Search

The intelligibility of sung text is an important component of listeners’enjoyment of vocal music and a central concern for singers and, for example, choral conductors. Expert listeners such as singers and singingteachers may be better than non-singers at perceiving sung text. We replicated and extended an earlier study investigating the intelligibility ofsemantically and non-semantically meaningful words performed soloand by a small group of trained soloists by carrying out an experiment inwhich we manipulated listening expertise, type of text, number of singers, and time of hearing. Participants listened twice to four songs withmeaningful and “scrambled” lyrics, sung in unison by a choir and solo,and wrote down as many of the words as they could discern. Singers werebetter at the task than non-singers; more words were recorded on thesecond hearing and when the words were meaningful. Sung text involvesdistortions of consonants and vowels to which singers may be more accustomed, so that they find it easier to discern texts even when scrambled. Choirs may be harder to understand than soloists because theirphonemes are more variable and less clear. In future research we will useoperatically trained soloists and polyphonic choirs.

An important aspect of perceiving sung music is understanding thewords. Previous research has suggested several factors affecting the intelligibility of sung text. This study investigates two of those factors: thenumber of singers and the singing expertise of the listener. We expectedmore singers to cause greater variability in the acoustic signal and beharder to comprehend. Listeners who are themselves experienced singersare more likely to be attuned to factors affecting singers’ diction and wereexpected to be better than non-singers at understanding the sung text.Forty eight participants, half accomplished singers and half self-reportednon-singers, listened to four 8-bar unaccompanied songs twice (in orderto test for familiarity) and wrote out the texts as they heard them. Twoperformances were given by a soloist, two by a trio of singers in unison.Participants were significantly better at understanding the words on thesecond hearing than the first, and singers significantly better than nonsingers overall. There was no effect of the number of singers. Hence familiarity and singing experience both benefited sung text understanding.An effect of the number of singers may be more apparent when comparing a soloist with a choir.

Wood smoke in the atmosphere often accounts for 20–30% of the ambient fine-particle concentrations. In communities where wood is burned for home heating, wood smoke can at times contribute the majority of the atmospheric fine-particle burden. Chemical mass balance receptor models that use organic compounds as tracers can be used to determine the contributions of different emission sources, including wood smoke, to atmospheric fine-particle samples. In order for organic chemical tracer techniques to be applied to communities across the United States, differences in wood smoke composition that arise from differences in the type of wood burned in various regions must be understood. A continental-scale accounting of particulate organic compound emissions from residential wood combustion has been constructed which helps to quantify the regional differences in wood smoke composition that exist between different parts of the United States. Data from a series of source tests conducted on 22 North American wood species have been used to assemble a national inventory of emissions for more than 250 individual organic compounds that are released from wood combustion in fireplaces and wood stoves in the United States. The emission rates of important wood smoke markers, such as levoglucosan, certain substituted syringols and guaiacols, and phytosterols vary greatly with wood type and combustor type. These differences at the level of individual wood type and combustion conditions translate into regional differences in the aggregate composition of ambient wood smoke. By weighting the source test results in proportion to the availability of firewood from specific tree species and the quantities of wood burned in each locale, it is possible to investigate systematic differences that exist between wood smokes from different regions of North America. The relative abundance of 10 major wood smoke components averaged over the emissions inventory in different regions of the United States is computed and then used to illustrate the extent to which wood smoke composition differs from region to region in North America.