Author |
Message |
triptonizer
Joined: May 22, 2004 Posts: 22 Location: wonderland
G2 patch files: 1
|
Posted: Fri Oct 29, 2004 4:57 am Post subject:
[slightly OT] math behind sine2 shape osc |
|
|
Does anyone happen know the math behind this? Like you can substract 2 saws (same pitch and amp, the one syncing the other), and by modulating the second one's phase offset you create your own pulse pwm. All I know is that with shape = 100% it looks a bit like the bipolar version of a rectified sine, but I don't know how to get to the intermediate shapes. It's really for something I'm building in Reaktor; Reaktor has a similar Parabol pwm module, but it doesn't have sync and I need one that can be synched. I suppose I could get there myself by trial and error, but I'd prefer a takeway solution - don't feel much like cooking tonight |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Fri Oct 29, 2004 5:44 am Post subject:
|
|
|
What I would try to do is make a sawtooth that can be made non linear. After that a table lookup or some simple math. That way the symetry of any waveform could be modulated by modulating the non linearity of the saw.
Just an idea,
Jan. |
|
Back to top
|
|
|
triptonizer
Joined: May 22, 2004 Posts: 22 Location: wonderland
G2 patch files: 1
|
Posted: Fri Oct 29, 2004 2:23 pm Post subject:
|
|
|
Nice idea, thx Jan. But I'm afraid lookup tables will cause tons of aliasing... I'll give it a try though. Incidently, I asked this because I spent a couple of hours with an EMS Synthi100 yesterday at the university here (Ipem). Osc 1-6 (in total there are 12 oscillators!) are syncable (through the patchboard) and each osc has two outputs, one of which is 'SIne with Shape Control'. Hence my sudden interest in the matter
|
|
Back to top
|
|
|
ian-s
Joined: Apr 01, 2004 Posts: 2669 Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 626
|
Posted: Fri Oct 29, 2004 3:27 pm Post subject:
|
|
|
I'm interested as well, did you patch it into the scope to see what the intermediate wave looks like? |
|
Back to top
|
|
|
triptonizer
Joined: May 22, 2004 Posts: 22 Location: wonderland
G2 patch files: 1
|
Posted: Fri Oct 29, 2004 4:23 pm Post subject:
|
|
|
g2ian wrote: | I'm interested as well, did you patch it into the scope to see what the intermediate wave looks like? |
yeah, and it's also described in the original manual: they look exactly like the Sine2 waveform on the G2's Shape Osc A, except that they have the full shape range (0...100%), so you get a sine with shape control in the middle position, turning the knob to the left is the same as moving the shape knob towards 100% on the G2 module, turn it to the right and you get the same output but with inverse polarity. |
|
Back to top
|
|
|
Rob
Joined: Mar 29, 2004 Posts: 580 Location: The Hague/Netherlands/EC
G2 patch files: 109
|
Posted: Sat Oct 30, 2004 10:38 am Post subject:
|
|
|
triptonizer wrote: | g2ian wrote: | I'm interested as well, did you patch it into the scope to see what the intermediate wave looks like? |
yeah, and it's also described in the original manual: they look exactly like the Sine2 waveform on the G2's Shape Osc A, except that they have the full shape range (0...100%), so you get a sine with shape control in the middle position, turning the knob to the left is the same as moving the shape knob towards 100% on the G2 module, turn it to the right and you get the same output but with inverse polarity. |
There is a big difference between the G2 shape osc and how that distorted sine wave is made on the EMS, on the G2 the shape is distorted in the horizontal direction by means of a FM-like technique, while on the EMS the sine is created by distortion in the vertical direction, which basically is AM-like.
According to the EMS schematics a triangular wave is put into a triangle to sine converter, which uses diode devices to 'softly clip the top and bottom' of the triangle. The sine waveshaping distortion is made by adding an offset value to the triangle wave before it goes into the triangle to sine converter, forcing the result to be asymmetric.
The example patch is my solution.
Drawback is that this solution only works with a positive shape control signal, while on the EMS the control signal can be both positive and negative. In practise there not really an audible difference, so this solution is fine enough for me.
There is some extra circuitry to keep the signal level ok and prevent the waveform to shift towards the clipping headroom of the G2 outputs.
Description: |
Emulation of the sine osc waveshaping on the EMS Synthi. |
|
Download |
Filename: |
SynthiSineOsc.pch2 |
Filesize: |
1.13 KB |
Downloaded: |
1734 Time(s) |
|
|
Back to top
|
|
|
ian-s
Joined: Apr 01, 2004 Posts: 2669 Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 626
|
Posted: Sat Oct 30, 2004 12:39 pm Post subject:
|
|
|
Thanks Rob, works very well.
And thanks Triptonizer, this wave sounds quite interesting when synced, specialy when you modulate the shape as well.
Is that a modified Synthi100 or do they all have sync? The VCS3 and Synthi-A need mods for sync and VC shape.
I hear Daniel Miller rescued a Synthi100 from a rubbish skip, lucky bastard! |
|
Back to top
|
|
|
triptonizer
Joined: May 22, 2004 Posts: 22 Location: wonderland
G2 patch files: 1
|
Posted: Sat Oct 30, 2004 2:38 pm Post subject:
|
|
|
wow thx rob, I'll dig into it tomorrow but at first glance I think this would be pretty easy to emulate in Reaktor, which was my original purpose.
Do you by any chance have a similar solution for the syncable saw/tri shape osc in the Synthi? I only have the users manual here, no schematics.
@g2ian: it's not a modded Synthi100, all oscillators are syncable but you need to patch the sync source on the patchboard, nothing hardwired.
btw I heard Aphex Twin recently bought a Synthi100 on eBay |
|
Back to top
|
|
|
|