Announcements at train stations are often difficult to understand due to the noisy surroundings, for example when a freight train is thundering through the station. Researchers at the Fraunhofer Institute for Digital Media Technology IDMT have developed software that improves the intelligibility of voice communications in noisy environments. In addition to train stations, it could be used for conferences and for mobile phone conversations.
The software, dubbed “ADAPT DRC”, uses a microphone to constantly analyze the ambient noise and adapt the speech accordingly in real time. Simply making the speech louder is not enough. Instead, the software amplifies specific frequency ranges that are crucial for good understanding. This gives extra emphasis to consonants such as p, t and k. The software also boosts the level of relatively quiet speech components to compress the dynamic range. The software is application-mature and can be used in new and existing equipment.