Brain controlled synthesizer

[pjbridger]

Hi guys,

i am at university in my last year of my degree,

i am doing a project called experimental sound and i decided to persue an idea id had previously

i saw a new controller that had been brought out by OCZ called the NIA

it reads signals from your brain and transforms them into keystroke commands to control computer games

i thought it would be good to use this device to control a synthesiser and do tests with subjects to see what kind of noises their brains would make

i wanted to use synthmaker because i have a small amount of experience with it and found alot more complications when trying to use Max MSP

i have now come up against a problem

the NIA outputs keystrokes from the qwerty keyboard which i was using with a stand alone synth that i had made

in addition to this i wanted to add drum samples from a sample looper

but i have found that these clash when using both an oscillator and a sampler at the same time

i also wanted to activate some controls such as filters etc using MIDI CC commands

i found a guy who had made a converter that changes the NIA messages into CC commands

the problem with this is that it is limted to 6 commands and overides the keystrokes

i have recently been trying to find a program that will enable me to convert the keystrokes into MIDI commands

i have tried the bomes software, MIDI ox and others but cant seem to get any to work

i am at the moment slightly panicing about this project

can anyone see a work around i could use

please help

love the forum

PJ

[DigitalWhiteByte]

Look here
http://www.musatcha.com/software/MindMasterMIDI/index.php

[MichaelBenjamin]

depending on what your thing gives as output you can also try translating to midi/mouse/whatever with glovepie: http://sites.google.com/site/carlkenner/glovepie

[pjbridger]

thanks guys

the mind master MIDI is what ive already been having problems with

i have been trying to use glove pie but cannot seem to get it to work at all!

does anyone know any tutorials for it?

[MichaelBenjamin]

i have used it with a wiimote, downloaded a sample script and tweaked that. worked good with surprisingly small latency and was not too complicated with an example. you can also setup the things with the wizard mode, but it depends on what your controller can output and if you can intercept it in glovepie and remap it. i am no expert with this however, so i can only hint vaguely.

btw can you post a link to that nia device? edit found it http://www.ocztechnology.com/products/o ... herals/nia

[MichaelBenjamin]

i just flew through the manual, and it seemed in all examples they use muscle contraction and map that to either joystick or keycommands in a quite formulaic way. i cant imagine you can play a 3d game with mouse/key precision, but maybe you also can output the seven unaltered realtime streams (muscle, alpha1,2,3 and beta1,2,3 (are these the brainwaves?)) to several joystick axis and intercept that in glovepie or the other midi software at dwbs link to make midi data out of it.

[pjbridger]

thanks for the reply

ye the problem is i cant get the midi commands to workas keystrokes to actually play the synth

i found a module on here hich will accept them called midi learn but i cant seem to hook it up correctly

[MichaelBenjamin]

what kind of data do you get inside sm? already midi? or do you use the key pressed primitive?

[pjbridger]

i can get actual midi messages into sm but i cannot convert these CC messages into keystroke commands

[MegaHurtz]

you could see if they are over a certain threshold, 0 for instance. via > comparrison.
Or if you interprit it as midi note numbers you could have a full keyboard.

reminds me of some eeg experimenting i did, drove the budget up waaay high.
Not feasable..

Edit: ow if you run it as note numbers be shure to have a key pressed for the gate.

[MegaHurtz]

Look here
http://www.musatcha.com/software/MindMasterMIDI/index.php

I think itd be slightly better placed, on the motor cortex.
Still its a fun idea, but seemst to lacking actual controller performance.