About the 4.0 version

[noise2sine]

Hello,

Will it be possible in the 4.0 version to record in the piano roll Midi datas, and being able to export them as Midi files ?


Thank u

[martignasse]

+1 for this question

[senso]

I hope I'll have time to do that soon.
If it's not in the 4.0 I'll implement in the first following update something like the 4.01?

[martignasse]

ok, ok

it could be for the 4.01 if we have the 4.0 sooner

hehe

[noise2sine]

:)

Perfect!!

[amiga909]

would love this too
[about the question: live midi input per keyboard works]

made a test regarding with a idea for the piano roll



1. would it possible to have the array common for all midi and data constructs? I am aware the piano roll would need far more information and sysex is quite another story again.. but maybe its possible to have better translations than we have now?

edit: dark green cases: did rescale the values and it works now (dark green -> light green)

2. btw. I believe the piano roll GUI is buggy: for example 'clear' needs to close and reopen the window..

[senso]

1. would it possible to have the array common for all midi and data constructs? I am aware the piano roll would need far more information and sysex is quite another story again.. but maybe its possible to have better translations than we have now?

Array's are totally compatible each other but you have to respect scales:
the StepMidi sends data's from 0 to 127. So if you connect a stepMidi to a SeqLine you have to change the Max parameter of the SeqLine to 127 or divide StepMidi data's by 127.
The only exception is the mapper curve which can receive only data's from 0 to 1, (for that you have to set the type of curves to 'user draw')

Also for the pianoroll input array, things are more complexes because we need, at least 4 array inlets
1) midi message
2) midi data1
3) midi data2
3) time position of the midi msg

actually it's an hard job to implement, perhaps in futures version, but not very soon!

[amiga909]

Also for the pianoroll input array, things are more complexes because we need, at least 4 array inlets
1) midi message
2) midi data1
3) midi data2
3) time position of the midi msg
actually it's an hard job to implement, perhaps in futures version, but not very soon!

thanks a lot for your reply senso. I understand much better now, and taking rescaling in account there is no problem at all for all non-pianoRoll modules.

for the piano roll: I see there are much more array inlets needed, if you want to access the CC curves, even more. I'll be patient to wait...
just a thought:
- only allow note events (no array inlet for event type)
- G. Mazzola seperates 4 dimensions of a sound: startTime/gate/noteOn (time pos1), length/noteOff (time pos2), pitch (midi data1), velocity (midi data2).
- imagine you would control these 4 array inlets via different data and midi generators or feed back e.g. piano roll startTimes to a seqSwitch module

other opinions about 'multidimensional' composing?

[senso]

in next version you will be able to create various algorithmic patterns in the piano roll for example:

connect a random generator to the ppQPos
connect a create Midi message to the midi input.
activate the Rec.

randoms notes will be recorded at random positions.
fun!

[gregh]

Hi, what random distributions(s) do you have available in Usine? I ask this because my favourite is the beta distribution as you can control it to give a very wide range of shapes to draw values from.
greg

[senso]

Hi, what random distributions(s) do you have available in Usine? I ask this because my favourite is the beta distribution as you can control it to give a very wide range of shapes to draw values from.
greg

actually two distributions:
- uniform (data/random module)
- user drawn distribution (data/random law distribution); with this one you can create your own.

but any distribution can be easily implemented as scripts.
feel free to post your favorites equations!

[gregh]

"Computer Music" by Dodge and Jerse has a good description and pseudo-code on the beta distribution on pages 351 and 352

Here is the c-code for all the distributions mentioned in Dodge and Jerse

http://sig.sapp.org/src/sig/Distribution.cpp

from Craig Sapp - there is alot of useful code there and elsewhere on his site

Greg