I sure didn't see this coming.
I was playing around with an idea I had. I was curious to see if I could get the same kind of results by off setting the start of a sample as I could when I detune. My idea was that I would get a flanged sound, but a stable flanged sound. Something that would loop easier. It was a thought that came about form experiments with two cd players and off setting one just a bit but keeping them the same speed. When you do that, you get a steady flange sound, or a highpass or lowpass kind of sound. Regardless it was stable and that is something I was looking for. I failed, but only to discover some interesting characteristics about certian waves. Specifically the ramp and sawtooth.
I loaded up the sawtooth and ramp. Set it up so I could get both playing on top of each other like I normally do and then I went into voice edit of the ramp. I went to start off set and just for laughs set it at one. when I hit the key I found that the timbre of the two waves changed dramaticly. It sounded like a square wave. I hook up my darla to my sampler and started recording. It was infact a square wave. Almost perfect. Just like the one I sampled from the emu or wave lab. Perfect all the way down the the little spike of noise at the beginning of each cycle.
So I started to play the same note, and record as I tweaked the amount of start off set. When I looked back that the recording in Cubase I found that each wave form was still a square wave, but was actualy PWM. Each wave form varied form a square to PWM of like 2:8 to 6:4 and so on. One difference though. On the most extreme setting at 127 I found that I had a very small spike, almost like a triangle wave but with long levels of zero. Also found the sequence repeating though out the value climb and also inversions of the wave forms.
I do know of the PWM tech using the sloop cord, but what I was interested in was the 'sstart chord to recreate the off set. And it worked. It has a note on property, unlike the sloop cord that is real time, but interesting all the same.
Square waves and presampled PWM's work well with this cord set up. One thing that was really interesting was using two sawtooths with this cord. What started to happen was a sort of PWM with those, but the out come was more of pitch shifting. what would happen was for the lenght of one cycle, as the 'sstart off set the start of the second copy, I would start to get various amounts of cycles inside the one cycle. At one point I had two cycles in the lenght of what one cycle was at normal start.
This cord is pretty stable, each value holds pretty true to what is should be. One thing is that the cycle seems to repeat about three times if you go through the full range and about the second cycle you get inversions of the wave form. The only problem is, you would have to map out what each value does to the wave form with this cord as it is a bit unpredictable. Such as going form the spike with long zero lengths to a PWM of 8:2 and so on.
This works a little with established timbre's like a sample of a violin, but do to the huge amount of overtones, the PMW really doesn't seem to be as noticeable as it is with raw test tones. But it did seem to clean up any reese-y type modulation a little. Might be worth playing around with.
Any way. Just wanted to share this with you guys. My day of discovery.
snh