w4ffl35

I'm doing a performance involving granular synthesis for uni and I've cobbled together a semi-satisfactory granulizer. What I'm struggling with is the various schemes for controlling/leading it in the right direction. From my reading I'm come across using ramps and tendency masks, I've got the ramps down (use lines!) but I'm not sure how to implement a tendency mask at all.

Basically I was wondering if you guys could point me in the right direction for setting up various levels of control for the granulizer. I've thought about using drunk, chaining lines, random, but what I'd like to do is automate most of the actual parameter changing yet at the same time be able to make a change to it on the fly and nudge it in the right direction.

controlling-granular-synthesis

sorry to state the obvious but does this sound like a textbook description of [rslider]?

Hi
sorry to state the obvious but does this sound like a textbook description of [rslider]?

I've been using [rslider] + [line] to sort of achieve what I'm looking to do, but I figured there was a better/different way to control the different parameters.

I did just get an idea of using the two in combination with [count] and a series of [gates] to get a semblance of automation. I'll try it out and see how it works.

This is my very rough mock up of how I could go about controlling it

But as I soon realized, for this to be effective I would have to preset all the ways I want the parameters of the grain (pan/pitch/duration/etc) to change which is not the route I want to go.

I'm not sure if I've described correctly what I'm trying to achieve, it might be more akin to using the output of algorithm to control the parameters.

Sorry if this is sounding increasingly vague

I'm not sure if I've described correctly what I'm trying to achieve, it might be more akin to using the output of algorithm to control the parameters. 

Sorry if this is sounding increasingly vague


Very vague indeed. I would use an existing granulator (for instance the one in the Max examples) and start from there. You might get what you want by just controlling the granulation parameters through the bpatcher's inputs. But before coding, you should certainly formalize the control you want (which parameters, which range, which controllers - midi hardware, etc.)

Very vague indeed. I would use an existing granulator (for instance the one in the Max examples) and start from there. You might get what you want by just controlling the granulation parameters through the bpatcher's inputs. But before coding, you should certainly formalize the control you want (which parameters, which range, which controllers - midi hardware, etc.)
All best.


A quick look at your patch shows some kind of randomised parameter generator, but from what you describe as your problem it sounds more like you need to establish some kind of meta-control of the various GS parameters, which given the potentially large number of parameters required makes some sense.

To this end I would recommend: Firstly working out a method for generating "randomised" values and then using pattrs for all the GS parameters (ie pitchmin, pitchmax, etc. etc.) & pattrstorage for parameter storage/recall for distinct states, even possibly incorporating something Nodes if you want to use or try multi-state interpolation. You can still have a randomised process (ie tendency mask) for parameter ranges , and then store the relevant parameters for these processes

As JF already suggested, it may be better to use an established example as a starting point...

A quick look at your patch shows some kind of randomised parameter generator, but from what you describe as your problem it sounds more like you need to establish some kind of meta-control of the various GS parameters, which given the potentially large number of parameters required makes some sense. 

To this end I would recommend: Firstly working out a method for generating "randomised" values and then using pattrs for all the GS parameters  (ie pitchmin, pitchmax, etc. etc.) & pattrstorage for parameter storage/recall  for distinct states, even possibly incorporating something Nodes if you want to use or try multi-state interpolation. You can still have a randomised process (ie tendency mask) for parameter ranges , and then store the relevant parameters for these processes  

As JF already suggested, it may be better to use an established example as a starting point...


Controlling Granular Synthesis