so we will control: contrast, saturation directly and rotation inversely, mapped to correct scales, via the midi signal input. This will be based on a single reward, no inhibit signal, asking for alpha activity in the occipital lobe. amp of alpha controls volume and pitch of midi signal. Alternately, we could create a 1 reward, 1 inhibit system in which the reward signal is blocked when another band is high than alpha. In this setup, alpha amp controls pitch, but if alpha is less than others, averaged together, block the midi signal.
in max, the volume will control volume, the pitch will control the pitched generation of new notes along a scale, with some tap-in rhythmic delay, so the sound fades out slightly.
One option for playback is load a single feedback movie that develops over time, adjust loop points as this fades into the feedback, so the simpler stuff is at the beginning, headier stuff at the end, as loops strung together, and then adjust the loop point presets so that it is drawing from whatever zone is called for based on the feedback signal. This would function like levels in a video game, so they wouldn’t be directly correlated with the alpha reward signal, but an average of the alpha reward performance over time, so if a minute of solid alpha performance was achieved without too much inhibit, then the system would morph to a new pattr set, with a more “alive” fb image, and more shimmering music. To achieve this perhaps, the midi signal is multiplexed as reward –> pitch, inhibit –> volume, and convert them to actual midi attributes in max.
Three states:
1. baseline
2. some alpha, some other
3. mostly alpha, stage one
4. mostly alpha stage two
5. realization
In the future, levels could include reward for alpha, theta synchronization.
Still working on the infinite zoom idea. I thought perhaps creating a video delay, adding a jit.rota object, and then xfading with the original would provide the effect of forever zooming in to the center of the image, but it did not. Instead, it created a delayed, zoomed in image, crossfaded with the original. suprise surprise. To actually produce this effect, perhaps I could take a quicktime movie, split it, create two rfilters for the loop points that were close to one another, and alternately cross fade each in with a zoom stage. Oh well, a project for another time.