Outdoor sound

Standard sources

The noise level inside the room will depend on the frequency spectrum of the sound source.

For instance, noise from nightclubs or bars will often contain a lot of low-frequency sound energy and lightweight building elements like windows will provide only a small amount of attenuation of that noise. It is therefore important that the frequency spectrum of the outdoor sound is either known or can be estimated with reasonable accuracy.

The Outdoor to Indoor Calculator includes a simple tool which can estimate the frequency spectrum for some types of sound source if the overall A-weighted noise level is known. For example, traffic noise can be estimated reasonably well in free-flowing conditions from a standard spectrum given in ISO 717. Simlarly, male speech spectrum can be estimated from ANSI S3.5.

Select Standard Sources to open to Source Spectrum Calculator tool. Choose the type of source from the drop-down box and set the overall A-weighted level.

The source levels are free field

The source sound level to be entered into the calculation sheet is the sound level incident on the building facade, but measured or assessed with the building not there.

It is the free field level.

If noise levels have been measured in the presence of building, then this measured level will include some reflection from the facade. In general, if the measurements are made 1 to 2 metres from a facade, the measured sound levels will be about 3 dBA higher than the free field level (i.e. with the building not there).

Angle of incidence

An often-overlooked factor with outdoor to indoor calculations is that all sound transmission loss data that is commonly provided is measured (or calculated) for a random incidence field on both the source and receiver side of the partition. Yet, as acoustic textbooks point out, the sound transmission loss of a partition will vary with the angle of incidence, reducing as the angle of incidence approaches grazing incidence.

It could be expected that the transmission loss of the facade and hence the internal noise level in the room should vary with the angle of incidence of the external noise. However, research undertaken by Rindel (1975) shows that for finite-sized partitions (i.e. not infinite sized partitions) and for low and mid frequencies there is little actual influence of angle of incidence on the "External" transmission loss.

To explain this in a different way, in classical acoustics the sound transmission loss (defined in terms of incident intensity) tends towards zero at grazing incidence. However, the incident sound power will also tend towards zero because the projected area that the sound wave is incident on tends to zero. Thus, it is not clear what wins, the incident level tending to zero or the transmission loss tending to zero. What will the internal level be?

Rindel avoided the confusion by defining the "External Transmission Loss" as the ratio of the energy densities of the sound fields on the source and receiver sides. The external energy density does not go to zero at 90° incidence and the External transmission Loss can be estimated from the traditionally measured random incidence transmission loss. These effects are included in the outdoor to indoor calculations carried out by INSUL. The same effects are included in EN 12354/3 in Section 4 when it describes the general principles of the calculation model. The apparent sound reduction index of the facade is given as equal to the random incidence sound transmission loss.

Thus trying to reach a simple conclusion, the answers calculated by INSUL are a good prediction for noise that is incident from a range of different angles of incidence (e.g. traffic noise) and furthermore is a reasonable approximation for most building facades for any angle of incidence. The answers for specific angles of incidence could be in error at frequencies above 1 kHz.

For either normal incidence sound or for grazing incidence sound, the maximum expected error at say 4 kHz would be ±3 dB, and for most noise sources these frequencies would not determine the overall loudness of the internal noise.