Solaris Rotors in G2???

peterkadar · Posted: Sun May 03, 2009 12:30 am Post subject: Solaris Rotors in G2???

Hey everyone,

I just saw this and thought it would be REALLY cool if someone with more patching experience than myself could build this for the G2.

http://www.johnbowen.com/videos.html

Scroll down to video #3 of 12, entitled Solaris - The Rotors.

I thought it was really cool when they pushed the rotor rate into the audio range.

Antimon · Posted: Sun May 03, 2009 1:57 am Post subject:

I think the basic functionality can be accomplished using an X-mux, as in the attached patch. I've added oscillator 1 in position 5 so that it will fade smoothly back to position 1. If you want to modulate this "G2 rotor" you can put something into the empty input in the module named "rotor pitch". I sensed a number of possible patch routings in the menus he showed - you can probably recreate most of those in the G2. Also check out the control sequencer for cool LFO-driven value changes (of course you can use whatever module instead of an LFO if you want).

It won't sound exactly like the Solaris - the control input to the mux is blue, meaning that (while it does go into audio range) it doesn't have as good a resolution as the red inputs. Also, the Solaris probably has some other DSP solution for the rotor to create that cool sound. An nice trick anyway - I like to use the x-mux in this fashion every now and then.

/Stefan

sbt_rotor.pch2

Description:

Download (listen)

Filename:

sbt_rotor.pch2

Filesize:

1.42 KB

Downloaded:

1996 Time(s)

Tim Kleinert · Posted: Sun May 03, 2009 10:37 am Post subject:

(deleted, see below)

Tim Kleinert · Posted: Sun May 03, 2009 10:47 am Post subject:

(deleted too, see below)

Tim Kleinert · Posted: Sun May 03, 2009 11:18 am Post subject:

I deleted the above posts, as they presented patches that were based on a wrong conceptual assumption. I wrongly assumed that the Solaris Rotor is based on a quadrature oscillator, which it isn't.

So here's the final version that does what the Solaris does, with the same audio quality (the Solaris clocks at 96kHz too). No aliasing. Smile

6.9% cycles, 12.5% memory usage. Nice building block. Smile

By toggling the "HiFreq"-button on the mixer on and off, you can select between audio and subaudio oscillation range.

Edit: different design below.

AudioRteRotor3TK.pch2

Description:

Final version. Finally :) works like the Solaris rotor.

Download (listen)

Filename:

AudioRteRotor3TK.pch2

Filesize:

1.58 KB

Downloaded:

1602 Time(s)

Tim Kleinert · Posted: Sun May 03, 2009 12:14 pm Post subject:

Gee, I managed to streamline it even more.

7.5% cycles, 9.4% memory

AudioRteRotor4TK.pch2

Description:

Even more streamlined implementation of the Solaris Rotor

Download (listen)

Filename:

AudioRteRotor4TK.pch2

Filesize:

1.58 KB

Downloaded:

1620 Time(s)

Antimon · Posted: Sun May 03, 2009 12:21 pm Post subject:

Wow!

Advanced... Er, could you take a little moment to explain how it works? Smile

I think that a value exists somewhere that moves like a reverse saw, but well, this is heavy stuff. Also, where do you adjust the xfade amount?

Nice surfing avatar, btw!

/Stefan

Tim Kleinert · Posted: Sun May 03, 2009 12:46 pm Post subject:

It's an audiorate DIY charge-reset sawtooth core, not unlike an analog oscillator (charging capacitor). Since the overflow isn't taken into account, the oscillations always have the same sample count and thus don't produce foldbacks (aliasing), but sacrifice tuning resolution, which isn't a big deal here.

In the first version, this drives a set of logic and switches that select the inputs, as well as being converted into triangle waves to perform the crossfading between those.

The second version is far simpler, doing away with that switching stuff. The sawtooth is converted into two unipolar triangle waves, 180 degrees apart, which drive fader modules with the inputs connected to them. Mix together, done. Cool

Edit: These audiorate DIY oscillators and counter circuits (which I invented some time back) are central to all my patches. Eg., the elastic audio patches are brim full of them.

cappy2112 · Posted: Sun May 03, 2009 5:04 pm Post subject:

Tim Kleinert · Posted: Mon May 04, 2009 2:39 am Post subject:

ehdyn · Joined: Sep 13, 2007 Posts: 33 Location: Caerdydd, Palau

Thanks for sharing Tim - excited to see what you've come up with.
Any chance you could post the quadrature patches in another thread?

Thanks again!

Modularmind · Posted: Wed May 06, 2009 1:54 pm Post subject:

I just bought the G2 and patches like this really gets me excited to learn a lot about modular synthesis! I have a question though - you say it's streamlined to use 7,5% cycles and 9,4% memory - but when I load it into my G2 (expanded, running OS 1.61) it uses 17,0% cycles and 14,1% memory. Am I missing someting here...?

Anyway, thanks for the patch!

/Mm

jima · Posted: Wed May 06, 2009 2:23 pm Post subject:

cappy2112 · Posted: Wed May 06, 2009 6:07 pm Post subject:

blue hell · Posted: Wed May 06, 2009 6:13 pm Post subject:

Modularmind · Posted: Thu May 07, 2009 2:43 am Post subject:

Pardon my ignorance - what is kernel modules and which would they be in this patch?

About the expansion and polyphony: If I have only one slot activated and have a patch in that slot that use more than 50% cycles I get 8 voices. If I didn't have the expansion I would get 4 voices. Right?

/Mm

iPassenger · Posted: Thu May 07, 2009 3:12 am Post subject:

Yes yr right, presuming all of that patch load is in the poly area.

By keeping the stuff you can live with in mono in the mono area you can save substantially on poly count.

iPassenger · Posted: Thu May 07, 2009 3:14 am Post subject:

Modularmind · Posted: Thu May 07, 2009 3:29 am Post subject:

Ok, the "processing" modules... Thanks for enlightening me!

blue hell · Posted: Thu May 07, 2009 3:51 am Post subject:

Tim Kleinert · Posted: Sat May 09, 2009 6:54 pm Post subject:

ehdyn · Joined: Sep 13, 2007 Posts: 33 Location: Caerdydd, Palau