Channel-per-note and Voice-per-channel?

Official support for: rogerlinndesign.com
RELATED
PRODUCTS

Post

I'm trying to understand the terminology here. "Channel per note" is what the LinnStrument can do. "Voice per channel" is what a synth must do to properly respond to "Channel per note". Is that right? or what?

Post

It's actually the same thing, we thought channel per note would be easier for musicians to understand. There's no official way to name this, but voice per channel seems to become the most popular one.
Moog software - LinnStrument - RackBlox - CableCube - Knobotron - Eigenharp Alpha/Tau/Pico - SendMIDI / ReceiveMIDI - MIDI Tape Recorder - Geco MIDI Leap - Steelstring Guitar - Electric Guitar - Vocals - Dynamod Games - Kung-fu

Post

Thanks, Geert. In my mind, they are two sides of the same coin, but not the same (since the LinnStrument knows nothing about voices, as I understand it).

Post

"Channel per note" and "Channel per row" seem pretty clearly descriptive of the two multi channel LinnStrument settings, but "voice per channel" does not seem clear to me at all. In Logic Pro, the corresponding setting, "MIDI Mono Mode", seems to provide multiple copies of the same voice spread over multiple channels, one note at a time (mono) per channel.


That seems rather confusing and difficult to find because it (MIDI Mono Mode) is hidden in the Logic Pro UI as an extended parameter of individual synth engines, such as ESX24, and it is obscured in the documentation by "keyboard mono mode".

Post

its a pity there isn't some standardisation, especially as if you are talking to a synth developer you have to describe it each time... to ensure there are no confusions.

I guess the note per channel or voice per channel, comes from your perspective... as a controller implementation, notes makes more sense, but from the synth side, its more about voices. ( in fairness, the midi specs also talk of voices)

I prefer voice per channel, as notes per channel kind of misses a subtle (or perhaps pedantic ;o) ) point, which is your are trying to get independent control of the voices, and that is actually not the notes at all... but the other messages (CC/pitchbend/channel pressure).
(but i can see why others prefer note per channel, as being more 'obvious')

oh well, its not like the only inconsistency, there was a similar topic here (on kvr) recently, debating poly pressure/ poly aftertouch... with arguments about if it was channel pressure and poly aftertouch (the midi specs say pressure=aftertouch its the same)

Post

Well technically voice-per-channel is more correct, since as soon as you exceed the number of MIDI channels that you reserve for this, you'll start reusing MIDI channels for additional notes. So imagine you have 4 MIDI channels reserved, as soon as you play the fifth simultaneous note, it will reuse the first channel. Mean that you don't have a different channel for each note anymore.
Moog software - LinnStrument - RackBlox - CableCube - Knobotron - Eigenharp Alpha/Tau/Pico - SendMIDI / ReceiveMIDI - MIDI Tape Recorder - Geco MIDI Leap - Steelstring Guitar - Electric Guitar - Vocals - Dynamod Games - Kung-fu

Post

yeah, I knew eigenD that, I wasn't sure about the linnstrument, so didn't want to confuse matters (in case it didn't :))

of course, in practice, in that scenario, voice per channel is also still 'misleading' as the synth will then have multiple voices per channel.
(e.g. if it has 32 voices, we might have 2 voices per channel ). but i guess we mean 'independently controlled voice per channel'

not that this ever happens for me... 16 voices is plenty :)
or we all just move to OSC :o

Post

It's my understanding that the main reason Apple added MIDI Mono Mode was to permit the On-screen keyboard in Logic Remote (the iPad app) to do poly bends with multiple touches, which is presumably why it's tacked on at the bottom of each synth. Regardless of the reason, it's very helpful for LinnStrument.

Post

I think I just found a counter example that shows that voice-per-channel is technically incorrect, a channel-per-note Reaktor ensemble, called "Two Junatiks with Y Mix" (http://www.marksmart.net/instruments/co ... aktor.html), that varies a mix of two voices per touch depending on Y-axis values. This ensemble is built on "ContinuumFrontEnd", a Reaktor macro also by Mark Smart, that "can be inserted into any Reaktor synth with 8-note polyphony" to support 8 channels of channel per touch. Thanks to technobear for sharing this on G+ Eigenharp.

Post

Roger, thanks for mentioning Logic Remote. That seems to be a useful and fun addition to Logic Pro.

Post

in regards to channel per note: is there a way to do a split of one side of linnstrument in mono legato mode while the other is left for polyphonic playing? is it as simple as only leaving one channel active on one side of the split while the other split has all asscoiated channels assigned for polyphony?

Post

LinnStrument can do it fine, but it depends on the synth(s). In LinnStrument, simply set one split to Channel Per Note mode and define which channels to use for the notes; then set the other split to One Channel using a channel that's not used in the other split. Then set 2 synths to match those settings, assuming the one playing the left split has Channel Per note capability. Unfortunately Logic Pro X can't do this currently without digging dig into the environment because a track currently must receive on any single channel or all channels but no choice of multiple channels.
Next month the 5 main Polyphonic Multidimentional Controllers (PMCs)-- LinnStrument, Continuum, Seaboard, Eigenharp and SoundPlane--are announcing a new MIDI Polyphonic Expression specification so that any synth that includes it will be able to have plug-and-play polyphonic expression over MIDI, and permitting 2 splits each with polyphonic expression.

Post

sounds interesting... could you possibly briefly outline what this covers?
Is this going to be based on the existing Continuum midi messages? also is it going to be 14 bit? and with 21bit extension for pitchbend... as per continuum?

the soundplane, linnstrument, and eigenharp ... already are using the continuum messages (or can be configured too), so its already a de-facto standard.
the exceptions really being, 14bit only supported by Eigenharp/Continuum, but should be supported by all given sensor resolutions. (21bit PB is required for large surfaces like Continuum to ensure reasonable resolution)

Also will this 'standard' be submitted to the MMA? and some how be related to the MIDI HD spec (should it ever arrive!)

Im very interested, as Ive written software for the Eigenharp which already supports this various PMC devices with EigenD, I think it would be nice if interested parties were given some information, rather than doing in private like the MMAs tendency.

Cheers
Mark

Post

Coming from a Midi Guitar background where we have had to suffer all this for years there is part of the midi standard and it's called "Channel Grouped MIDI mode 4 (Omni Off/Mono)". Doesn't roll off the tongue as easily as channel-per-note though!

In this mode the Mono On midi message allows you to set the number of channels, so for example for a midi guitar this would be 6. If you set it to 0 then the synth is meant to use as many channels as it has voices.
Bitwig, against the constitution.

Post Reply

Return to “Roger Linn Design”