Different results in Max & Live with tapin~out~ w/ feedback

[spectro]

Hi, Firstly I will say I am a long time MaxMSPJitter user, but relatively new to Max for Live. I have created a Max Audio Effect Device that incorporates a network of recirculating, cross coupled tapin~ tapout~ pairs (think Karplus-Strong plucked string algorithm). In developing in Max, I have been using an I/O vector of 512 and Signal Vector size of 64. I have the same audio buffer size in Live (512) and from some searching on this board I have been led to believe (hopefully correctly) that the Signal Vector in Live is also 64 samples.

When I try the Device in Live, I get discrete echoes (which is not what is supposed to happen - as the sound should be resonating comb-filter like tones) and leads me to suspect that possibly Live doesn't like/allow this recirculating network and is injecting samples (buffers worth?) into the feedback loops, as when I edit/preview the same Device inside MaxMSP it sounds as expected.

I'm wondering if there is anything that I may have missed along the way in translating from Max to Live and also hoping that I can indeed realize device as I'd hoped to. Any help or suggestions would be much appreciated.

[broc]

AFAIK, in Max for Live (performance mode) the I/O vector is 64, independent of the audio buffer size in Live.