Oh, I'm not going to let this place die off and go dormant. Hello again. ;-)
Sometimes you're going to want to create a control with a nonlinear taper. This can be handy for controlling the frequencies of an LFO, for adjusting the smoothness of a granular sampler, and lots of other applications where you want to taper off to very fine values without cursing at your mouse and getting wrist strain.
Here, I've created a macro that scales the output of a knob that goes from zero to one, and connected the output to a numeric display. I've duplicated the macro a few times so you see how it handles values:
And what is the magical mechanism that produces this effect? It's a simple multiplier module:
So the output is merely the input multiplied by itself, or squared. Don't go all math phobic on me now. This is easy stuff. You're looking at Y = X² which is a formula that creates a parabolic curve:
That's what it looks like in an XY scope. Here's the guts of the XY scope displaying the pretty curve: 
Fire up Reaktor and build the scope mechanism yourself. It's pretty straightforward. Make sure the settings on the XY module are as follows:
Object type should be "scope", you don't want any cursor, and the fade time should be 80 or thereabouts. Set the XY control to "always active" on the gears tab. So what happens if you multiply the value by itself three times instead of twice - cubing it? Try it and see what happens to the curve.
Tuesday, October 23, 2007
Scaling Values
Subscribe to:
Post Comments (Atom)
2 comments:
Kudos to you for trying to set up a blog for learning reaktor.
But, sorry to say that to all noobies, if you want to use Reaktor you will need some pretty good knowledge of math. If you already chicken out at x² then better go play with another tool. 40% of editing is putting up chains of blocks that mangle (streams of) numbers and you will be lost if you don't know what happens if you (for example) multiply a number with it's negative counterpart or why it is better to multiply something with 0.5 instead of dividing it by 2.
The genius of Reaktor is that the user can approach it at different levels of abstraction. You can do interesting things without thinking about math at all. When I was getting started I intuited my way through it, dissecting instruments, using multiplier modules where others had used multipliers and so on.
Until you get down to the level of designing or implementing DSP algorithms in the Core level, like FFT or filters, there's almost no need for anything more than the most basic, basic math and what I'd even hesitate to call algebra.
However, I can think of a handful of useful and simple math tricks that users ought to know in order to build fluently in Reaktor. I wish someone had laid them out on a cheat sheet for me in the beginning. And you've just given me an idea for my next tutorial. Thanks!
Post a Comment