Clicks/artifacts when increasing pulse width on osc/pwm

#1

Finally getting to sit down a bit with the axoloti, and I'm noticing that increasing (but not decreasing) the pulse width on an osc/pwm module there is a lot of crackle, like what you would get if you jumped between two distant values in a wavetable for example; I forget the specific name for this kind of distortion. It's easy to hear in the tutorial patch 05_oscs by switching to the osc/pwm (4: asymmetrical square on the radio button module) and sweeping the pulseswidth value. It's also clearly audible in cpwitz's polysynth patch. Leaving it on the default settings (turning down the suboscillator makes it even easier to hear) you can hear the same crackling every time the LFO value that's modulating the pulsewidth rises. It's a bit more subtle in the context of a complete patch like this, but it's still pretty loud. It definitely makes natural sounding pulsewidth changes or modulation impossible.

Here:s a simple patch demonstrating it:
pwm_bug.axp (1.8 KB)

When the LFO is rising you can easily see the edge of the square wave glitching as well as hearing it.

Sorry if this is already a known bug, I couldn't find anything about it when I searched on here.

Thanks!

#2

the very first bug on github smile
might be interesting to attach it to a scope object, and see if its obvious whats going on.
(though it could be its too fast for the scope to show) - as you say sounds like some kind of discontinuity, perhaps not staying at a level for the correct time.

#3

Cool, thanks. Should have checked there first.

#4

no problem, it just 'rang a bell' when i read your post... so I thought Id let you know its been logged.

#5

It's actually kind of a useful effect sometimes, just not all the time. I hold onto an ancient Akai s612 specifically because it has faders to set the loop points and they make a similar sound if you adjust them while a sample is playing.

#6

aww thats my bug! smile