Many modern auditoria pose a significant challenge to the acoustic designer. This is because the need for ultimate flexibility often means that the same space must be suitable for both speeches and musical recitals. Unfortunately, the acoustic requirements of these two activities are not only different, but often conflict, leaving the designer with many difficult choices to make.

Requirements for Speech

The aim here is for good speech intelligibility at all seats within the space, such that each spoken syllable is heard separately and not blurred together, and that the speaker is loud enough to be heard without strain. The sound heard by an auditor is a blend of direct sound straight to the seat, early reflections from the walls and ceiling, and reverberant sound which results from multiple longer term reflections within the space.

Direct sound is a function of the distance between speaker and seat and will be louder when the distance is reduced. Reverberant sound has been traditionally measured in terms of the time it takes for the sound to fall by 60dB (i.e.: to one millionth of its initial value). This time period is usually independent of speaker-seat distance, depending instead on the volume of the space and the sound absorbing properties of its surfaces and contents. It also varies with frequency.

Research shows that a Reverberation Time (RT) of around 1.1 seconds at mid-frequencies is appropriate for speech in small to medium spaces, and that the RT should not rise too much at low frequencies as this lowers intelligibility by an effect known as 'masking'. With good early reflections this figure becomes more flexible.

An audience's expectation regarding the actual quality of the speech signal is usually not too critical, as long as the speaker's voice and accent are clearly recognisable and any vocal information is understandable.

Requirements for Music

Music audiences, on the other hand, have inherited quite a developed expectation of particular sound qualities for various styles and eras of music.

There are many subjective elements involved in listening to music, and a plethora of terms exist by which these elements are expressed. A problem of much acoustic research has been to relate these subjective terms to quantifiable aspects of design. Such terms include warmth, blend, clarity, etc. Barron in his book on auditorium design notes:

"…clarity should be adequate to enable musical detail to be appreciated,…reverberant response should be suitable, so that the listener perceives himself to be surrounded or enveloped in sound, the listener should sense the acoustic experience as intimate and he should judge it as having adequate loudness."

????? Barron

Whilst definition is a prerequisite for speech, with music, we expect more blend within the separate bursts of sound. Excessive clarity in music auditoria gives the subjective impression of brittleness or dryness and accentuates unwanted bowing or fret noise, making the musicians job even more difficult.

One immediate conclusion from this is that rooms for music will be expected to have longer reverberation times than rooms for speech. RTs up to 2.4 seconds occur in large concert halls, and a value of 2s is typical in recent work. The currently accepted optimum range is 1.8-2.2s, with a volume per seat of 8-10m³. Even this is, to an extent, a matter of debate as music in which clarity of line is paramount would benefit from shorter times, and chamber music halls are typically much dryer.

Research in the last few years on subjective impression has emphasised the importance of early lateral reflections, that is sound from the side walls which reaches the listener within the first 80ms after the arrival of the direct sound. This aspect explains why rectangular spaces get better audience support than fan shaped, for instance. Warmth in a space is associated with bass response, and music auditorium tend to have a longer RT at the lower end to encompass this.

Another difference is that music can consist of a great range of frequencies (20 Hz to 20 kHz) whilst speech is basically a narrow band signal (500 Hz to 4 kHz).

The Conflict

It is clear that there can never be a space which is totally ideal for both speech and music, unless it is adjustable either in volume or in the nature of its surfaces to change its reverberation time. One use usually has to take precedence. Unless electronic amplification is to be in constant use, this tends to be the speech application.