THIS IS ABOUT MIDI - NOT AUDIO

Let's say i have one 4 voices midi track.
And i want it to be played by 4 monophonic instruments

how can i easily edit (convert) the polyphonic midi track to 4 monophonic tracks.

the track is quite long so manual editing (piano roll) is out of the question.


[ PC - Win7 - Ableton Live ]

 

when there is space between the voices (each voice has their own range of notes without overlapping with other voices) you can use an instrument rack, there you can set the noterange for every chain

also possible with only 1 chain and 4 instrumentracks in their own track with their own noterange

 

If you want to "explode" polyphonic to monophonic, check out Xhip NoteLogic VST plugin:
http://xhip.net/effects/releases/ad_notelogic_b1.zip
http://xhip.net/effects/
Split up into as many voices as you need, each new voice becomes a new midi channel (so 16 is maximum).

 

@GanjaRa : wont do , the track is clusters chord

@Baxter : Not much information on that one. Nothing on Xhip site, google or youtube ...
i'll have to wait and test that on old PC ( just in case)
Don't even know if its 32 or 64 bits ( i know, jbridge will tell me ... )

Thks Guys

And Yes i want to explode the track, but it would be nice to use it in live performance too.
like playing chords with vsl solo string or any brass from Sample Modeling

 

then learn max and build a chordsplitter in max for live (maybe there is allready a similiar patch in community)

but im sure, splitting by hand is faster ;)

 

If you delete unwanted notes its usually a lot faster than copying, so manual method ain't too bad

 

pizmidi midiPolyphony - For spreading polyphony across MIDI channels, i.e. if you play 4 notes, each one will go to a different channel. Polyphony per channel can be set from 1 to 16 notes for more flexibility. You can also use it to convert polyphonic input to monophonic output.
http://www.thepiz.org/plugins/?p=pizmidi

 

http://xhip.net/effects/?p=Notelogic

Sorry, this really needs to be improved.

It is far faster than manually allocating voices and it does so in a way which would be nearly impossible for a human to keep track of. It doesn't merely "split" a part, it operates exactly like any polyphonic synthesizer would when allocating voices. In fact it is the voice-allocator from Xhip!

In the near future I'd like to add some additional features such as handling glide mode by sending MIDI CC to switch it on and off. Ideally any good MIDI->CV should have a glide pedal output jack or similar, or perhaps implement glide by processing the CV itself. In order to set the glide start position for a voice, immediately before turning it on the glide can be disabled and the start position sent as an "overlapping" note. After this, both notes can be shut off and the glide re-enabled and new note triggered.

Such a feature will allow the same glide modes as presented in Xhip: Follow, Retain and Voice mode. Legato is another feature I'd like to add once this is working.

In my opinion doing this would be absolutely insane. In fact I spent so much time trying to do this sort of thing by hand I wrote a whole plugin just for my own use!

 

this could be helpful...

https://www.youtube.com/watch?v=Pvoxrk8uot8

 

An example of what we're talking about here:
https://soundcloud.com/xhip/megamonopoly

It isn't excessively difficult of course to manually arrange tracks while you're composing. Of course you'd just treat the instrument as monophonic and have the "flute" play one part, the "bass" another and so on.

That isn't what we're talking about in this thread, though. The issue is recording a polyphonic part from live performance or taking a polyphonic sequence from some other source (a midi sequence as in the soundcloud link) and automatically allocating the notes in the sequence to "voices". We're not splitting by pitch, not by sound or octave or anything like that. This isn't really possible to do manually unless you spend countless hours keeping very careful track of things, essentially you are becoming the computer executing the program.

Now the problem should be clear. We're not computers. At least not all of us...

We're human, we're prone to error and have trouble keeping track of hundreds of different variables in our minds all at once. I believe the common understanding is that human capability is typically limited to from eight to twelve "objects" in memory.

This may sound fine, you might say "well then we can handle eight voices all at once no problem, right? Sure we'd be doubly-sub-sub-turtle, you know like some sort of xzibit turtle speed but we have all the time in the world, no?"

Well, no unfortunately usually we can't. You have to take into account the amazing overhead of this process. A "voice" isn't a single object or variable, it has a pitch associated with it as well as velocity, age and potentially other parameters such as polyphonic pressure. (Which isn't yet supported.)

Then you have the additional overhead of trying to remember keyboard or mouse maneuvers while you're trying to execute this horribly complex task. "Okay, this goes to voice six, now wait which channel was that? Ah, I have to name all my tracks like analog bass mono 6." This sort of thing may only enter into the mind for an instant but it removes whatever element was already stored in that channel of your memory. Often the result is corruption rather than complete loss which results in confusion, which leads to error.

Now, error is a beautiful thing. Art is all about errors and the human touch. When you're trying to remember which note goes to what channel with what velocity and so on however you don't want to get all mixed up. This is frustrating, it requires a massive amount of energy to do assuming it is even possible, let alone practical to actually do on any significant scale.

Enter the computer... a device we use for computing. You know, that sort of thing I just finished describing which we're really terrible at doing ourselves.

Now you can perform that task in real-time as you input notes from the keyboard and send them to multiple monophonic synthesizers. Oh my gosh we've just invented the polyphonic synthesizer.

 

To me the question sounds more like "i've got a midi drum track, how can i make separate voice for snare, kick, etc... ?"

So the answer is really simple, and maybe ableton (like Cubase does) can split the MIDI channel during the import or export.
It's the easy way to split midi... velocity and quantization SHOULD remain the same, the only thing to re-set is the tempo track

By the way, maybe I'm completely off the tracks! But the question itself sounds more simple than the biggest replies that it gets...

 

The original post seemed like he was talking about splitting parts, although we don't know what those parts are as the post didn't go into much detail.

My impression was actually that the part would have a mixture of different notes and that only four could occur together at any one time. That means we still need a voice allocator to do the work for us if we want to assign the voices in any complex way.

The allocator works by allocating to the "oldest" voice for each new event, unless the event is the same note played multiple times in which case it is assigned to the same voice. So it keeps a memory of both the age of each voice since the event it contains began as well as which note at which velocity and so on was assigned to it.

I'm sure we could be talking about a few chord progressions that loop through a track, in which case yes I'd bet you could quite easily split the chords by hand if you could manage to "chop" the progression into repeating elements.

If on the other hand the sequence the OP refers to is more complex such as improvisation it gets way more hairy to do by hand.

 

Thank you all for your answers.

The question was more about playing chord and each notes goes to a different midi channel ( because instrument is, for example, a monophonic trumpet )
So exploding the track like you would for drum track (cubase,ableton,etc...) wont do.

i was looking for PizHere vst's but the site seems to be down for a while ( i just remembered , i might have them on an old HD )

And for Xhip, seems it would make it Thx. ( but i didnt read everything ... sorry no time now ... i'll come back in a day or two )

See ya soon. Thk ya all !