Speech intelligibility in Swedish forests
Regular paper
Saint-Gobain Ecophon
Tuesday 2 june, 2015, 11:40 - 12:00
0.9 Athens (118)
Abstract:
For thousands of years we have developed our hearing in an outdoor
environment full of natural sounds, as babbling brooks, wind from trees,
bird songs and human voices. The problem is that students and teachers spend
a major part of their time indoors, in a sound environment with very few
natural sounds.
The effect is problem for students to understand what the teacher is saying
and voice problems for teachers. It´s important that teaching places provide
good speech intelligibility for listeners and good speak comfort for
speakers. Being able to listen without effort is important for good learning
and we know that bad room acoustics is a burden that impedes learning and
affect teacher voice health.
A good example is the Swedish forests where we can talk to each other over
long distances without having to raise our voice. I have made several
listening tests in the forest and also measured the sound reflections in
different forests.
The results are interesting and I mean that “forest acoustics” should be the
goal in terms of acoustic conditions in our schools.
My presentation will focus on why the forest acoustics provides so good
speech intelligibility, and which acoustic parameters what are interesting
to measure.
Many national sound standards put requirements on room acoustics in
classrooms. One requirement is reverberation time, according to ISO 3382-2,
and it´s often evaluated with T20. Unfortunately this is a very blunt
measure, because we start T20-evaluation first after the sound pressure
level dropt 5 dB. This “waiting time” is often quite long and it´s a problem
because we miss a lot of important information from the early part of the
decay curve. Therefor I mean we have to add C50 according to ISO 3382-1, to
control if the room acoustics is good enough for teaching.