Before trying other waveforms I tried intervals. Given two adjacent strings I would drone one and slide up and down the other. E/A, A/D, didn't seem to have any problems, but D/G, G/B, B/E did. It gets a little weird in the 2nd and minor 2nd intervals and the weirdness increases as the strings get higher.
On to the other waveforms - I went through triangle, square, and saw. Each got increasingly worse, as I would expect, since that is like distorting a guitar signal. Triangle was a little worse and with square trying the 4 note chords as before I couldn't distinguish the intervals. Since there are even more harmonics with the saw tooth that was the worst. I then applied a low pass filter to the two OSCs and lowered the cutoff to around 36. I then worked my way back through the waveforms - saw, square, triangle, and sin. For the most part everything was good. Since I can filter the waveforms back to sin waves and they sound OK it seems that the pitch recognition is fine. The fundamental frequency is fine.
I'll have to think about this a bit, right now I'm thinking it may just be that the upper level harmonics aren't playing nice with each other. Like distortion. I put an fx box on the direct out and used OD/DS, Muff Fuzz, Drive 60, Tone 0, Lvl 50, Bottom 0, Direct Lvl 0, Solo off, Amp off, and NS off and that's pretty close to a square wave. What happens with the droning is really similar to the square wave; the 4 note chords sound like mush, etc.
Brak(E)man wrote "I'm comparing the guitar waveform , ( not guitar direct thru ) with other waveforms
The guitar waveform at least when not distorted works MUCH better recognizing the chords , no graininess or turning the chords into mush. "
If you're talking about the Input setting for the OSC I'm not sure it's treated the same. You can't change the pitch, fine tune it, sync, ring modulate, sequence or "layer" that signal.
That's about what I figured, since there's no hex prossesing there'll be intermodal distortion, and I'm talking about the settings for the guitar , not the guitar wvwform
But I guess that the main issue pitch recog works in a way and that should be the same with guitar waveform as well as the other, the problem is the graininess of the waveforms and most likely that it's not " poly distortion ". it's a huge dissapointment as such , the machine itself is great , but with absence of hex
all the vgs and gr300 are better in this aspect.
All input to work 'round the problem is welcome