That’s what Roland thought many years ago. “Hey, we have the best pitch-to-voltage system in the world, so let’s take that voltage and convert it to a midi note.” After all, MIDI itself had been an idea from Roland’s president at the time, Ikutaro Kakehashi. (All of us musicians owe him a great deal of gratitude, by the way.)
The A4 note has a value of exactly 4 volts. So, an output of 4 volts = A4. Simple and easy, right?
No. That only happens with a keyboard.
When it comes to the guitar, it almost never ever outputs exactly 4 volts when one plays an A4. Then, to make it a midi note, the pitch bend value has to be added. Worse yet, even a loose string when played has a fluctuation of pitch, so the pitch bend value has to be continuous. Then velocity decreases. Then the player’s finger moves slightly over the string and new pitch bend values are created. Then the player touches the string with the left hand before it actually plays the note (which is almost always the case), so there’s a hammer-on, but wait! an actual note is coming right after, so you can’t play the same note twice. Then the instrument also produces noises that get output in the guitar signal. Then each string sound slightly different than the other, and each pick creates a certain shape to the wave, and every pickup has its own sound characteristics. Then… you get the idea.
Pitch to midi conversion relies on guessing a lot of the time for the sake of speed – or else they wait long enough to detect the actuall note, but that comes with a considerable latency. The best systems are the ones who guess better, and most use divided pickup to separate the sound of each string.
Therefore, your logic is correct, but implementing it is not that simple. And MIDI doesn’t help either, it’s a rather imprecise specification for a lot of things.
