Long-time average spectrum characteristics of different choirs in different rooms

Long-time average spectra (LTAS) were made of a boys' choir, a youth choir and an adult choir, singing in a large church, a basement room and a rehearsal hall. The effects of room acoustics, vocal effort, musical material and choir type on the LTAS were evaluated. Large spectrum effects were caused directly by differences in room absorption. Such differences also induced compensatory changes in the effort levels of the youth and adult choirs, but not of the boys' choir. In addition, all choirs modified their voice usage, apparently by raising the larynx in the more absorbent rooms. The maximum level difference between pianissimo and fortissimo varied from about 12 dB for the boys' choir to 20 dB for the adult choir. Boy singers had a somewhat steeper spectrum slope than the other singer types. The LTAS influences of the musical material were negligible. The familiar spectrum-slope effects of power changes were measured in detail and described using a frequency-dependent gain factor. The frequency characteristic of this gain factor was similar but not identical for the three choirs. A change in the sound propagation path, induced by holding up the music binders, resulted in a 3 dB loss of treble but also in a 2 dB boost in the upper bass region. The subjective effect of this change was very different in different rooms.