Polyphonic Aftertouch: why isn’t it always represented?

[e-crooner]

I took classical piano lessons for 7 years before playing any synths. Obviously, the piano is mightily expressive without offering AT. So, since my synth day 1, I have never cared about AT.

Thinking about it, AT is also pretty rare with real instruments. Once one has hit the key, string, drum skin or whatever, usually the only thing one can still manipulate to a certain extent is the duration of the sound. Most of the player's sensitivity goes into the attack, not the sustain.

[Kevin63101]

Having owned an Ensoniq VFX-SD (1990) with poly AT, my main use was emphasizing a single note or chord I was sustaining without affecting other notes being held. (eg. Opening the filter or adding LFO only on the bass note of a chord or top melodic note). I never really triggered differing amounts of AT on each finger.

In practical use, AT is most beneficial when playing a monophonic solo -- therefore have not missed it much as mono AT suffices playing monophonic solos. I usually prefer to control modulation by multiple pedals anyways ... but mono AT is still on my keyboards. (I kept the VFX just for purpose as a Poly AT controller, but never once pulled it out for that purpose).

I would agree a MPE controller would have more useful features such as a Seaboard or Lininstrument ... but do not own either. I've tried a seaboard and its truly multi dimensional. Vids on Lininstrument are equally or more impressive too.

In short, MPE rules if you need expressive flexibilty. Poly AT on a standard keyboard seems like an unappreciated additional cost for most users. If you saw the mechanics of achieving, its a physical hardware difference to add poly AT to a keyboard.

[felis]

triple mess

[felis]

edit - not quote

[felis]

....
with MPE you are not stuck with mono-timbral, with roli (and i believe in other MPE controllers, like joué and sensel morph, and then the expensive ones), you can set the roli in different modes,
for the CS-80 V3, for example, i use single channel, change the setting for poly aftertouch. and it works.

so a MPE controller can also be used for non-MPE synths. articulation for a Mini V3, for instance, but one note, one sound....
....i know only about roli, because i bought the roli...

What if you play 10 notes, with 15 sounds layered across the whole keyboard (150 notes triggered), and each separate key is supposed to have its own midi channel, and you only have 16 channels. Something's got to give.

[zerocrossing]

polyphonic aftertouch, not many (midi)keyboards have it.
i think only the CS-80 V(3) responds to polyphonic aftertouch, of the arturia synths? pigments 2 perhaps, has certainly MPE.

MPE has always polyphonic pressure. but you have to have a MPE capable midi keyboard, the GEM S2 has a polyphonic keyboard. it is not a MPE keyboard.
MPE works different (with 16 midi channels, 15 for seperate notes, 1 channel for global).
polyphonic aftertouch are CC messages, 1 channel.

but it is a good question, why not implement poly aftertouch. it are only CC messages... but the synth must be capable to whatever the aftertouch controls/modulates, to do it per note. it is not only midi implementation, also the design of the synth itself.

Most of the Arturia synths respond to it.

[zerocrossing]

The main reason is, that implementing it needs per voice modulation. MPE synths can do it in principle, the others mostly not, and why bother if there are almost no keyboards which support it...
Of course any CS-80 emulation has to, I wonder if Arturia will add MPE at least to the CS-80 V. They support it in Pigments, they know how to do it, and it would be a logical enhancement...

I can’t think of a poly synth that doesn’t have per voice modulation.

[zerocrossing]

Same here, my keyboard does not support any AT, my previous one did but I never used it because my playing is just not good enough, I am already happy when I manage to hit the right keys

Maybe so, but there are a few synths and controllers that are now starting to feature it.

[WasteLand]

polyphonic aftertouch, not many (midi)keyboards have it.
i think only the CS-80 V(3) responds to polyphonic aftertouch, of the arturia synths? pigments 2 perhaps, has certainly MPE.

MPE has always polyphonic pressure. but you have to have a MPE capable midi keyboard, the GEM S2 has a polyphonic keyboard. it is not a MPE keyboard.
MPE works different (with 16 midi channels, 15 for seperate notes, 1 channel for global).
polyphonic aftertouch are CC messages, 1 channel.

but it is a good question, why not implement poly aftertouch. it are only CC messages... but the synth must be capable to whatever the aftertouch controls/modulates, to do it per note. it is not only midi implementation, also the design of the synth itself.

Most of the Arturia synths respond to it.

yes, but not full MPE, only pigments 2 has MPE. but you can use single channel (perhaps even how it is called, multiple channels).

The main reason is, that implementing it needs per voice modulation. MPE synths can do it in principle, the others mostly not, and why bother if there are almost no keyboards which support it...
Of course any CS-80 emulation has to, I wonder if Arturia will add MPE at least to the CS-80 V. They support it in Pigments, they know how to do it, and it would be a logical enhancement...

I can’t think of a poly synth that doesn’t have per voice modulation.

per voice modulation, so also the filter? effects? the complete routing? perhaps there are, or certainly synhts that can do it, but o well, i should investigate... maybe i don't get it.

....
with MPE you are not stuck with mono-timbral, with roli (and i believe in other MPE controllers, like joué and sensel morph, and then the expensive ones), you can set the roli in different modes,
for the CS-80 V3, for example, i use single channel, change the setting for poly aftertouch. and it works.

so a MPE controller can also be used for non-MPE synths. articulation for a Mini V3, for instance, but one note, one sound....
....i know only about roli, because i bought the roli...

What if you play 10 notes, with 15 sounds layered across the whole keyboard (150 notes triggered), and each separate key is supposed to have its own midi channel, and you only have 16 channels. Something's got to give.

i don't really understand what your are saying. 15 channels is MPE (1 channel is global). stacked voices, for one oscillator. you mean.
15 sounds layered across the whole keybaord, a sampler, some things synths can do it, but 15 sounds?
it is per key MPE. maybe my terminlogy is of... you strike a key on a roli, strike a next one, both send via seperate midi channels (2 = 2 keystrikes), velocity, pressure, slide, glide (and lift).
strike 4 = 4 midi channels.

[zerocrossing]

polyphonic aftertouch, not many (midi)keyboards have it.
i think only the CS-80 V(3) responds to polyphonic aftertouch, of the arturia synths? pigments 2 perhaps, has certainly MPE.

MPE has always polyphonic pressure. but you have to have a MPE capable midi keyboard, the GEM S2 has a polyphonic keyboard. it is not a MPE keyboard.
MPE works different (with 16 midi channels, 15 for seperate notes, 1 channel for global).
polyphonic aftertouch are CC messages, 1 channel.

but it is a good question, why not implement poly aftertouch. it are only CC messages... but the synth must be capable to whatever the aftertouch controls/modulates, to do it per note. it is not only midi implementation, also the design of the synth itself.

Most of the Arturia synths respond to it.

yes, but not full MPE, only pigments 2 has MPE. but you can use single channel (perhaps even how it is called, multiple channels).

The main reason is, that implementing it needs per voice modulation. MPE synths can do it in principle, the others mostly not, and why bother if there are almost no keyboards which support it...
Of course any CS-80 emulation has to, I wonder if Arturia will add MPE at least to the CS-80 V. They support it in Pigments, they know how to do it, and it would be a logical enhancement...

I can’t think of a poly synth that doesn’t have per voice modulation.

per voice modulation, so also the filter? effects? the complete routing? perhaps there are, or certainly synhts that can do it, but o well, i should investigate... maybe i don't get it.

....
with MPE you are not stuck with mono-timbral, with roli (and i believe in other MPE controllers, like joué and sensel morph, and then the expensive ones), you can set the roli in different modes,
for the CS-80 V3, for example, i use single channel, change the setting for poly aftertouch. and it works.

so a MPE controller can also be used for non-MPE synths. articulation for a Mini V3, for instance, but one note, one sound....
....i know only about roli, because i bought the roli...

What if you play 10 notes, with 15 sounds layered across the whole keyboard (150 notes triggered), and each separate key is supposed to have its own midi channel, and you only have 16 channels. Something's got to give.

i don't really understand what your are saying. 15 channels is MPE (1 channel is global). stacked voices, for one oscillator. you mean.
15 sounds layered across the whole keybaord, a sampler, some things synths can do it, but 15 sounds?
it is per key MPE. maybe my terminlogy is of... you strike a key on a roli, strike a next one, both send via seperate midi channels (2 = 2 keystrikes), velocity, pressure, slide, glide (and lift).
strike 4 = 4 midi channels.

I’m also after MPE, but that’s not really the same thing as polyphonic aftertouch. I don’t need every synth to be MPE... I mean, it would be nice, but that’s another can of worms. There are many types of sounds that I don’t need individual pitch control of. It always seems odd to me that Equator has things like piano style patches, as if one could pitch bend a piano or any sort of chime or tine type sound. Sometimes I just like the feel of a traditional keyboard. My Rise 49 is awesome, but not for everything.

As for per voice modulation, yes, filter, but many synths offer polyphonic LFOs and per voice EGs. Think of what you can do with an FM operator per voice. Tons of control. Equator has a little bit of FM and it can do amazing things. Anyway, I’m just saying that there are a fair amount of synths that do automatically understand polyphonic aftertouch. I guess I’m asking if this was a super difficult task to implement, or they just did it because they figured a few people would really appreciate it and it wasn’t that big of a deal.

[felis]

...i don't really understand what your are saying. 15 channels is MPE (1 channel is global). stacked voices, for one oscillator. you mean.....

OK - that's a bit much.

Say you have two sounds layered across the keyboard. You use both hands and play a 10 note chord.

You've just triggered 20 notes and each is supposed to have its own midi channel, so that polyAT can be applied to each note separately.
There must be something I'm still not understanding.

These two posts are what I'm having trouble with:

With MPE you could totally have 16 different sounds on each MIDI channel. Fun stuff.

....i believe it is 15, one midi channel is for global stuff.
but 15 different sounds, is already mayem...

Edit: took a quick peek at the MPE spec, and I think I get it:

"....An MPE controller assigns every new note its own MIDI Channel, until there are no unoccupied Channels available. An occupied Channel becomes unoccupied when its sounding notes have all sent or received Note Off messages.

When there are more Notes than unoccupied Channels, a new note must share a MIDI Channel with an existing note. Since Control Change and Pitch Bend are Channel Messages, they then affect both notes on that Channel...."

[WasteLand]

...i don't really understand what your are saying. 15 channels is MPE (1 channel is global). stacked voices, for one oscillator. you mean.....

OK - that's a bit much.

Say you have two sounds layered across the keyboard. You use both hands and play a 10 note chord.

You've just triggered 20 notes and each is supposed to have its own midi channel, so that polyAT can be applied to each note separately.
There must be something I'm still not understanding.

These two posts are what I'm having trouble with:

With MPE you could totally have 16 different sounds on each MIDI channel. Fun stuff.

....i believe it is 15, one midi channel is for global stuff.
but 15 different sounds, is already mayem...

Edit: took a quick peek at the MPE spec, and I think I get it:

"....An MPE controller assigns every new note its own MIDI Channel, until there are no unoccupied Channels available. An occupied Channel becomes unoccupied when its sounding notes have all sent or received Note Off messages.

When there are more Notes than unoccupied Channels, a new note must share a MIDI Channel with an existing note. Since Control Change and Pitch Bend are Channel Messages, they then affect both notes on that Channel...."

here is more information:

https://support.roli.com/support/soluti ... at-is-mpe-

[tapper mike]

AT is not a global parameter. AT is singular per channel as is bend.. The Yamaha CS-80 was not standard midi. Poly AT requires sending signals across multiple channels. IE MPE.

Poly AT is not genre specific, nor is AT common on many midi keyboard devices due to cost constraints. In many situations it's more suitable to assign that function to an expression port such as a knob or an expression pedal.

Although sysex is transferred on channel one along with other data it is still capable of sending various local data as well including note on etc. Channel 10 for drums was simply instituted for simplicity and instant compatibility with other devices. In short all midi channels can side chain proprietary data along with specific midi per channel midi data.

I have a linnstrument and I've done several workarounds to attain mpe with non mpe enabled devices. It's work. It's a lot of work. Mostly it's about creating several channel inputs with the same instrument set up with their own midi channel ins.

A great deal of thought has to go into whether you want mpe or not and what you can assign to the various cc values. While one may not be able to fake at on a non at hardware output one can with reason assign cc values to a specific "group sound" Not all virtual instruments play nice with MPE. Just as some sounds only work well with mono rather than poly output.

The development of mpe was in development long before the phrase came about. The greatest hindrance was as it is now in Development cost and hardware production costs. It's still expensive (especially on hardware keyboards) to configure all the wiring, components processing and more. Then there is the market adaptability. If the wider market refuses to see the added value for the added cost then there will be limited production. We see that with the market here at KVR as well. Every time someone says "I want cheap" the market will respond with cheap. That's the only way the market can stay in business.

If the public really wants mpe, it's out there buying into mpe will eventually lower prices for mpe but it still wont be cheap. There are some inexpensive mpe devices on the market... Namely the artiphon instrument one. But the truth of the matter is that the instrument one was priced out cheap and most features of mpe were not implimented. As well it has serious design issues. (Yes I have one)

The best mpe enabled device on the market is... The Linnstrument.
Yes: I have one, Yes I have a fb group dedicated to it and I try to offer lessons on the FB Linnstrument Lessons page. It costs money, The greatest limitation of it is familiarize yourself with the physical act of playing witch takes time and practice. Not for the emotive values as much as getting down chords and scales. I've seen a few go on the used market simply because buyers who wanted all the power that the linnstrument has to offer weren't willing to make the commitment to learning how to play the thing. Regardless of your prior instrument experience (I play keys, guitar, bass, drums etc) you will need to commit yourself to at least the basics. If you want it, it's out there but it's going to take more than wanting or wishing and after you have it... It's still going to take effort to maximize your potential.

[tapper mike]

Also important to note. Roli has it's own proprietary method of dealing with mpe as does RogerLinnDesign and other brands claiming mpe capabilities. Which is why they all supply their own drivers.

[Noumena]

I disagree that the Linnstrument is superior to the Roli. I've had both and I've found that for me the opposite is true. I feel that it is true that one has to spend the (considerable) time with either to clear to play them. I didn't feel more comfortable on the Roli for more than a year... and I sold my Linnstrument (to fund a Seaboard 49) a year after that. I'm not saying that the Roli is better... just that it is a matter of preference.

There are plenty of examples to be found online of excellent playing that takes advantage of Poly-AT, and it is used very extensively on one of the most popular electronic music recordings of all time: The Blade Runner soundtrack. Vangelis used Poly AT to control amplitude and ring mod on patches with a slower attack.... which, if you think about it is something that exists in synthesis but not in non-electronic keyboards.

I find poly-AT to be most useful on the Roli to control amplitude. I program patches that map AT directly to amplitude, bypassing envelopes and velocity completely. This enables my to swell and fade with very precise control... having worked on the technique quite a bit I find that I can play in a way that is as expressive as playing a woodwind (which features very precise control of amplitude and *subtle* pitch modulation) polyphonically.

But all this is about performance. Vangelis was improvising and recording straight to tape, and this is the way that I play, too. If you're someone that isn't focused on performance the obsessive nature of people that are interested in more kinds of realtime control might seem misplaced. For me, it's all about performance, and playing live and having this level of control puts me in a whole different place than playing note on + velo triggering an envelope.

Last thing I'll note is that MPE is bringing about an age of synths that allow control of parameters per note, not just modulation. This is huge. In Bitwig I can assign lfos, sequencers, envelopes, random mods and combinations of these in the HOST to a VST synth *polyphonically* (i.e. a different host based LFO for each note played) because of poly AT and MPE. This is a very different way to work than having to rely on the mod sources and setup provided by the synth's developers. And with VCV, Voltage and Bitwig Grid supporting MPE things are really going places they have never gone before. Polyphonic, expressive modular synthesis.