VSL goes into the studio and records a flute playing every possible note. All these note recordings are then saved into a single file.
When you press a key on your piano, VSL's software looks up what note you've pressed and plays black the recording of that note. When you release the key, VSL's software stops playback of that note.
MIDI, or Musical Instruments Digital Interface, is simply a standardized set of commands (a "protocol") that tells the software how and when to do this, i.e. start playing A#; stop playing A#; make A# louder etc. Thus, MIDI is not audio, it is a simply a set of instructions about how to play existing audio. The MIDI protocol therefore must encompass a variety of different actions which correspond to different aspects of sound making , such as:
- note on: start playing a note
- note off: stop playing a note
- velocity: how hard a note was played
Your keyboard speaks "MIDI-ese" and so does VSL's software too speak "MIDI-ese", and it's because they both speak this same language that they are able to work together to produce the coordinated dance of instructions that equates to a musical performance.
Now to your question:
Your sequencer -- whether Cubase, Garageband, Logic, etc. -- is simply is a recorder of this MIDI data. It sits in between your keyboard and VSL's software. When you play your keyboard, you generate MIDI data and your sequencer remembers this data which then passes it along to VSL's software. Us humans need some way to look at this data so that it makes sense to us. Sequencers typically displays it as "blocks", but there's no real reason that it has to look like blocks, except that it happens to be easy for humans to look at. So, yes, your sequencer generates these blocks. The start of a block corresponds to "note-on" and the end of a block corresponds to "note-off" and by looking at the length of the block we can get an idea of how long the note is.
When you think about it, traditional musical notation is no different than MIDI blocks. Traditional notation is just MIDI blocks represented in a different way. An eighth note is simply a "note-on" message that lasts one-eighth of a beat. In your sequencer, however, this eighth note looks like a short little MIDI block.
Both musical notation and MIDI blocks contain no audio, they are simply a set of instructions which indicate a musical performance for a receiving device. In the case of notation, the receiving device is a human being. In the case of a sequencer, the receiving device is VSL's software. Thus, your sequencer (being the recording device), is akin to what a blank sheet of paper is to music notation: a canvas on which to record and display data.
This is a gross over-simplfication but I hope it communicates the basic idea.
Best
MOH