No tags yet.
A key aim I have for this project is to design a system in which I am able to produce a large amount of musical content from Markov chains, browse through these to moderate, and eventually pick successfuly candidates. All generation and moderation will take place on my laptop, which I am afraid may not be powerful enough to play many different tracks worth of Serum presets. I am currently considering that I could generate and moderate the musical information through a 'demo device'. In stead of the laptop using Serum, it would use a less CPU intense synthesizer that will demonstrate the essence of the instrument and its melodic information, but without the bells and whistles. The audience would not hear the output of this laptop and therefore would be for my ears only, and once a successful candidate is sent to the performance PC, the generations fancier older brother will be presented to the audience in all its CPU-intensive glory. Perhaps melodic content can be previewed before it's dropped, but the preset can be totally randomly selected! I have also considered that this would leave an element of surprise for me to be subject to as the performer. I would have a rough idea of the melodic, timbrel and rhythmic content of the generation to come, and once the successful candidate was being performed with, I would be equally surprised with the outcome as the audience members would be. I think this would help maintain a level of anticipation and a rewarding performance flow, instead of knowing exactly what is going to happen and sucking the life out of the performance. After all, how could the audience enjoy what they are experiencing if the performer is bored and uninterested!? again creating anticipation and excitement for the performer.