In this video, composer Josh Madoff discusses MIDI Polyphonic Expression (MPE), a standard that enables per-note expressive control over pitch, volume, timbre and more.
Topics covered:
00:00 Intro
00:15 What does MPE mean?
00:28 A simplified history of the keyboard
02:00 Hear MPE in action
02:43 MIDI vs MPE
04:15 MPE & software synths
05:54 MPE & hardware synths
07:42 Discovering the MPE arpeggiator
08:40 Creating an MPE super synth
09:52 Different kinds of MPE instruments
11:25 MPE playing examples
12:55 The benefits of MPE
…and for those with a 6th grade reading level… a video! myyyystreyyyyy….
Wow! Nice! I recently saw an excellent Jordan Rudess demo of the Lumi & its infrared controller add-on. Between that and Josh’s piece, I think we have a near-ideal pair of real-time descriptions that answer “What is MPE?” If you still don’t get it then, buy a kazoo.
I think it is important to note that the original MIDI 1.0 Specification had polyphonic aftertouch, and several early MIDI keyboards have provided this. It’s only recently that it has made a resurgence.
Any history of MPE should include the fact that MIDI 1.0 uses 16 channels for multi-timbral purposes. And that MPE involved sacrificing that original intended purpose, and did a clever workaround. In essence, where the 16 multi-timbral channels originally were to allow a different synth sound/device on every channel. Where MPE cheats the system and instead uses those channels to separate out one synth voice per channel. So you lose multi-timbral capability, but gain polyphonic expression.
And finally, I think it is also worthy of mention that the new MIDI 2.0 spec would allow you to have multi-timbral and polyphonic expression without having to sacrifice one for the other.
Although I think it’s always good to have more options, I think in practice having a controller with polyphonic aftertouch is more than enough.
One of the issues with even monophonic aftertouch is, that you need to remember what the effect of using it per patch. For one patch it would open the filter, for the other it would speed up the LFO, etc. Now, you will also have to remember what happens if you wiggle a key.
MIDI 1.0 also supports Note-off velocity. I’m not sure if this is implemented in controllers at all. And if anybody uses it. I guess not, because it is too confusing.
Many of the expressive examples you see in the video above can be easily accomplished with good old MIDI 1.0, polyphonic aftertouch, a pedal, mod wheel and pitch bender.
And some good patch programming.
I think there are plenty of keyboards with note-off velocity. Hydrasynth de luxe and Explorer, and Deepminds, (though the latter doesn’t transmit polyphonic aftertouch) to name a few. But yeah, I don’t know of any midi controller, that transmit note-off velocity
I agree that polyphonic aftertouch is adequate, unless you need maybe two dimensions of control– like press to open filter, wiggle for vibrato.
One drawback ANY kind of aftertouch is that all CC values must start and end at zero; i.e., start at 0, ramp up to some value, do some funny business, then let go and it goes back to zero before the note ends. There are times when having the CC already at a higher value when you start or end is useful. Which is why breath controllers and pedals can be nice.
As for release velocity– in my experience, release velocity is fairly common in controllers, even cheap ones– even when they don’t mention it in their specs. This is because the same switches that measure attack velocity can measure the release. It is very useful for thing like having long release times for low release velocities, or louder release-thuds for piano sounds when the key-up is fast. Or even for timpani rolls where you assign notes to trigger on both note-ons and note-offs (works really well).
Release velocity is an unsung hero. It feels like MPE on its own, to some extent. Being able to define the total sound event with a more measured touch is good for everything, but especially synth sounds. It’s fun to just lift your fingers more slowly and have the sound blossom & fade naturally. Try it with bowed strings. It’ll teach you a lot in just one sitting.
I guess release velocity is nice when combined with the fx section.
Lift hand quick – play lots of delay …
Even just assigning release-velocity to the release time of the AMP envelope, or filter is quite useful. And if I have sounds assigned to trigger on key-release, then release velocity can control it’s volume (at least I can do this with a Kurzweil keyboard).