MIDI Polyphonic Expression has been updated with one powerful feature: better documentation

At the beginning of May, The MIDI Association quietly announced MPE 1.1. What’s new? Absolutely nothing. It’s the same spec. So why the new version number? They listened to your questions, issues, and feedback, and made the whole document easier to read, understand, and implement. And that might be better than any new feature.

For those of you just joining MPE, the idea is simple: instead of a pitch bend or aftertouch or vibrato impacting all keys at once, MPE can assign per-note expression for polyphonic instruments. It’s actually a bit comical to me that people wonder whether this is important. That’s easy to answer: if it’s a monophonic instrument, totally not, if it’s a polyphonic instrument, totally it doesn’t make any sense without it. Imagine if you added vibrato on one string and all the other strings… yeah, basically the way electronic instruments have conventionally worked makes sense to keyboardists and organists and absolutely no one else. Enter MPE, and you have one very easy-to-implement solution.

What can it do? Things like this (pictured at top):

LinnStrument and Moog Model 15 go perfectly together

But technologies don’t become reality becomes someone published a paper or defined a spec. They get hashed out by humans. Humans have questions. Questions need answers. Enter MPE 1.1 – enabling better interfacing with those humans.

MPE Working Group, with representatives of the MIDI Association, Moog, Yamaha, Keith McMillan Instruments, and others – including our hero dev and MPE evangelist Geert Bevin – met and made this happen.

There is literally not a single technical change here, but that’s okay – MPE is working reasonably well where it’s implemented. What could matter to users is, better documentation means it’s more likely you’ll see MPE in software and hardware, and implemented in a more consistent way.

As the authors put it:

Every point of confusion or misinterpretation we’ve encountered over the past few years has been addressed. All the implementation rules are now gathered together into section 2 “Detailed Specification”. Further, the best practices and appendices have been elaborated to provide greater detail and examples.

I almost didn’t post this when I got that far, but then started browsing through and thought to myself, self, uh, no, this is exactly the time to talk about MPE, especially knowing various hardware and software engineers read CDM. So it’s well worth a perusal of the updated documentation, especially if you looked at MPE in the past and didn’t get anywhere. Check it out from

They’ve also given a clue to what they’ll work on next – auto-configuration between MPE senders and receivers. It all uses MIDI-CI Profile, a gee-whiz feature that we saw with MIDI 2.0 that sounds great (instruments automatically “knowing” how to work with other hardware or software), but that we haven’t really yet seen in action. So this is potentially exciting, especially as MPE is a bit more involved than plain-old-vanilla MIDI. And they do promise it’s coming in 2022, which (checks note) would be this year.

The announcement

And check some Geert TV – though, of course, the next star of MPE could be you. Aherm. Okay, let’s watch:

Source link

Share this post