What order should a vocal chain go in, and why?

The order, and the reason for each step

1. Pitch correction. Tuning works best on the raw recording, before anything has changed its level or added harmonics.
2. Clean-up. A high-pass filter at 80 to 100 Hz, and removal of noise, clicks and breaths that are too loud. Whatever is left in will be made louder by the compressor.
3. Subtractive EQ. Cut what should not be there: boxiness around 300 to 500 Hz, a nasal peak, a room resonance. The compressor then reacts to the performance and not to the fault.
4. Compression. Evens out the level so every word is heard. Two compressors doing 3 dB each are less audible than one doing 6.
5. De-essing. Compression raises sibilance along with everything else quiet, so the de-esser comes after it.
6. Additive EQ. Presence at 3 to 5 kHz, air above 10 kHz. Placed after the dynamics so the boosts stay as set.
7. Saturation. A little harmonic density to help the voice hold its place in a busy mix.
8. Reverb and delay, on sends. Space goes last and in parallel, so the dry vocal stays clear and the effects can be treated on their own.

Where the order bends

• De-esser before the compressor when the sibilance is severe enough to trigger the compressor by itself.
• Saturation before compression when the saturation is a sound, not a polish, and the compressor should control its result.
• No saturation at all on a vocal that is already dense.

The fixed points are these: corrections before compression, enhancements after, time effects in parallel.

In Zammoth

Tune Notes on the clip corrects pitch note by note before the signal reaches any device. In the track's chain, devices are freely ordered: drag a device's header sideways to move it. The De-esser, Compressor, Saturation and EQ are separate devices, and reverb and delay live on FX returns fed by the track's send knobs. Or ask for the whole chain:

The engineer measures the vocal first, phrase by phrase, so the chain is set for that recording. See the manual on the device dock.