218
D. Cotoros et al.
Several explanations were provided starting with physiological reflex or willing
compensation, analysis of the starting point, and probably most important the
influence on speech recognition [4, 5].
There are controversies regarding the adaptation of the people to communication
in noisy environments, some researchers consider there is no active adaptation, and
people just speak louder but others tried to find proofs that there are significant
involuntary changes not only in the sound intensity but also in the fundamental
frequencies, higher emphasis on vowels, etc [6].
Some interesting approaches aim at creating some systems able to manipulate
natural speech into Lombard’s speech in order to enhance intelligibility [7] or
studying the effect of outside noises upon the communicative phenomenon [8], even
improving speech recognition with forensic purposes.
2 Materials and Methods
The experiment was performed due to the involvement of several persons, aged
between 25 and 60 years old, three males and seven females, with no reported issues
regarding hearing loss or speech problems. They are all native Romanian speakers,
and eight of them have no clue regarding Lombard’s effect. Only two of the subjects
are vaguely aware of the phenomenon, but the purpose of the experiment was revealed
only after the recording.
The subjects were asked to read a short text (around 300 words) from a
book review, while wearing headphones. They were subjected to four possible
disturbances:
• Quiet headphones—the headphones are used only to camouflage any environmental sound and create a silent area.
• White noise—It is created by a continuum of frequencies equally distributed
over the whole hearing range. In healthcare applications, white noise is used to
treat increased sensitivity to normal environmental sounds, or to camouflage the
annoyance caused by tinnitus.
• Construction site noise.
• Pop music.
A Shure SM7B microphone was used to record them while reading. This type
of microphone was selected as it is a microphone that focuses on capturing a clean
recording of a single sound source, while filtering out everything else, making this
polar pattern perfect for clearly recording a subject’s voice, while being surrounded
by progressively more excessive background noise [9] (see Fig. 1).
The experimental setup also includes a pair of Audio-Technica ATH-M50 X
headphones with a closed back, keeping the sounds playing back to the subject
from bleeding into the microphone, allowing for a clear recording. Everything was
recorded into a program called REAPER, a digital audio workstation (DAW)-type
D. Cotoros et al.
Several explanations were provided starting with physiological reflex or willing
compensation, analysis of the starting point, and probably most important the
influence on speech recognition [4, 5].
There are controversies regarding the adaptation of the people to communication
in noisy environments, some researchers consider there is no active adaptation, and
people just speak louder but others tried to find proofs that there are significant
involuntary changes not only in the sound intensity but also in the fundamental
frequencies, higher emphasis on vowels, etc [6].
Some interesting approaches aim at creating some systems able to manipulate
natural speech into Lombard’s speech in order to enhance intelligibility [7] or
studying the effect of outside noises upon the communicative phenomenon [8], even
improving speech recognition with forensic purposes.
2 Materials and Methods
The experiment was performed due to the involvement of several persons, aged
between 25 and 60 years old, three males and seven females, with no reported issues
regarding hearing loss or speech problems. They are all native Romanian speakers,
and eight of them have no clue regarding Lombard’s effect. Only two of the subjects
are vaguely aware of the phenomenon, but the purpose of the experiment was revealed
only after the recording.
The subjects were asked to read a short text (around 300 words) from a
book review, while wearing headphones. They were subjected to four possible
disturbances:
• Quiet headphones—the headphones are used only to camouflage any environmental sound and create a silent area.
• White noise—It is created by a continuum of frequencies equally distributed
over the whole hearing range. In healthcare applications, white noise is used to
treat increased sensitivity to normal environmental sounds, or to camouflage the
annoyance caused by tinnitus.
• Construction site noise.
• Pop music.
A Shure SM7B microphone was used to record them while reading. This type
of microphone was selected as it is a microphone that focuses on capturing a clean
recording of a single sound source, while filtering out everything else, making this
polar pattern perfect for clearly recording a subject’s voice, while being surrounded
by progressively more excessive background noise [9] (see Fig. 1).
The experimental setup also includes a pair of Audio-Technica ATH-M50 X
headphones with a closed back, keeping the sounds playing back to the subject
from bleeding into the microphone, allowing for a clear recording. Everything was
recorded into a program called REAPER, a digital audio workstation (DAW)-type