I'm trying to create a simple instrument that will prolong the sustain of a live instrument. I've been tinkering with different approaches for the past few days and I'm wondering if anyone can suggest anything else. If I use pvsfread/write I can get a nice warm sustain going with the caveat that I have to use two different instruments because I can't access a file being written until the instrument writing the file has stopped. If I use pvsbuffer to apply the sustain the audio is warbly and not very convincing. Here are the csds which I've being using. Remember to change the audio devices if you want to try them. I've also tried using pvsfreeze but I don't want to freeze the spectrum as I want to preserve the spectral evolution of the captured audio, even if it is very short. Any tips or suggestions are most welcome. Rory. ;--------------------------------------pvsfread/write example----------------------------------- -odevaudio1 -b10 -idevaudio2 sr = 44100 kr = 44100 ksmps = 1 nchnls = 1 instr 1 asig inch 1 ; input Sname sprintfk "soundin-%d.wav", p4 fsig pvsanal asig, 1024, 256, 1024, 1 ; analysis pvsfwrite fsig, Sname ; write file endin instr 2 k1 linen 1, 0.1, p3, 1 Sname sprintfk "soundin-%d.wav", p4 idur filelen Sname ; find dur of analysis file kpos line 0,p3,idur ; to ensure we process whole file fsigr pvsfread kpos+0.5, Sname aout pvsynth fsigr out aout*k1 endin f1 0 1024 10 1 i1 2 .3 1 i2 2.5 20 1 i1 6 .3 2 i2 6.5 20 2 ;---------------------pvsbuffer/read example------------------------------------------ -odevaudio1 -b10 -idevaudio2 sr = 44100 kr = 44100 ksmps = 1 nchnls = 1 ;capture 1st second of line audio and sustain it.... instr 1 asig inch 1 k1 linseg 0, 1, 0, 1, 1, p3, 1 ktime line 0, p3, p3 if(ktime<1) then fsig1 pvsanal asig, 4096*2, 2048*2,1024*8, 2 ibuf1,kt1 pvsbuffer fsig1, 2 endif khan init ibuf1 if(ktime>1) then kpt phasor 0.1 fsb1 pvsbufread kpt+0.5, khan aout1 pvsynth fsb1 out aout1*k1 endif endin f1 0 1024 10 1 i1 0 100