The sound clip below of metal chimes softly swaying in a gentle breeze is both real and artificial at the same time, walking the tightrope between nature and creation. Our Modeled Sound Synthesis engine (MSS) can dynamically ingest a few seconds to minutes of high quality audio, decode and identify all of the distinct features within the audio segment, and use this information to train its own internal sound models.
These sound models can then be leveraged by MSS to generate new audio content by seamlessly combining and rearranging sound patterns and sequences to produce a unique sound experience every time it is played. In fact, a mere 6 minutes of source audio can be used by MSS to synthesize over 400 years of unique audio — slightly longer than most rainstorms — resulting in the most natural sound experience available (short of standing in the middle of a downpour!)
Why Does It Matter?
Our brains are adept at finding patterns and we’re naturally drawn to auditory irregularities and unexpected changes. Perceivable sound issues in games, sound generators, and even office noise suppressing systems are easily detectable by the human brain. Loops or consistently repeating artifacts disrupt the smooth flow of sound, forcing our attention to constantly shift and detect these anomalies. This continuous effort can be mentally taxing and lead to fatigue, irritation, and ultimately, stress. This loop awareness can activate stress centers in the brain, lower cognitive function, and even sour moods.
The usual fix? Increasing the amount of unique audio but that’s too bulky for today’s downloadable software model. The developer is then left with a difficult choice of having more artificial environmental sounds or using lower sampling rate, introducing new sets of quality problems and audio artifacts. With the MSS engine, neither sacrifice is required. Simply ingest a very small sample of high quality audio into the engine to enable it to generate countless hours of unique atmospheric audio, so authentic sounding that the listener cannot help accept it as anything but natural and melting perfectly into the background.
Hearing is Believing
Want to hear a longer demo for yourself and hear how generative audio can work for you? Visit the Apple App store and try one of our Sonici Soundscape apps, powered by MSS, or use our contact form to arrange for a formal demo. Your ears will thank you!
References
“The Effects of Background Noise on Cognitive Performance: A Review” (2010)
“The Relationship Between Unwanted Sound and Emotional Distress: A Review” (2011)
“Auditory Distraction: A Review of Psychological and Neurological Mechanisms” (2015)
“Environmental Noise and Psychological Health: A Review of Recent Research” (2017)
“Soundscapes and Their Effects on Human Psychology” (2017)
“The Neurophysiology of Unpleasant and Pleasant Melodies” (2019)