Why static saturation isn't enough: the idea behind ODD ENGINE
Static saturation is honest about what it does: pick a transfer curve, run everything through it. ODD ENGINE started from asking what happens if the plugin is allowed to know what's coming in first.
One curve for every signal
Static saturation applies the same curve regardless of level or dynamics. A whispered vocal and a screamed one get identical treatment, at whatever level Drive puts them at. That is not a flaw exactly — it is what "static" means — but it puts the entire burden of matching character to material on the person turning the knob.
ODDDIST ships with that mode. Turn ODD ENGINE off and you get exactly that: one curve, bit-exact, zero added cost. It is a real setting, not a placeholder for the interesting one.
What changes when it's on
React mode reads the signal hitting the plugin, continuously, and uses that reading to steer the drive stage while it processes — not a pre-pass, not a lookahead, a control path running alongside the audio. The same Drive and Saturation settings end up behaving differently on a quiet passage than on a loud one, because the input itself is part of what decides the outcome.
We are not publishing the exact analysis method. That is deliberate — it is the one part of ODDDIST that isn't a straightforward implementation of ideas everyone already knows, and the product page says so plainly. What we can tell you here is the shape of the decision, not the formula.
Why bypass had to be real
An adaptive mode is only worth trusting if turning it off actually turns it off. If Off quietly ran a cheap version of the same analysis in the background, every A/B test against static saturation would be comparing two shades of the same thing. ODD ENGINE's Off mode skips the analysis path entirely: no CPU spent on it, no code executed for it, the audio takes the same route it always did.
One fader, not a rabbit hole
The adaptive behavior sits behind a single macro, ODD, from 0 to 100 percent. It doesn't add new destinations to reach; it scales how strongly React mode leans on what it hears, on top of whatever Drive, Tone and Saturation are already set to. That fader turned out to be its own design question, worth a post on its own.