The technology based on decreasing mask effect by frequency-dependent amplification of a useful signal. Simultaneously the algorithm takes into consideration individual hearing thresholds, frequency characteristics of a certain device, and incoming sound level.
The spectral estimation of the the power spectral density of the ambient noise produced by the external microphone. To customize using the built audiometry implemented.
This technology provides contact centers and teleconference systems a useful function for users, which talk in a very noisy environment, hear a collocutor clearly.
The effect is reached by a real-time analysis of surrounding noises, and subsequent increasing intelligibility of a collocutor speech in accordance with the type of noise. The technology protects user's hearing because it provides only selective amplification of only those signal components which are masked by noise, allowing a user to achieve high intelligibility with a relatively small increase in the overall volume.
As a side effect the technology improves both a user, and his/her collocutor comfort while a talk or teleconference.
Implementing the "hearing through ambient noises" technology can be effective for both phone and VoIP systems. It can be implemented both as an App (Conference Denoiser, UrbanDenoiser Player), Windows/Linux driver (Petralex Speech Communication), or server-based module.
The basic technique of Denoiser is automatic adjustment to the type of surrounding noise and balanced amplification only of those signal elements which are muffled by ambient noises. The patented technology which yields in efficiency to no active noise reduction devices allows communicating in aggressive noise environment by either using standard headset or not. The applied technique allows keeping high quality of audio signal of 44,1 kHz (16 bits) and provides the best performance in constant noise environment (public transport, streets, working areas, near noisy household appliance, etc.)