Aftertouch, polyphonic aftertouch and MPE: what key pressure really does
Channel aftertouch, polyphonic aftertouch and MPE: three ways to send finger pressure, three different promises. What to check before you buy.
Three terms on a spec sheet look alike but do not promise the same thing: aftertouch, polyphonic aftertouch and MPE. All three are about what your finger does after a note starts, once the key is already down. In recent weeks, the MicroMonsta 2 firmware 3.0, the keyboard on the Behringer Polyeight and the FabFilter Twin 3.10 update have all made it a selling point. This guide draws on the MIDI Association’s reference for MIDI 1.0 messages and on its MPE page.
Channel aftertouch: one pressure value for the whole keyboard
Aftertouch is the pressure you put on a key that has already bottomed out. In its most common form, the keyboard sends the MIDI Channel Pressure message: a status byte and a single value, from 0 to 127, for the whole MIDI channel. Play a chord and press harder on one note, and the whole chord responds: the filter opens or vibrato comes in on all three notes at once.
This is the type you get on most keyboards that simply list “aftertouch”. It is all you need on a monosynth, where only one note plays at a time, or to add vibrato to a pad played without much finesse.
Polyphonic aftertouch: one pressure value per key
The Polyphonic Key Pressure message carries two pieces of data: the note number and its pressure value. Each key has its own curve. In the same chord, you can open the filter on a single note, swell the fifth and leave the third untouched. The message has been in MIDI 1.0 from the start, but few keyboards generate it because it needs a sensor under every key, which is what makes the Polyeight’s 49-key keyboard notable.
The synth also has to receive it. The MicroMonsta 2 gained it with firmware 3.0: a preset set up for aftertouch responds to channel and polyphonic aftertouch alike.
MPE: one MIDI channel per note
MPE (MIDI Polyphonic Expression) is a MIDI Association Recommended Practice, now at version 1.1. It works around a MIDI 1.0 limitation: pitch bend and most controllers apply to a whole channel. MPE gives every note its own MIDI channel, so anything sent on that channel affects that note alone.
In practice, an MPE controller sends three gestures per note:
- pitch bend, often from sliding a finger sideways: vibrato or a glide on one note of a chord;
- pressure, sent as Channel Pressure on that note’s channel;
- a third dimension, CC74, often labelled “Timbre” or “Slide”, driven by moving the finger forward and back on the key.
The keyboard is split into zones. Each zone has a manager channel (channel 1 for the lower zone, channel 16 for the upper zone) that carries shared messages such as the sustain pedal, plus member channels assigned note by note. One zone can therefore hold up to 15 independent notes. By default, member channels use a pitch bend range of 48 semitones, so a slide from one key to another lands in tune. On the hardware side there are dedicated controllers such as the Roli Seaboard, the Expressive E Osmose and the Linnstrument, and compatible synths such as FabFilter Twin since version 3.10.
MIDI 2.0 builds per-note controllers and pitch bend into the protocol itself. MPE remains the method that works over MIDI 1.0, which is what the vast majority of gear in use still speaks.
The three modes side by side
| Channel aftertouch | Polyphonic aftertouch | MPE | |
|---|---|---|---|
| MIDI message | Channel Pressure | Polyphonic Key Pressure | Note, pitch bend, pressure and CC74 on one channel per note |
| What varies per note | Nothing: one shared value | Pressure | Pitch, pressure, timbre |
| MIDI channels used | 1 | 1 | Up to 15 per zone, plus a manager channel |
| On everyday keyboards | Common | Rare | Dedicated controllers |
| Check on the synth | Almost always received | Confirm reception in the manual | MPE mode on, enough voices |
Before you buy: four things to check
- What the keyboard sends. “Aftertouch” on its own almost always means channel aftertouch. Polyphonic aftertouch and MPE are always spelled out.
- What the synth receives. A polyphonic aftertouch keyboard does nothing for a synth that only reads Channel Pressure. The MIDI implementation chart at the back of the manual tells you in one line.
- Where pressure can go. Pressure is only useful if you can route it to the filter, the level or an LFO. A modulation matrix, or at least an aftertouch destination, is the minimum. The envelope, covered in our ADSR guide, shapes the attack; pressure takes over while the note is held.
- The voice count. Per-note expression only makes sense on a truly polyphonic synth. A paraphonic synth shares one filter across its notes, so it cannot open the filter on just one. Our guide to polyphonic, paraphonic and monophonic synths explains the difference.
Which one you need
- Monosynth, bass, lead: channel aftertouch is enough, since only one note sounds at a time.
- Expressive pads and chords on a traditional keyboard: polyphonic aftertouch, with a synth that receives it.
- Playing closer to an acoustic instrument, with per-note vibrato and glides: MPE and a controller built for it, after checking that the synth or plugin supports it.
For beginners, all this comes after choosing the type of synth, which our guide your first synth: analog, digital or hybrid covers.