5] ); - shows the issue g10.coefs( [0., 0.5, 0.] ); - doesn't show the issue Hooking ChucK up to a spectrum analyzer it's easy to see the difference. In case there are odd harmonics, there are a lot of unwanted partials/sidebands, instead of having only the components specified in .coefs(). Gen9 doesn't seem to have this issue. Cheers, Mario -- Electronic Musician, Creative Coder, QA Engineer https://vimeo.com/creativecodingsalerno http://mbuoninfante.tumblr.com/ https://github.com/mariobuoninfante https://bitbucket.org/mariobuoninfante/