Tim Lloyd

Just wondering if anyone knew how I could create a slide~ like abstraction that smoothed the signal in an s-curve?

I need a signal transition of 0 to 1 to create the 2nd half of a hann-function, and a transition of 1 to 0 to create the 1st half of the curve.

That would mean that the successive 0 to 1 to 0 transitions would create an inverted hann-function that could be used for really smooth crossfading.

I'm trying to make a crossfading selector~ abstraction similar to Eric Lyon's [el.clean_selector~] but fully signal controlled.

Using slide~ nearly works, but as the below patch shows, the slide down has a "sharp edge" at the beginning of the ramp, and the slide up has a "sharp edge" at the end of the ramp. Also the two ramps do not "mirror" eachother as I need them to.

Using slide~ nearly works, but as the below patch shows, the slide down has a "sharp edge" at the beginning of the ramp, and the slide up has a "sharp edge" at the end of the ramp.  Also the two ramps do not "mirror" eachother as I need them to.

signal-based-hann-function

A little less sophisticated attempt uses rampsmooth and slide:

Hi thanks for the replies and patches. The rampsmooth~ + slide~ combination is what I was trying to use to begin with, but it doesn't produce quite the right kind of curve. The tiny difference between it and an actual hann function is pretty small, but I was looking for the smoothest fade possible.

I've used parley's signal rate method, and it works brilliantly. I just had to adjust it to create a range of exactly 0 to 1. I've now sorted out the random distortion and completely eliminated clicks using the hann crossfading, so the patch is finally a useable concept!

Hi thanks for the replies and patches.  The rampsmooth~ + slide~ combination is what I was trying to use to begin with, but it doesn't produce quite the right kind of curve.  The tiny difference between it and an actual hann function is pretty small, but I was looking for the smoothest fade possible. 

I've used parley's signal rate method, and it works brilliantly.  I just had to adjust it to create a range of exactly 0 to 1.  I've now sorted out the random distortion and completely eliminated clicks using the hann crossfading, so the patch is finally a useable concept!

http://dl.getdropbox.com/u/288305/St.ools%20for%205.zip

signal-based "hann-function"?