Reshape the waveform of an audio file to add harmonics that were not in the source. Three curves are available, each with a different character.
The three curves
Soft clip passes the signal through a hyperbolic tangent, which rounds the peaks progressively rather than cutting them. Harmonics build in gradually as drive rises, which is the behaviour valve circuits are usually described as having.
Hard clip flattens anything beyond the threshold and leaves the rest alone. The corner it creates in the waveform generates stronger high-order harmonics, so the result is sharper and more aggressive than soft clip at the same drive.
Foldback reflects the signal back down when it passes the ceiling instead of flattening it. The reflection is not harmonically related to the input, so the result sounds synthetic and works better on sound design than on music.
Drive and output level
Drive scales the signal going into the shaper, from unity up to 31 times. More drive pushes more of the waveform into the shaped region.
Soft clip divides by its own response at full scale, so a signal that arrives at full scale leaves near full scale regardless of drive. Hard clip and foldback are bounded by their shapes. Peak level stays controlled in all three cases, but the average level rises as the waveform fills out, and that reads as louder.
Tone
A single-pole low-pass filter sits after the shaper, sweeping from 400 Hz to 12 kHz. Distortion generates most of its new content in the upper register, so the tone control is what makes the difference between a warm result and a harsh one.
The filter runs on the shaped signal only. The dry side of the mix is unaffected, so lowering tone darkens the distortion without dulling the source underneath it.
Aliasing at high drive
Reshaping a waveform creates harmonics above the Nyquist frequency, and there is nowhere for them to go. They fold back into the audible range as tones that bear no relation to the source, heard as a gritty, slightly metallic edge.
The effect runs at the file’s own sample rate without oversampling, so this is most audible with high drive on bright material such as cymbals or sibilant speech. Lower drive, a darker tone setting, or a source with less high-frequency content all reduce it.
Mix
Mix balances the shaped signal against the original. At 0 the audio passes through untouched and the shaper is skipped. Around 50% the distortion sits behind the dry signal, keeping the attack of the source while adding harmonics underneath, which is often more usable on a full mix than the fully wet setting.