Self resonating filter - impulse response length

[aciddose]

No.

A diode clamp is not going to be frequency dependent or have any of those sorts of effect in the audible range.

The thing that makes it marketing nonsense is the fact they claim there is some limit applied to the level of the signal, which is completely untrue. There is nothing limiting the level of the signal other than the level you pass through the filter. If you pass a large enough signal, it will absolutely result in clipping.

The amount of saturation in even the most saturated filters in common use is not nearly high enough to limit the peak signal level low enough to prevent clipping.

What makes it double marketing nonsense is that no engineer should ever have said anything remotely resembling this. A current limiter is a specific sort of circuit and it makes very little sense to use this particular term to describe what is going on, unless you want to simplify it to the point of being technically useless as a description.

To someone who has no idea though, I'm sure it fills its role as marketing.

[itoa]

I found another place for this phenomenon:
nonlinear one pole filter - integrator.. (difference of saturated input and buffer). This causes something like "cutoff reduction" for large signals.

[aciddose]

You can describe it that way, but it isn't accurate to do so.

What really happens is:

capacitor_charge += input_current - diode(input_current)

or, in pseudo-c (this works just the same if I recall?):

b += diode_curve(input - b) * c;

So "non-linear filter" is actually a spot-on description. Why should we use nonsense like "current limiter" or "cutoff modulation" that already have well defined meanings for something they don't even accurately apply to?