Phase Distortion Oscillators?

stefan · 12 Apr 2005

Uhm. I thought phase distortion basicly just was phase modulation?

lalo · 12 Apr 2005

phase distortion is a distorting function applied to the wavetable itself affecting the speed at wich the wavetable is readed... if you own it you can find some info in the 'Computer Music Tutorial' at page 252 or

http://www.answers.com/topic/phase-distortion-synthesis or

http://www.cems.uwe.ac.uk/~lrlang/music/synthesis.pdf

a very synthetic ( 😀 ) resume of many synthesis techniques

lowe

lalo

stefan · 12 Apr 2005

Yeah, ok. Well. That is doable. Try setting the frequency of an oscillator to zero, connect a ramp to the phase input of the oscillator. Now try fiddling with distorting the ramp.

Edit: But remember to keep it within the range of 0..1 or whatever 🙂

lalo · 12 Apr 2005

stefancrs wrote: Yeah, ok. Well. That is doable. Try setting the frequency of an oscillator to zero, connect a ramp to the phase input of the oscillator. Now try fiddling with distorting the ramp.

Edit: But remember to keep it within the range of 0..1 or whatever 🙂

wow! so the ramp (or ramp segments) pitch will control the oscillator pitch?

thanx

lalo

stefan · 12 Apr 2005

lalo wrote: stefancrs wrote: Yeah, ok. Well. That is doable. Try setting the frequency of an oscillator to zero, connect a ramp to the phase input of the oscillator. Now try fiddling with distorting the ramp.

Edit: But remember to keep it within the range of 0..1 or whatever 🙂

wow! so the ramp (or ramp segments) pitch will control the oscillator pitch?

thanx

lalo

Yep. Unfortunately, at the moment, the built in oscillators in SM will not be "antialiased" when playing them back using the phase input instead of the frequency input. This is solvable though, and me and jorge has discussed aliasing noise quite a lot so later on, this might get fixed. But it's probably not something that hinders you in your work on PD oscillators 🙂

lalo · 12 Apr 2005

stefancrs wrote: lalo wrote: stefancrs wrote: Yeah, ok. Well. That is doable. Try setting the frequency of an oscillator to zero, connect a ramp to the phase input of the oscillator. Now try fiddling with distorting the ramp.

Edit: But remember to keep it within the range of 0..1 or whatever 🙂

wow! so the ramp (or ramp segments) pitch will control the oscillator pitch?

thanx

lalo

Yep. Unfortunately, at the moment, the built in oscillators in SM will not be "antialiased" when playing them back using the phase input instead of the frequency input. This is solvable though, and me and jorge has discussed aliasing noise quite a lot so later on, this might get fixed. But it's probably not something that hinders you in your work on PD oscillators 🙂

good to know stefan! thanx for the great info.... ..sometimes i love aliasing 😀 i'm a nasty boy...LOL

lalo

Sk · 12 Apr 2005

This may look stupid but... what is exactly the pitch? I thought it was the same as the frequency...

Shifrin · 12 Apr 2005

AFAIK Pitch = Frequency (more or less) in the description above 🙂

Shif.

lalo · 12 Apr 2005

Shifrin wrote: AFAIK Pitch = Frequency (more or less) in the description above 🙂

Shif.

well yes..more precisely frequency 🙂 ..i have to figure out how to 'segment' the ramp tho 🙂

lalo

stefan · 13 Apr 2005

lalo wrote: Shifrin wrote: AFAIK Pitch = Frequency (more or less) in the description above 🙂

Shif.

well yes..more precisely frequency 🙂 ..i have to figure out how to 'segment' the ramp tho 🙂

lalo

What do you mean by segment?

lalo · 13 Apr 2005

stefancrs wrote: lalo wrote: Shifrin wrote: AFAIK Pitch = Frequency (more or less) in the description above 🙂

Shif.

well yes..more precisely frequency 🙂 ..i have to figure out how to 'segment' the ramp tho 🙂

lalo

What do you mean by segment?

hi stefan! i mean a segmented line.. a function that changes its slope at a certain x value...this will change the actual reading speed of the wavetable during time...maybe a multi-segmented line too....slope changes in the ramp make a reading speed change in the wavetable.. that's pure PD 🙂 lalo

stefan · 13 Apr 2005

lalo wrote: stefancrs wrote: lalo wrote: Shifrin wrote: AFAIK Pitch = Frequency (more or less) in the description above 🙂

Shif.

well yes..more precisely frequency 🙂 ..i have to figure out how to 'segment' the ramp tho 🙂

lalo

What do you mean by segment?

hi stefan! i mean a segmented line.. a function that changes its slope at a certain x value...this will change the actual reading speed of the wavetable during time...maybe a multi-segmented line too....slope changes in the ramp make a reading speed change in the wavetable.. that's pure PD 🙂 lalo

Oh. ok. I'll try to knock something basic up and get back to you. I thought it be more interesting to just apply some kind of distortion function (the parabolic for instance) to the ramp 🙂

lalo · 13 Apr 2005

yes..this is a more..complex thing in terms of synthesis... as i said the 'classic' PD (as it was implemented in casio synths down in the '80 🙂 ) was done with segmented ramps.... another interesting thing could be crossfade between 2 different waves at sample rate 🙂 ...easy with SM i guess.. anyway i'm very courious about a segmented ramp....i guess you'll choose to do it with the code component 🙂

lalo

stefan · 13 Apr 2005

You guessed right. But it's easily doable without the code module as well, very basic operations in action. Here we go:

Three different params to control the segments behaviour, is that enough? This should get you started atleast. Schematic: stefan phase dist.osm

Do tell me if I misunderstood PD 🙂

lalo · 13 Apr 2005

thanx stefan! i go learn 🙂

lalo

lalo · 13 Apr 2005

thanx stefan! a little thing... i've not yet check the math of your code ...but i've analyzed its output graphically....and doesn't seems to me like the following... but maybe i'm wrong...

ideally i would need a paramter to set the point from wich the second segment start and another parameter for the slope of the second segment obviously a paramter for the first segment's slope would be nice.... i suppose is only a matter of multiplying and scaling... damn i would i have studied 'serious' math when i did gymnasium ...only latin/greek/philosophy and some addition 😀

can you suggest me some math text where i can study this pretty basic stuff and the needed calculus for dsp?

what i'm after is to obtain a segmented ramp like this

starting from a plain ramp like this

the first ramp (the segmented one) read the wavetable more slowly in its second part..that's the phase distortion effect

stefan · 13 Apr 2005

That's what the first slider is supposed to do. Maybe I should've looked at it's output, but when I set the other two sliders to the middle that is the result I think I'm seeing. Am I doing it wrong?

edit: I took a look now, and I see what I'm doing wrong. I'll post another version asap 🙂

lalo · 13 Apr 2005

stefancrs wrote: That's what the first slider is supposed to do. Maybe I should've looked at it's output, but when I set the other two sliders to the middle that is the result I think I'm seeing. Am I doing it wrong?

edit: I took a look now, and I see what I'm doing wrong. I'll post another version asap 🙂

thanx stefan 😉

in the mean time i'll try to figure out what the function for a segmented line can be math-wise 🙂

lalo

stefan · 13 Apr 2005

Here we go: stefan phase dist 2

lalo · 13 Apr 2005

stefancrs wrote: Here we go: stefan phase dist 2

you are too fast man 🙂

thanx

when i was around with my doggie (spike) i got the right idea i guess 🙂 even if i have some trouble about frequency 🙂

i'll check yours 🙂

lalo

stefan · 13 Apr 2005

lalo wrote: stefancrs wrote: Here we go: stefan phase dist 2

you are too fast man 🙂

thanx

when i was around with my doggie (spike) i got the right idea i guess 🙂 even if i have some trouble about frequency 🙂

i'll check yours 🙂

lalo

Taking walks and visiting the toilet makes wonders to the idea machine.

lalo · 13 Apr 2005

stefancrs wrote: lalo wrote: stefancrs wrote: Here we go: stefan phase dist 2

you are too fast man 🙂

thanx

when i was around with my doggie (spike) i got the right idea i guess 🙂 even if i have some trouble about frequency 🙂

i'll check yours 🙂

lalo

Taking walks and visiting the toilet makes wonders to the idea machine.

visiting the toilet the most 🙂

lowe lalo

Tom7777 · 13 Apr 2005

Thanks stefancrs 🙂

Shifrin · 13 Apr 2005

Yet more stuff to get my head around, very interesting guys! Thanks for the .osm Stefan 🙂

Shif.

Solstice · 9 Apr 2009

stefan wrote: Here we go: stefan phase dist 2

The link is dead. Can you reupload it, please ?

Thank you very much ! 🙂

stefan · 14 Apr 2009

Solstice wrote: stefan wrote: Here we go: stefan phase dist 2

The link is dead. Can you reupload it, please ?

Thank you very much ! 🙂

Oh. Ouch. I don't have it any longer, I'll see if I can dig it up somewhere...

stefan · 14 Apr 2009

Sorry, they seem to be long gone. Anyone else got those old files stored away somewhere?

theun · 14 Apr 2009

stefan wrote: Sorry, they seem to be long gone. Anyone else got those old files stored away somewhere?

Yep. I found it on my drive:

I hope this is the right file.

Cheers!