Depth lives in the cable run
A modulation source — LFO, envelope, random voltage — produces whatever it produces at full amplitude. That amplitude is rarely the right amount to move a filter cutoff, a VCA level, or a pitch input. The depth of modulation, meaning how far the destination actually swings, is set by whatever sits between source and target. If nothing sits there, you get the full amount, which is almost always too much.
This is why an attenuverter or an inline VCA is often the most consequential module in a patch. It is not the oscillator, not the filter — it is the quiet scale factor governing how much the interesting thing happens.
Unipolar versus bipolar attenuation is where the real decision lives. A plain attenuator scales a signal toward zero; it cannot invert it or shift it below the zero line. This is fine for an envelope controlling VCA level — the signal is already unipolar, sitting between zero and some positive voltage. But an LFO swings both positive and negative. A plain attenuator reduces the swing symmetrically. An attenuverter — an attenuator with a polarity inversion built in — lets you scale and flip: useful when you want modulation to push a parameter down rather than up, or when mixing two sources and one needs to oppose the other.
The practical implication: when a patch feels uncontrollable, trace every modulation cable and ask what determines its depth. If the answer is "the source module's output level," that is a missing parameter. Inserting even a passive attenuator changes the feel of the patch more than swapping any sound source would. Depth is the control that goes unbuilt until you notice its absence.