Play layered multiple voices from one midi channel

Hello Everyone !

I was wandering if it was possible to play two ( or more ) voices from the same keyboard from a unique midi channel ( let's say channel 1) .
I'm not talking about splitting, but more about layering.

It tried to add  "  MULTI_TIMBRALS = All   " to  my configuration.txt


and
Set,Channel,Prog,Voice,
0,0,0,1,#
0,1,0,2,#


to MTchannelmap.csv in the right folder ( 0 mysampleset )

But it didn't work ...

I don't really know if this is possible, but I'be great to do that in order to create more complex sounds.


Cheers !
P

I used a trick in my previous setup made with python and mido library.

I copied every incoming note message from channel one to channel 2 ( or 3,4...) and sent them all.


Something like this

for msg in keyboard:
     if msg.channel == 0:
         for channel in [1,2,3]:
             output.send(msg.copy(channel=channel))

Hello Pete,
It's in my future wishlist to enable multiple voices on one channel (e.g. for facilitating organ stops).
But: that's future.

If you are able to provide multiple channel input to samplerbox, the multitimbral option will give you some room for getting nearer to your goal, however typical samplerbox features won't be available which is a serious drawback.
Only mixed sound effects are awarded and these are controlled from the master device(s) which cannot be the same as multitimbral devices.

Next release will introduce a MULTI-WITHMASTER mode giving a device combining a master channel with 15 MULTI-TIMBRAL mode channels giving some more options.
But I'm afraid you need the "that's future" thing.

What can I say.. work in progress, so much nice things to include 
Hans

I'm not talking about splitting, but more about layering.

Hi Pete,
While thinking a little longer about the goal behind your question (as well as the "copy channel" post):

Do to want to mimic the hardware method of changing nature of the sound with the velocity - as instruments have have different sounds depending on the force of playing ?
So if you have 2 or more recorded sounds, they should be mixed differently depending the velocity = force of play.

In that case this info can give an alternative approach, given you have at least recordings of 4 levels.
This is the basis of the grandpiano demo set - I tried with 8, 4 and 2 levels and (subjectively) concluded 4 levels to be enough for this, and 8 should be the ultimate (unless you are a magician on keyboards, but then you can afford more professional stuff than DIY on PI ).

For most instruments, if you have only two they should be mixed as the switch between two will be noticed even by less experienced users.
The good news: I can testify that such mixing gives real nice results on Roland equipment, in other words "it works". They call that layering.
Despite using this technique actively myself I never thought about implementing it on samplerbox...
But it might be that results are even better than the implemented technique and since only 2 layers needed it saves memory and load time.

But I have a hunch doing this software wise is far more complex than doing it hardware wise. So before digging into this: did I catch the background of your question?
Regards, Hans

In fact what I want to achieve within samplerbox is what I do with my actual setup.

It's very simple actually. I just want that what I am playing live on my keyboard is sent to several channels at the same time. So It can be played by a piano and an organ for example. This would be the simplest use.

Once I have the ability to do that I would also be able to create Kontakt like "multi nstruments"

Like I could build a multi voice with the close sound of piano on voice1 and on voice 2 the room sound of the same piano.

Let's say I want to add different release samples too. I could do the same thing.

Ok, so that was covered fully with my initial understanding of 17-12.
Alas I already determined that is very complex and still a future fantasy.
However your goal could also be achieved by requeing the notes on (a) different channel(s). Also not really easy, but perhaps easier to realize ... with sidenote it may cause small latency of the duplicated notes.
I assume you read the docs, but be sure: other channels than the master channel will only have the "All sounds" controls/features as detailed in More info. This may be a showstopper for you.
Thanks for clarifying and giving me food for thought.
Hans

Hey Hans thanks for replying.

As I said before I already have this kind of script in my old raspberry + sampler setup and it works really well.
Of course when you duplicate the notes on two many channels it can be messy but used with care it's ok.

In fact I am so noob that I don't know how I can run my script on startup just before yours. I could make believe to samplerbox that I am a sequencer sending many notes

hhhmmmm... Maybe the "concept" to achieve what I describe ( kontakt multi instruments ) would be possible by launching multiple instances of samplebox at the time ...
Maybe rpi 4 would be able to handle it ?

launching multiple instances of samplebox at the time ...

Hi Pete,
You can try, but I expect complexity in the area of competition for crucial hardware (both the callbacks of midi and audio won't like it).

I think of investigating note copying within samplerbox itself (third variant of techniques used for autochord and chorus).
Alas this will give a voice switching complexity within that single call back.
But it would also be a first step in development of church organ stops mentioned above, so this discussion gets the future fantasy closer to tangible things to investigate=>create.

Compared to preprocessing methods, timing will be perfect because multiple midi messages (serial by nature) are not needed. However at cost of CPU strain in the midi callback.
How many simultaneous instruments are at stake here: 2-4 or complete orchestra's ?

Hans

How many simultaneous instruments are at stake here: 2-4 or complete orchestra's ?

Just 2-3 would already be cool !