Location: Columbia, MD
True Mel, but what about attempt to use some sort of average "volatility" and/or combine some with "volume", to gauge some kind of proximity.
There has to be a manner or technique to going beyond in search of this process. For example, similar tuning to finding approx. "cycles" and removing noise, its a multiple step process.
For now, I am approx. fib. nos. for ticks to time, but, as you mention, its arbitrary, not a process.
For NTB Range, possibly using a "volatility"-based measurement for proximity to come up with a process. But I don't know how to formulate.
I am simply suggesting to thinking out of the box. I am sure there has to be a way :-).