nice... and back to the topic too
but maybe you should use a function for the noise part which has a distribution function that doesn't focus so much on 1 and -1. like a triangle
i tried expr~ $v1*1+0*abs($v1) after the noise sin and think it sounds warmer
and maybe a second control to soften the generated noise at lower noise values
because nobody understands what i've written here i made a quit and dirty patch