Author |
Message |
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Sun Mar 14, 2010 5:16 pm Post subject:
Grain Cloud Delay Subject description: Creates dense granular ambient drones from incoming audio |
 |
|
This is a further development of the GranuSustainer patch, offering a more powerful granular engine for a denser sound.
Incoming audio is fed into a 2 sec delay line with feedback. The delay line is "read out" by a powerful 8-grain engine, creating a dense ambient granular "cloud". The delay line can also be "freezed", which puts it into 100% feedback and simultaneously shuts off the audio input, thus producing an everlasting complex drone.
The patch is based on a new granular design, which wrings 8 grains from one 4-tap delay via multiplexing the four 96kHz taps into eight at 48kHz and combining this with an efficient new octal grain envelope generator. With this large grain count (PolyGranuSyn_X only has two) there was no DSP power left to include any kind of grain pitching. Subtle grain pitch jitter however is an important ingredient for making the sound thicker and less flangey. I discovered that -in this specific scenario- i can get the same effect by simply adding subtle pitch modulation to the incoming audio. Since the grain position-randomising range covers the whole delay line, this will result in individual grains "chorusing" with each other pretty much as they would with jittering. Does the trick. The grains are also randomised in panning, creating a complex lush chorus-y stereo "sound-cloud" from whatever you sent into the patch. Add copious reverb and you've got the ultimate ambient soup generator. Instant Brian Eno.
Controls:
Page A:
INPUT Source: Select source stereo bus
CHANNEL Source: Select L or R channel from the selected bus
EFFECT Send: This patch is intended to be used as an external send effect. So I added this control.
INPUT COMPRESSOR: Compress inputted audio. It's a good idea to iron out the dynamics considerably in order to generate a more uniformly sustaining grain cloud.
Page B:
GRAINS Rate: Grain clock base frequency. Multiply x8 for actual grain frequency.
DELAY FB: Feedback amount of the source 2 sec. delay line.
FB LPF: Lowpass filter within the feedback loop.
FB HPF: Highpass filter within the feedback loop.
FREEZE Hold: Freezes the delay line for as long as you hold this button.
FREEZE Toggle: Freezes the delay line, toggle on and off.
INPUT MOD Rate, Amount: Pitch modulation of the incoming audio, for generating a chorus-y sound.
FREEZ TRANSITION: Crossfading transition time when going into freeze mode. You can probably leave this as it is.
ACTIVE Grain1-Grain8: Turn on/off the individual grains. They are played back in a round-robin fashion (1 to 8 ). So, if you eg. want a smooth 4-grain engine, only activate grain 1, 3, 5 and 7. Or if you want only a 2-grain engine, only activate 1 and 5. I don't presume anyone wants to do this anyway, since 8 grains sound so much fatter. But if you want to, here it is.
Page E:
Reverb and output section
Fits on a DSP.
If you prefer a smaller delay size, you can simply change it in the patch. This doesn't affect the functioning at all. However 2 seconds is the maximum -except if you remove the reverb (which you are free to do). Interestingly, even with 2.7 seconds, there appears to be some space left for the 5ms delay line used for pitch modulation.
The patch is more or less an one-trick pony. But it's quite a nice trick (at least for me as an ambient lover), so I thought I'd share it.
cheers,
t
Description: |
Creates a thick grain cloud from an incoming audio source. Perfect ambient drone maker. Powerful 8-grain engine. |
|
 Download (listen) |
Filename: |
GrainClouDelayTK.pch2 |
Filesize: |
4.73 KB |
Downloaded: |
2895 Time(s) |
|
|
Back to top
|
|
 |
lefev

Joined: Apr 04, 2007 Posts: 12 Location: NYC
|
Posted: Sun Mar 14, 2010 8:55 pm Post subject:
wow 8 ) |
 |
|
Thank you so much for this cool and wonderful patch!
Like how it developed from the earlier one . Very nice inflections in the ambience , nice even without reverb. Appreciate you sharing your discoveries, inspiring stuff for sure! |
|
Back to top
|
|
 |
BobTheDog

Joined: Feb 28, 2005 Posts: 4044 Location: England
Audio files: 32
G2 patch files: 15
|
|
Back to top
|
|
 |
iPassenger

Joined: Jan 27, 2007 Posts: 1068 Location: Sheffield, UK
Audio files: 5
G2 patch files: 78
|
Posted: Mon Mar 15, 2010 2:50 am Post subject:
|
 |
|
Hey nice one Tim. I've been wanting to do something with grains on the Nord for ages but not really got round to it and now I don't have to.
With regards to the pitching limitations, could you not get round this by using another slot and turning it into a performance? Just if case you wanted something else to scratch your head over.
Not played with the patch yet but I will soon. Bob's mp3 sounds cool though. _________________ iP (Ross)
- http://ipassenger.bandcamp.com
- http://soundcloud.com/ipassenger |
|
Back to top
|
|
 |
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Mon Mar 15, 2010 8:10 am Post subject:
|
 |
|
iPassenger wrote: |
With regards to the pitching limitations, could you not get round this by using another slot and turning it into a performance? Just if case you wanted something else to scratch your head over.  |
Obviously, overstepping the boundaries of one DSP opens up all kinds of extra possibilities. In this scenario however, one runs into a new bottleneck, which is only having 4 interslot busses. 8 grains equals 16 control signals (amplitude and readout pointer per grain). Of course, one could theoretically again multiplex these in order to squash them through the 4 busses. But dynamic grain readout pointers necessarily need to clock at audiorate for the sound quality to be good, so this doesn't work out. Also, the required mux/demux scheme would eat it's share of DSP. And most importantly, really good grain pitching (meaning: which has decent exponential tracking) is quite expensive to pull off. This was the reason why PolyGranuSyn_X only had 2 grains per voice (it's predecessors all had 4, but didn't sound good).
As I'm more and more getting into patches that max out the complete G2 anyway , I've been thinking about making a monophonic "monster patch" spin-off from PolyGranuSyn_X ("MonoGranuSyn", I guess ), which then would have 14 simultaneous grains while flattening out an expanded G2. Same for ElasticAudio_X, which would then have 7 dynamic wavecycle clustering engines on top of each other. Whacking those out of sync via jittering should give a very dense sound.
cheers,
t |
|
Back to top
|
|
 |
iPassenger

Joined: Jan 27, 2007 Posts: 1068 Location: Sheffield, UK
Audio files: 5
G2 patch files: 78
|
|
Back to top
|
|
 |
GovernorSilver

Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
|
Posted: Sun Jan 23, 2011 7:35 am Post subject:
|
 |
|
I tried the patch using the mic but wasn't able to get a sound out of it. Does the patch default to a line in instead of the mic.
My G2 user skills are admittedly quite rusty. _________________ Current and recent work on Soundcloud
Some old stuff on VIRB |
|
Back to top
|
|
 |
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Sun Jan 23, 2011 7:57 am Post subject:
|
 |
|
I checked quickly and yes, you're right. The input of the preset variation (there's only one) is set to interslot bus 1. That's because I ran audio from a patch in another slot into the algorithm when designing it, rather than an external signal.
The two upper leftmost modules are in charge of audio channel selection. "Input" selects the source stereo bus and "channel" selects L or R. You can change it there in the editor. Alternately the controls are also assigned to the panel page A1.
cheers,
t |
|
Back to top
|
|
 |
GovernorSilver

Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
|
Posted: Mon Jan 24, 2011 5:22 pm Post subject:
|
 |
|
As it turns out, the Browser window was covering up the Input module in the G2 Editor - that's why I couldn't figure out how to switch the input to the mic.
This patch sounds awesome! Thanks, Tim! _________________ Current and recent work on Soundcloud
Some old stuff on VIRB |
|
Back to top
|
|
 |
taylor12k
Joined: May 04, 2005 Posts: 56 Location: brooklyn
|
Posted: Fri Feb 18, 2011 8:43 am Post subject:
|
 |
|
this is highly, highly, impressive...
i have a Freeze pedal (2 actually!) and love it... and love all things ambient... this patch sounds sort of like a combo of the Freeze and something you'd get out of an Eventide..
beautiful work! |
|
Back to top
|
|
 |
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Fri Feb 18, 2011 10:46 am Post subject:
|
 |
|
Thanks.
There are 2 ways to perform audio freeze: time-domain (sampling/granular) and frequency-domain (FFT/resynthesis). AFAIK, those freeze pedals use the latter method while this uses the former.
Both have their charms. Unfortunately, comprehensive (meaning: really usable) FFT/resynthesis is impossible on the G2, as there are no dedicated modules to perform this, and patching it the DIY way eats up all DSP very fast.
I implemented subtle pitch modulation to mask potential artefacts of the granular method and add chorus-y lushness. This might not be "politically correct" , but that's the kind of sound I like. |
|
Back to top
|
|
 |
|