An all too common situation: the speaker voice tells about change in departure time, but the important details are lost in the sound noise in the station premises.
Now, language researchers at the Fraunhofer Institute for Digital Media Technology IDMT developed a software program that improves the audibility of the loudspeaker announcements. A technology that can improve the sound in mobile phones when calls are implemented in noisy environments.
The technology behind software Adapt Drc, as it is called, consists of over a microphone constantly analyzing the ambient noise and adjust the speaker message to ambient sound in real time .
– It is enough then not to raise the volume on the speaker system to drown out ambient sound, explains Jan Renneis-Hochmuth.
Such solutions sometimes use the car radio manufacturers, but the result is usually that the speakers are pressed to max, which does not improve audibility.
The researchers instead analyzed the spoken language for frequencies, noting that the vowels often have low frequencies and pronounced a long time, while the consonants p, t and k is short sound images with high frequencies.
The consonants are the hardest to hear in noisy environments, but they are important for language comprehension. The algorithm used in the program enhances the sound that risk drowning in omgivningssorlet at any particular time.
The program also analyzes what is called the language dynamics, the strong and weak parts of the spoken language. Stronger parts may need to be tempered, weaker highlighted to be heard in a noisy environment.
The program is ready for commercialization. For mobile phones and conference facilities already have microphones, no additional equipment beyond the program. For railway stations and airports have microphones mounted in order to take up ambient sound for analysis in a hall.
No comments:
Post a Comment