k y m a • t w e a k y — the kyma collective || view the current website || February 2015 Archive

/ WebHome / Tools / Share.LiveMarkovKBDDiscussion

Search


Share Section


Sounds
Samples
Timelines
Tools
Microsounds

Home 
Topics 
More... 

All Sections


Products
Order
Company
Community
Share
Learn

Login / Register 
Change password
Forgot password?
%SESSION_IF_AUTHENTICATED% Site Map
%SESSION_ENDIF%

Symbolic Sound


Home
Kyma Forum
Eighth Nerve

TWiki Links


%SESSION_IF_AUTHENTICATED% TWiki Shorthand
TWiki Formatting FAQ
What is TWiki?
%SESSION_ENDIF% twiki.org
This Tool generates new, monophonic sequences based on the likelihood of each transition (or chain of transitions) in a monophonic sequence from the MIDI input, Continuum, Wacom Pen (eraser end), or previously recorded text file.

On the first page, decide how long the TrainingSequence should be (MemoryLength), how long each algorithmically-generated sequence should be (SequenceLength), and how long a pause there should be between algorithmically generated sequences (BeatsBetweenPhrases). BeatsBetweenModels controls how often a new model should be created (since the TrainingSequence is continuously changing).

Set the tempo and the number of subdivisions and, if you'd like to hear the metronome, press Metronome. You can also practice for a while before you start the training.

When ready to start the training, click the Train button. Or, if you prefer to initialize the TrainingSequence from a previously saved text file, click Read instead. (To start with, try reading the file called pentatonic.txt)

On the training page, you can begin playing a monophonic sequence. Once you have at least SequenceLength events in the TrainingSequence, the Tool will generate and play a sequence of SequenceLength (or fewer) note events, separated by pauses of BeatsBetweenPhrases. Continue playing along with the generated sequence (and/or during the pauses). When TrainingSequenceLength matches MemoryLength, new input will be added and the oldest inputs will be "forgotten". Click Read to read a previously recorded sequence into the TrainingSequence. Click Save to save the last sequence played.

Every BeatsBetweenModels, the Tool creates a new MarkovModel based on the new state of the TrainingSequence. If you change the chainLength, DivisionsOfBeat, or pitchGrid, a new MarkovModel is created right away.

The Tool uses the likelihoods of transitions of length chainLength in your training sequence as the likelihoods for generating new, random sequences.

-- CarlaScaletti - 08 Mar 2007

Thankyou Carla, very very cool sequencing tool - you know I love sequencers, this tool is very inspiring ... Any chance we could get this within a sound, as an alternative/add-on to the AnalogSequencer? sound?

-- CristianVogel - 08 Apr 2007

 
 
© 2003-2014 by the contributing authors. / You are TWikiGuest