There are three parts to Audio Analytic's patent pending Sound Classification and Analysis system.
- Sound Pack™ – Holds a concise robust summary of the time & frequency characteristics of a particular class of sound
- CoreLogger™ – Compares a buffer of sound to a particular Sound Pack™
- CoreTrainer™ – Automatically generates a new Sound Pack™ for subsequent classification or comparison.

Sound Pack™
To enable the acoustic signature of a sound to be detected, even in the presence of background noise, we have developed a system that reduces thousands of comparable sounds into their core constituent parts and then summarises them into what we call a Sound Pack. These Sound Packs often represent the summary of hundreds of hours of audio of a particular sound class yet are significantly smaller than the average text document on a computer. Simply add a new Sound Pack to your installation to gain new sound analysis and classification capabilities. You can learn more about how these are made here.
CoreLogger™
To enable you to classify and analyse sounds we set out to develop a system that was not only light weight, in terms of memory usage, but also extremely quick. Our software can run under the tightest of memory and processor requirements or can analyse hundreds of hours of audio in minutes easily adapting to multi-core systems, different architectures and operating systems. Using a Sound Pack the CoreLogger compares a new incoming sound buffer to the compact summary of the Sound Pack enabling the detection and analysis of sounds even in the presence of background noise.
CoreTrainer™
To summarise the complex differences between sounds into something as compact as a Sound Pack we have developed a completely new type of sound analysis tool based on our own cutting edge research. CoreTrainer scans through example audio accumulating the most salient properties as it goes. It then reduces and concentrates these properties into a Sound Pack which the CoreLogger can then use to classify new sounds.
Technology Overview
This two minute video features Audio Analytic's VP of Engineering Nick Brereton discussing the unique Sound Recognition Technology. He covers the ease with which CoreLogger can be integrated, the hardware requirements of the system as well as the simplicity of Sound Pack configuration.