Cocktail Party Effect and Room Acoustics

Why are bars and restaurants sometimes so unbearably loud? So loud that you have to scream at the person across the table to hold a conversation?

The reason is something called “the cocktail party effect.” Imagine you are in a space with a number of other people. You talk. This raises the sound level, so the next person has to talk louder to be understood. Someone else now tries to be understood, plus your voice, plus the other person’s voice. This can escalate until everyone is screaming to be heard.

Some spaces are more prone to the cocktail party effect than others. Some restaurants spread their tables out, are carpeted, and also may have sound-absorbing wall and ceiling treatments. These restaurants are less prone to the cocktail party effect. Alternatively, other restaurants pack their tables close to one another, are not carpeted, and often have few, if any, sound-absorbing wall or ceiling treatments. These spaces tend to be more prone to the cocktail party effect.

The cocktail party effect is dynamic in the sense that a lively space may not be difficult to communicate in as long as there are few people in the space talking at the same time. Furthermore, if everyone were disciplined and spoke quietly, multiple conversations could still go on without compromise to intelligibility. As an example, in a noisy restaurant, if a waiter drops a stack of dishes, people stop talking and the sound level drops. Eventually, as people start to converse again, the sound level starts to rise back up again, as the discipline to keep conversation levels low rarely lasts.

How does this apply to music rehearsal rooms? If a large ensemble is rehearsing in a room that is undersized for the ensemble, the musicians have to sit closer together. These rooms can be excellent rehearsal spaces for individual practice or even for small ensembles, particularly if those ensembles are choral or made up of instrumentalists who play less powerful instruments, such as acoustic guitar. But for larger ensembles, these spaces can become dangerously loud. The reason is that musicians, like restaurant patrons, will play as loud as necessary to hear themselves. As in a restaurant, the average sound level rises; piano sections start to rise to mezzoforte levels, and forte sections become intolerable.

As in a restaurant, the situation is dynamic—highly skilled musicians can still control sound levels in a room that is too small or too live and responsive, compensating by how loudly they play their instruments. In general, the more highly skilled the ensemble, the greater the dynamic range it expresses. A highly skilled ensemble will not play louder than the same ensemble made up of beginners…but it will play more quietly than the less-skilled group.

As acoustics designers of music rehearsal spaces, we understand that the greatest determinant of a room’s acoustics is its size. In this case, “size” reflects the total interior volume of the room, as determined by its plan dimensions and height. The plan dimensions of the room are typically set by the size of the largest ensemble that will use that space. The taller the room, the greater its size in acoustical terms. Therefore, it is always easier to control excessive loudness in taller (i.e., acoustically larger) rooms.

The room’s finishes also affect loudness control, but it is not possible to fully compensate for a room that is too small by adding more sound-absorptive finishes. One way to describe this challenge is having to choose between a room that is too loud and a room that is too dead. Musicians do need to hear themselves and they need to hear other sections of the ensemble. This is difficult to do if the room is too dead. Because most music rehearsal rooms are used for a variety of different ensembles, varying in both number of players and loudness of the instruments, we often recommend adjustable acoustics elements in the form of curtains, hinged panels, etc., so that the room loudness can be adjusted to best suit each ensemble.

It is important to recognize the dynamic nature of the cocktail party effect, not only in relation to restaurants and bars, but also to music rehearsal rooms. One can calculate the loudness level of a room or its reverberation time based on the room volume and finishes. However, the actual impact of adding more sound-absorbing material to a room may be far greater than calculated. For example, if the amount of sound absorption in a room is doubled, the sound level can be calculated to drop 3 decibels. But in reality, the average sound level at which an ensemble plays in the room with more sound absorption may be only half that at which they’d play without, and a halving in loudness generally correlates to a drop of 10 decibels. In other words, the help provided by sound-absorbing treatments may be far greater than calculated.

The cocktail party effect applies to other types of spaces, too. In future communications, I’ll explore how the cocktail party effect impacts the goal of creating an acoustically advantageous worship environment that encourages congregational participation.