Acoustic perception of an approaching ambulance siren

In Germany many ambulances on duty are involved in accidents. For 1998 the Federal Office of Road Traffic published a statistics stating that in 322 accidents with ambulances 9 people were killed and 541 injured. More than 40 % of the accidents occured at major intersections and more than 20 % at minor junctions. Working Group Noise Measuring Technology investigated the question how the siren of an approaching ambulance close to an intersection is heard by a driver inside another car. The results will be included in the training of ambulance drivers which is currently developed by two German organizations, the "Deutsche Verkehrssicherheitsrat" and "Johanniter-Unfall-Hilfe e.V."

Practical road traffic experience shows that it is very difficult to define the direction from which the ambulance is approaching when the siren is heard for the first time. This is, among other things, due to the fact that localization is considerably disturbed by sound reflections on houses and cars. The perception of the direction is the result of the evaluation of sound signals on both ears. Both the sound level differences and the running time differences between left and right ear are evaluated by the brain. To demonstrate this process, so-called dummy head recordings were taken. Two microphones applied to the ear channel entrances of a dummy head record the sound. Contrary to conventional recordings, the acoustic behaviour of the auricles and the influence of head and shoulders on the sound were also taken into account. When these recordings were listened to with a suitable headset, a three-dimensional room impression results which allows the listener to evaluate the origin, direction and distance of sound events.

Fig. 1: Dummy head  to measure and evaluate sounds inside a car

In our investigations, an ambulance approached an intersection with its siren switched on. A car was parked close to the junction at a distance of 15m to a possible crash scene. The ambulance used the following siren: city siren, country siren, pneumatic siren with a measurement distance of 45m, 30m, 15m and 0m to the possible crash scene. The ambulance was visible to the car driver only at a distance of less than 15 m.

The acoustic dummy head was installed in the car (small to medium size car). Additionally, the sound pressure level inside the car was measured with a sound level meter.

The dummy head recordings impressively confirmed that reliable localization of the ambulance by the car driver is possible only within the visual range.

Audibility measurements were performed as follows: First sound pressure levels of all different sirens were recorded seperately with the engine switched off. After that, the interior noise was recorded at speed and with the radio turned on. The levels of the single sound sources were tuned seperately with the aid of suitable signal processing programs and adapted for hearing comparisons.

By this method it can, for example, be investigated at which radio level the ambulance siren can not be heard any more. Likewise, various conditions can be simulated ( standstill with engine at rest, slow pace at 50 km/h, with and without radio turned on, etc. ) and the audibility or non audibility of the siren can be demonstrated. It can even be shown how the fan noise of the heating system, which often drowns out the engine noise, influences the acoustic situation inside the car. As all sound samples were dummy head recordings, an accurate spatial impression of the situation inside the car is obtained when mixed signals are listened to via headset.

The following sound pressure levels were measured inside the test car, the expanded uncertainty (2u_c) amounted to 2 dB:

The sound pressure level in dB (A) takes the frequency-dependent sensitivity of the human ear into account, whereas the sound pressure level in dB (B) rates all frequencies identically. Inside the car, the measured sound pressure level is strongly influenced by low frequency noise. This is why the dB (B) value exceeds the dB (A) value when the engine is switched on.

By comparing the measured values it was found out that inside the car there is often hardly a warning effect or no effect at all. In addition it has to be considered that the sound pressure levels very much depend on the car itself. Especially in new and expensive cars great importance is attached to external sound insulation which further reduces the audibility of external warning signals which further reduces the audibility of external warning signals. The car used here had no extra sound insulation.

The following sound samples demonstrate the hearing impression of a passing ambulance inside a car. First, the car had a speed of 50 km/h ( local traffic )

(Sample 1, approx. 2.7 MB).

With then radio switched on and a sound pressure level of more than 76 dB (A) inside the car the siren of the passing ambulance is hardly noticeable.

(Sample 2, approx. 2.9 MB ).

The perception of the siren at a speed of 100 km/h is also hardly noticeable ( even with the radio switched off )

(Sample 3, approx. 1.6 MB).

Summarizing it can be said that of the three ambulance sirens neither the city siren nor the country siren had a level advantage.

The pneumatic siren consists of two alternating sounds. Because of its rich sound spectrum it appears not only subjectively louder, but also shows a mean level advantage of up to 4 dB on both ears.

Contact person:

Ingolf Bork, Dirk Ratschko, FB 1.7, AG 1.72, E-Mail: dirk.ratschko@ptb.de