• ### Trying to reproduce a sound with Pd

Hello,
Despite that every thing is explained, it seems to me that the parameters values are very important (especially fof lFO rate and ampliture) and I do not manage to get anything close enough ( what is 0.5 on a knob in Cut off Frequency ?) .

Would someone help ?
Thank you a lot

• Posts 8 | Views 582
• @JMC64 You would need to look up the specs of the VCF on the manufactures website and find its frequency range. They will most likely give the range in hz, so if they say it goes from 8hz to 16khz that would roughly be C1 to C11, so halfway would be C6 or roughly 262hz, remember that frequency is not linear! You can use a chart like http://subsynth.sourceforge.net/midinote2freq.html to convert between note names and hertz or you can have pd do the math for you with the `[mtof]` and `[ftom]` objects. Also remember that you do not need to get perfect frequencies, those small knobs have poor resolution and being analog have a fairly wide tolerance, if you went through a dozen of those modules you would find that 0.5 would be slightly different on each of them, so just get close and then tune by ear to perfection.

You can do the same for any module and any parameter, you just need to convert the numbers differently, a VCA will likely have its specs in decibells and a CV input or output in volts.

EDIT: Should mention, you need to look at what the frequency range of the cutoff knob is, not the VCF. Sometimes a VCF or VCO will have a larger range than can be reached by just its frequency knob, the CV inputs can often extend it, so the knob may only go bring you to C8, but a +5 volt signal into the CV input would bring you to C13 because frequency goes up one octave for every volt in a V/oct system like the Doepfers. Generally they will give the tuning range of the knob separately if the VCF can sweep beyond the range of the knob, most modules do not do this though and you should not have to worry about it unless it gives two different values.

• @JMC64 The example sounds don't 100% match the patch diagram IMHO. Consider just the two square wave generators with modulated pulse width, detuned +/- 4 cents, then mixed together: they should periodically slide past each other 180 degrees out of phase, thereby cancelling each other out. I'm not hearing that, and there's nothing between that first mixer and the output that would fill that temporary gap in. I wonder if the documentation of that patch is wrong?

• @jameslo said:

they should periodically slide past each other 180 degrees out of phase,

They do, you can hear it reasonably well in the second sample. The out of phase PWM downplays it since it is rare they both will be the 180 degrees out of phase and the same width, so most of the time it will just sound like a hiccup in the PWM.

• @oid My Pd simulation was giving much deeper cancellations no matter how much out of phase PWM I applied, but thinking about both your comments led me to set the PWM offset of each oscillator slightly off-center. That made it match much better, thanks.

Edit: I also just now discovered that if you distort one of the square waves (or are they pulse waves?) using an all-pass filter, that also prevents them from ever cancelling each other out.

• @jameslo Analog components all have tolerances, +- 10% is not uncommon for pots, so any of the knobs between the two modules could be 20% off from each other given a worse case scenario, plus all those other components in the module could make it even further off, Generally circuits are designed to minimize the effects of component tolerances and high tolerance parts will be used in critical locations like frequency but amplitude and PW are not critical location, our ears will not notice small differences here. On top of that components age and their value changes, so two 30 year old modules can be very different despite once being identical. Things like phase and frequency are not constants in the analog world, things drift and oscillators which share a power source tend to sync when they get close. Most have to be very close in frequency to sync this way, the beat frequency will be well under 1hz but under certain situations you can hear it if you listen close. When trying to copy a patch from the analog world in the digital, knobs can only be used as a rough guide, need to use your ears.

When you distorted the wave you added harmonics and changed the strengths of the old harmonics, only harmonics of equal strength but opposite phase will cancel fully. You most likely did have cancellation just not full cancellation (depends on how you distorted it). Put both oscillators to the same frequency and opposite phase, listen to the distorted one and then add in the undistorted oscillator, you should hear a decrease in harmonic content from the cancellation. This will be most apparent with low to moderate levels of distortion, the more distortion, the more harmonics you are adding, the less the two waves have in common, and as a result less gets canceled.

Edit: didn't quite finish that last sentence.

• @jameslo Hello,

Thank you guys for answering my question. I would be glad if you could share your patch.. I still do not get something sounding the sound examples...
I will also upload what I came to.. but still very far to my ears.test_birth_planet.pd

• @JMC64 I didn't do the whole patch because I was only interested in the detuned square waves, so I hope this isn't too disappointing.
BOP.pd

Posts 8 | Views 582
Internal error.

Oops! Looks like something went wrong!