KBD_TRACKER wrote:Thanks for putting up these graphs.
But looking at the 2 curves the minimoog wavelength is 10 units, that of the voyager is 9.5 units so already there is a frequency setting discrepancy of 5%: they are not tuned. If I remember correctly the difference in frequency between A and A# is only about 4.3 % ......
Um... so? You should be looking at the shape
of them, which is visibly different. The first one is showing visible "shark-finning", has otherwise good linearity of the slopes, and no visible reset glitch. The second has better symmetry, significant nonlinearity in the slopes, and a visible reset glitch at the bottom.
KBD_TRACKER wrote:As far as the (bi) amplitude that of the minimoog signal is 5 units, that of the voyager is 6 units, so that there is an amplitude discrepancy: their sound level is not properly matched: the voyager raw oscillator will sound 20% louder.
End result: of course these oscillators won't (and can't) sound alike: they are not properly set to sound alike.
I honestly have no idea what you're trying to say here.
But to be fair there is a bottom notch on the voyager wave that's not present on the minimoog's. So here there is a difference that's not due to mismatched settings, and which might lead to an intrinsic sound "difference".
The question remains: is this notch going to be more than barely perceptible, particularly after post-osc. processing... ?
As has been said repeatedly by everyone else in the thread: yes
. Along with the aforementioned other significant differences
between the two. And that's before you get into frequency-domain stuff like pitch variation (fluctuations in the expo converter etc.), tracking and scaling, bandwidth limitations, etc, etc...
I'm about 90% of the way through designing a discrete oscillator at the moment, and I can list pages of points in a design where you can get variation of the output. And differences can be audible well before you can pick them out easily on a scope - the human ear is an incredibly fine instrument.