While working on another audiobook, i decided to make this sound. It's 38 voices, each saying different things, panned around and mixed together, creating a "wall of sound" that speaks like 38 radio channels at once. Recorded with a zoom h2 via usb into ardour2. Mixed and exported to flac with ardour2. Ps: it's all polish (with some possible german shout-outs), but the amount of noise makes it almost completely incomprehensible. Only a few words that are being yelled in a different voice can be understood. No sound repeats here, no recycling - every voice and every second of this recording is unique. Yes, it required quite a lot of work to record so much talking in quality! it's almost an entire audiobook squeezed into 5 minutes. Strangely (or not) listening to this makes my mind rest, because the noise blocks all other sounds from the environment - making my mind free of stimulation, allowing for sleep-like rest state. The signal is so much modulated that it appears to be not modulated at all - like static you get from a fm radio of you tune it wrong. The brain receives less data when you listen to this, than when you sit in a room hearing even faint (but distinct) noises from outside, other rooms, other people or yourself. This is sound masking in action. A very interesting psychoacoustic property of human hearing. Also: this is an interesting material to study of my voice's spectral energy distribution while speaking (as opposed to singing). As you can see using the spectrogram view, most energy is present in the band below 600 hz.
著者: Unfa
