Harmonics, warmth, and clipping character.
A perfectly clean digital signal contains exactly the frequencies you put into it. Nothing extra. Saturation adds frequencies that were not there before - harmonics generated by gently distorting the waveform. Push a signal into tape, a valve amplifier, or a transistor circuit, and the peaks of the waveform get softly rounded off. That rounding creates new overtones. The sound becomes richer, thicker, more present.
This is the difference between a raw digital recording and one that feels "warm" or "analogue." The warmth is not magic. It is extra harmonic content, particularly even-order harmonics (the 2nd, 4th, 6th), which the ear perceives as musical and pleasant. Every analogue recording ever made has some saturation baked in, because the tape and the circuitry themselves add it at every stage.
The machine below lets you control exactly how much and what kind of saturation gets applied. Press play and start with the Drive control.
A clean sine wave has smooth, rounded peaks. As you increase drive, those peaks get squashed flat - the waveform starts to resemble a square wave. That flattening is the distortion, and the new corners it creates are the extra harmonics you hear.
Drive controls how hard the signal is pushed into the saturator. At low drive, you get a subtle thickening - a few extra harmonics hovering just below perception. Crank it up and the waveform starts to visibly flatten at the peaks. More harmonics appear. The sound gets grittier, denser, louder in character even before you touch the volume.
Warmth is a low-pass filter on the saturated signal. It rolls off the harsh upper harmonics that heavy saturation creates. Low values let everything through, including the fizzy, brittle overtones. High values keep only the lower harmonics - the ones that sound round and warm rather than aggressive.
This is the difference between tape saturation (naturally dark, because magnetic tape absorbs high frequencies) and transistor clipping (brighter and edgier).
Clip changes the shape of the distortion curve. At low percentages you get soft clipping - the waveform peaks are gently bent, like tape. The transition from clean to distorted is smooth. At high percentages you approach hard clipping - the peaks are chopped flat, like a signal overloading a digital converter.
Hard clipping sounds harsher and more buzzy because it generates odd-order harmonics (3rd, 5th, 7th), which the ear reads as dissonant and aggressive.
Mix blends the clean (dry) signal with the saturated (wet) signal. This is called parallel processing. At 0% you hear only the original. At 100% you hear only the saturated version. Somewhere in between, the clean signal preserves the transients and clarity while the saturated signal adds body and harmonics underneath.
© ectoplasma.org