According to the relevant literature, suicides caused by people with psychological problems tend to occur around 4am, mainly because the impulsivity of suicide is linked to the human subconscious. On the other hand, the human sleep produced by the sound of snoring, turning over sound, sleep talk, and brain waves formed by the brain can objectively feedback the person's subconscious situation. Therefore, the work is based on the data collection of the interviewees' voices during sleep and the data collection of brain waves during sleep, and the visual "algorithmic generation" technology is realized through Touch Designer (TD), trying to visualize the interviewees' "inaudible sounds" (snoring, gnashing teeth, talking in sleep) during sleep." Invisible data "(the amplitude trend of brain waves) is visualized. Then, the narrative expression of human subconscious is carried out to achieve the possibility of suicide tendency in special time period through the summary of subconscious visual graphics. On the other hand, the "generative art" obtained by the algorithm has a certain uniqueness, which is just like the function of identity card, representing the individual identity of different interviewees. The results also show that the recognition rate of interviewees to the visualization results representing themselves is as high as 80%.

Goal: Inspired by the above research results, this set of works tries to visualize the invisible subconscious data and visualize the amplitude sound during sleep, so that more users can perceive their subconscious, understand their own bodies, and monitor their physical and mental health.
Data: Sound capture based hardware system to obtain the sound of 20 people sleeping; The brainwave data were collected by TENOCOM BCI device.

Methods: The collected voice and brainwave data were cleaned and classified, and information graphics were generated through the visualization function of TD software. On the premise of ensuring the objectivity of data, the functional relationship between data and TD software was adjusted to map the voice and subconscious data of different respondents.