[Csnd] RT MIDI Phase Vocoding
| Date | 2014-04-15 14:24 |
| From | hlolli |
| Subject | [Csnd] RT MIDI Phase Vocoding |
I'm trying to Phase Vocode my voice in real time. But the results I have now are very distorted. I'm just curious if anyone has done such an instrument or they have some advice for me when developing such an instrument. Here's the instrument I've made so far. What I would like to do is really to hold down C-Major cord or any cord an have my voice split up to the frequencies. Open for any ideas :) massign 0, 1 instr 1 ain1, ain2 ins aut delay ain1, 0.001 aut butterhp aut, 60 aut butterlp aut, 10000 ifftsize = 1024 ioverlap = ifftsize / 4 iwinsize = ifftsize iwinshape = 0 kstatus, kchnl, kdata1, kdata2 midiin knota cpsmidib fPhase pvsanal aut, ifftsize, ioverlap, iwinsize,iwinshape, 2 fPitch1 pvshift fPhase, knota, 50 fPitch2 pvshift fPhase, knota*2/3, 50 fPitch3 pvshift fPhase, knota*2, 50 aOut1 pvsynth fPitch1 aOut2 pvsynth fPitch2 aOut3 pvsynth fPitch3 aOut sum aOut1, aOut2,aOut3 outs aOut, aOut endin -- View this message in context: http://csound.1045644.n5.nabble.com/RT-MIDI-Phase-Vocoding-tp5734236.html Sent from the Csound - General mailing list archive at Nabble.com. |
| Date | 2014-04-15 14:26 |
| From | hlolli |
| Subject | [Csnd] Re: RT MIDI Phase Vocoding |
Sorry, I would like to post the instrument like this, I was just experimenting, forgot about that. massign 0, 1 instr 1 ain1, ain2 ins aut delay ain1, 0.001 aut butterhp aut, 60 aut butterlp aut, 10000 ifftsize = 1024 ioverlap = ifftsize / 4 iwinsize = ifftsize iwinshape = 0 knota cpsmidib fPhase pvsanal aut, ifftsize, ioverlap, iwinsize,iwinshape, 2 fPitch pvshift fPhase, knota, 50 aOut pvsynth fPitch outs aOut, aOut endin -- View this message in context: http://csound.1045644.n5.nabble.com/RT-MIDI-Phase-Vocoding-tp5734236p5734237.html Sent from the Csound - General mailing list archive at Nabble.com. |
| Date | 2014-04-15 19:57 |
| From | peiman khosravi |
| Subject | Re: [Csnd] RT MIDI Phase Vocoding |
Hello,
pvshift is a spectral shifter, which distorts the spectral structure of the sound. You need to use pvscale to transpose the input. Also, if you're feeding in vocal sounds you need to use it with formant correction...
P On 15 April 2014 14:24, hlolli <hlolli@gmail.com> wrote: I'm trying to Phase Vocode my voice in real time. But the results I have now |