Kaufman, Trading Systems and Methods, 4th edition, chapter 17 is full of good ideas about this. Two that are especially intuitive are the Kaufman Efficiency Ratio (p.732) and the residuals from a linear regression calculation, r-squared (p.739). But you could just as easily make your own definition, Price = Trend + Noise and then define "noise" to be "the absence of trend". This enables you to employ conventional measures of trendiness (such as ADX) and, in a sick perverted way, call them measurements of "the absence of noise". For example, Noise = (100 - ADX). When ADX is high (and the trend is strong), noise is low. When ADX is low, noise is high. You could do the same thing with RAVI or other trendiness monitors.