0

I'm currently programming a tape delay plugin using C++ and the RackAFX dev kit for uni. After discussing with my professor, I found that the best way to go about this was to make a fractional delay with nonlinear saturation in the feedback path and to add some wow and flutter I would add some perlin noise to the delay.

My issue is, working out when and where to add the noise.

Do I add the noise and do I add it to the samples themselves or the delay variable for instance?

Asciant
  • 2,130
  • 1
  • 15
  • 26
Dimitris Christou
  • 1
  • This will likely get closed as it is not a coding question as such. It may be better to post this at the [Signal Processing Stack](https://dsp.stackexchange.com). From the sounds of things, you would be unlikely to add actual noise to your signal. What is more likely is the index at which you access a circular delay buffer would change based on some perlin noise. So, if you have a delay index `i`, it would be offset by some value `delta` (i.e. delay = `i + delta`) – fdcpp Dec 30 '19 at 21:43
  • For best effect you will have fractional delay values, which will mean interpolating between samples, linear, Lagrange or spline depending on how cheap / accurate you need to be. – fdcpp Dec 30 '19 at 21:45
  • The wonderful Dan Shiffman has a tutorial on [Perlin Noise](https://www.youtube.com/watch?v=Qf4dIN99e2w). You could also try heavily low passing white noise and using that as your base for offsetting the delay index – fdcpp Dec 30 '19 at 21:47

0 Answers0