A commonly quoted definition of noise is that it is simply unwanted sound. This seems to fit quite well since it is a judgement of the listener that defines noise. We can all identify with the statement that one person’s music is another person’s noise and for most people this open ended subjective definition is adequate. However, as a consultant in acoustics, I am often trying to better understand the reasons for people’s irritation at certain types of sounds and it strikes me that defining noise is not necessarily that simple.
It is well understood that control can have a lot to do with someone’s level of annoyance at a noise. In fact B. F Skinners theories about ‘operant conditioning’ would suggest that all human behaviour can be described by a desire to modify or control the surrounding environment. This is quite clearly demonstrated going back to the analogy of one person’s music becoming another person’s noise, because it could be the case that the music is enjoyable, but it is being played in the middle of the night whilst you are trying to sleep. If the ability to control a sound can affect how much you are annoyed by it, then this could clearly become a factor in determining whether or not it is noise. This poses another question. Since control is an interactive process and requires a level of awareness – can a sound be defined as noise if a person is not aware of it? It is not uncommon for people living with high levels of road traffic noise or airport noise to filter it out to the point where they say that they don’t really notice it, but does that mean it is no longer noise? The obvious answer is that yes, noise is noise whether or not you are aware of it; but I’m not sure if that is the case because a lot of the human auditory system operates at a subconscious level.
Product designers are aware of the fact that certain aspects of design contribute to a user’s enjoyment of the product, and that this is not always intuitive. A good example of this is engine noise within a car. For years car manufacturers were reducing noise levels to provide a more luxurious and relaxing experience, but it is now understood that there is a balance between luxury and enjoyment. Removing all engine noise from a high end sports car results in a car that is not so enjoyable to drive, which would suggest that there is a level at which engine sounds become engine noise. The exact level at which this sound becomes a noise will obviously depend on a number of factors including the type of car, the character of the sound and the individual driving the car, but the key point is that a subconscious judgement relating to the level of sound is taking place.
Another point relating to product design is aesthetics and this can relate to sounds which are produced that might otherwise be described as noise. Going back to car design, the sound of a car door shutting is something which has been engineered to be aesthetically pleasing. If there were no cushioning in the seals and shutting a car door resulted in a sound similar to a lump hammer striking an iron girder. This would not be pleasing. Conversely, if a car door were completely silent, we may be left wondering if it has actually shut properly and it is likely that a level of enjoyment would be lost. It is quite possible that a certain amount of emotional attachment is related to the sound of shutting a car door whereby we feel safe and secure as the sounds of the busy world outside are shut out but I think it is also clear that the character of the sound plays a part in the aesthetics.
Low frequency sound is sometimes described as ‘warm’ and high frequency sound is often described as ‘harsh’ which would suggest that pleasing sounds may tend to have more low frequencies than high frequencies. However, birdsong and wind chimes can contain quite a lot of high frequencies and would generally be thought of as peaceful relaxing sounds. The character of a sound which defines how pleasing it is cannot easily be described and it is also inextricably linked to how we feel about the source that generates it.
There are also everyday noises that are subconsciously processed to provide feedback which we find useful or reassuring. The car door (or any other door for that matter) can let us know that people have arrived or are leaving. The hydraulics of the rubbish truck operating down the street can let you know that your rubbish has been collected or that you need to rush down to put the bins out. Creaks and groans in your pipework can let you know that the heating has come on and it will soon be warm enough to venture out from under the covers to have breakfast. These sounds can all be processed subconsciously but if the level were to become too high or the character more annoying then we may become more aware of them and they may start to become noise rather than sound.
To summarise, it would seem that there are four main aspects that contribute to the ‘N’ factor defining a specific sound as a noise:
It would seem that if some of the subjectivity in these aspects can be isolated then a more precise definition of noise could be achieved. The level of a sound is something which can easily be measured but the listener’s perceived level relates to background noise and context within the noise environment. Although there is an element of human perception to this, it would seem that sound level is a reasonably objective measurement for defining noise. Character, on the other hand, is something which can be very subjective, particularly when a sound is identifiable but frequency, tonality and impulsivity can all be objectively measured. If samples could be generated which exhibit varying levels of tonality and impulsiveness at different frequencies without producing any kind of identifiable sound, then it may be possible to examine a relationship between these characteristics and a definition of noise. A listener’s awareness of a sound clearly has subjective elements relating to their ability to filter out certain types of sound. However, for unidentifiable sounds, it should be possible to examine a link between, level, character and awareness. Although the ability to control a sound would seem to be entirely subjective, the extent to which it can be controlled may affect the point at which it is defined as noise.
Since all four aspects contributing to the ‘N’ factor can be interlinked and have varying levels of subjectivity, it is not surprising that there is no clear way to define noise. In principle, it seems that with enough resources, a research project could be undertaken to examine the objective elements of these features, which could conceivably help to narrow the definition of noise. It may be that someone in a university somewhere is already working on this and as I am lacking the resources to undertake such a project I can only hope that someday I shall be beholden to a professor with some interesting findings in this field of research.
by Seth Roberts
30 Nov 2016Back to news