You drag a piano part from an old project into a new project. Same notes, same timing, and yet it suddenly sounds completely different: the new instrument responds much more fiercely to your touch and the sustain pedal seems to do nothing. Nothing is broken. You are hearing exactly what MIDI is: a series of instructions, not sound. The part describes which keys are pressed when and how hard, and which control data goes with them. What sound comes out is decided by the instrument that receives those instructions. Once that clicks, your whole way of working changes: you know where to look when something sounds different than expected, and you know which layer to edit to make a part come alive.
What you will learn in this chapter
- Distinguish MIDI from audio and explain why a MIDI note has no sound until an instrument performs it.
- Recognize the core messages and use them deliberately: Note On, Note Off, velocity, Control Change, pitch bend, sustain and program change.
- Tell latency apart from jitter, and know which of the two you can compensate for.
- Place MPE and MIDI 2.0 in broad terms, without marketing language.
Who this chapter is for
For every producer who works with a piano roll or a MIDI keyboard. You need no prior knowledge; a controller is handy but not required, everything can also be done with the mouse. If you know Chapter 7 on latency and buffers, you will recognize a number of concepts. If not, we explain everything you need again here.
A preview: score and performer
Technical fact: a MIDI message contains no audio. Think of sheet music: the score does not sound, the orchestra sounds. MIDI is the score, your instrument is the performer. That is why you can polish a MIDI part endlessly without ever losing sound quality, because there is no audio to damage. And that is why the same part sounds different on every instrument. In the next lesson we work this model out fully, from individual messages to a workflow that lets you solve any MIDI problem systematically.