Photo: DTU

New tool improves acoustics and the working environment

Tuesday 21 Nov 17


Gerd Marbjerg
DTU Electrical Engineering
+45 45 25 39 33
An acoustic simulation tool is making it easier to calculate and work with acoustics in small rooms such as classrooms and offices.

When teachers take sick leave due to voice problems or students find it difficult to concentrate and hear what the teacher is saying, the problem could well lie with the classroom’s acoustics. Poor acoustics can also lead to stress and more frequent mistakes for staff in large open plan offices and hospitals. It is therefore important to take acoustics into account when designing and working with workplaces of this nature. This task has now been made easier with a new calculation tool developed at DTU Electrical Engineering.

Many new concert halls have been built around the world in recent years, and people naturally talk a lot about acoustics in relation to these. But acoustics are definitely not limited to musical experiences. Acoustics are also an important—often critical—part of our daily lives, in offices, conference rooms, schools, hospitals, etc.

So what are good acoustics? Associate Professor Jonas Brunskog, who researches sound and acoustic technology at DTU Electrical Engineering, says:

"In a company, you mostly work with short time horizons, so we were pleased to be able to involve DTU in this more long-term research project."
Erling Nilsson, an acoustics specialist in Ecophon’s R&D department

“We cannot define it precisely, as a lot depends on the situation and the people in the room. It was previously assumed that the more a classroom was acoustically dampened, the better. But our studies have shown that the reverberation time (the time it takes sound to disappear) should be around 0.6 seconds. If you dampen the sound more than this, the teacher’s voice gets too little support from the room to be able to carry. In most situations, it will be necessary to consider other acoustic parameters in addition to reverberation time in order to work towards a specification of good acoustics.”

Excellent tools are available for calculating the acoustic environment in large rooms such as concert halls. But Ecophon, a Swedish company that works with sound-absorbing ceilings for ordinary workrooms, felt that a tool to calculate and simulate acoustics in small rooms was sorely lacking. It is therefore participating in an industrial PhD project in collaboration with DTU.

“In a company, you mostly work with short time horizons, so we were pleased to be able to involve DTU in this more long-term research project,” says Erling Nilsson, an acoustics specialist in Ecophon’s R&D department.

Gerd Marbjerg was employed jointly by DTU Electrical Engineering and Ecophon for the three-year project. During this time she has developed a model that can handle all the parameters that influence how people perceive the acoustics in a room: reverberation, speech intelligibility and noise level.

From numbers to experience

Sound exists as waves emitted from a source and reflected by whatever they encounter in the room. The purpose of an acoustic room model is to express this using numbers and describe what happens to the sound under various conditions.

The more parameters the model has to include, the longer the calculations will take. To avoid the model becoming too calculation-intensive, existing commercial models have built-in assumptions about how sound will behave. This is acceptable when planning large rooms, but if you want to optimize the working environment in small rooms, it is necessary to describe in more detail what happens when the sound encounters the room’s surfaces and furnishings.

Absorbent surfaces can change the phase of the reflected sound compared to the sound that hit the surface. Similarly, the way the sound is reflected depends on the angle at which it hits the surface. Gerd Marbjerg’s model takes all these factors into account. Using her acoustic simulation tool, it is therefore possible to calculate exactly what effect various adjustments (such as adding Ecophon’s absorbent materials) will have. The tool calculates an ‘impulse response’, i.e. how the room reacts to a brief sound containing all frequencies, like the sound of a clap or gunshot. The impulse response is the room’s fingerprint, and can be used to determine the reverberation time etc.


But the numbers alone are not enough to describe how people actually perceive the sound. It is only when you combine the impulse response with recordings of speech in an anechoic chamber that you get the full picture of how a voice will be perceived in the room. This is called auralization, the audio equivalent of visualization, and you can listen to it using headphones.

“But the headphones do not take into account variations in the shape of people’s ears, head and shoulders. We are therefore now going one step further and recreating the modelled room in DTU’s audio-visual immersion lab, which has 64 speakers arranged in a spherical pattern. In the lab we can test how changes in the room’s acoustics affect the experience of sound, even for people with a hearing aid,” says Gerd Marbjerg, who is still at DTU on a postdoc project financed by the Oticon Foundation and Ecophon, to work with this type auralization.

Ecophon also aims to fully exploit Gerd Marbjerg’s model, and therefore plans to create a room similar to DTU’s audio lab, just with fewer speakers.

“You can use the model without having such a room, but it becomes much more tangible when we can also show the effect of adding an absorbent element or otherwise adjusting the acoustics. It’s amazingly effective,” says Erling Nilsson.


The walls, ceiling, and floors in DTU’s sound lab are covered with long wedges made of porous, sound-absorbing material. These help create a completely silent, anechoic room.
14 DECEMBER 2017