KVR Audio

stefancrs · Post by **stefancrs** » Fri Mar 12, 2004 11:00 am

Then you won't get a pulse wave as a result, you need to have double the frequency on the second saw and half the inverted amplitude, so do it the way I mentioned afterwards :)

no_signal · Post by **no_signal** » Fri Mar 12, 2004 11:21 am

stefancrs wrote:Then you won't get a pulse wave as a result, you need to have double the frequency on the second saw and half the inverted amplitude, so do it the way I mentioned afterwards

But
then I have to mess with 2 freq calculations, 1 more interpolation (the freq table), and some wierd ecuations to calculate the phase_2 for the freq/2, and add the phase offset to it.
If i do Y1-Y2, than it's a square wave+ dc offset (from the PWM). Is this DC offset bad ?

no_signal · Post by **no_signal** » Fri Mar 12, 2004 11:48 am

Now I have the same problem with my asymetric triangle waveform, which is made from parabola waveform. Reading your code, but don't see the difference.

stefancrs · Post by **stefancrs** » Fri Mar 12, 2004 12:28 pm

But
then I have to mess with 2 freq calculations, 1 more interpolation (the freq table), and some wierd ecuations to calculate the phase_2 for the freq/2, and add the phase offset to it.

Yes. That is totally correct :)

If i do Y1-Y2, than it's a square wave+ dc offset (from the PWM). Is this DC offset bad

Hms, I can not see any dc offset in my tests when doing the same, so it seems like a good method to me. I dunno why I missed that one out. Skip the whole double frequency thingy... :) Maybe you just confuse the whole thing about dc offset, to _not_ have a dc offset the waveform should have an average of zero. Which this one does have. that means that with very low pulse widths most of the wave is very near zero, except for the short "peak" of the pulse. On the other hand, at a pulse width of 0.5 (symmetrical wave) the wave is as equally far from zero all the time. The average is maintained at zero though.

no_signal · Post by **no_signal** » Fri Mar 12, 2004 12:33 pm

I tryed to fix that once, to have a symetric waveform, and i did make it symetric, but than i had problems with the unison. Huge dc offset.

no_signal · Post by **no_signal** » Fri Mar 12, 2004 12:37 pm

tranceinstitute wrote:Now I have the same problem with my asymetric triangle waveform, which is made from parabola waveform. Reading your code, but don't see the difference.

And the problem is the parabola waveform, which aliases.

Oscillator->Phase Offset->Aliasing ???