In a choir group, many people are singing the same piece at the same time. In this case the normal mechanisms of the auditory localization of single voices are disrupted, this is called the "chorus effect." This phenomenon is even stronger in rooms with different acoustics. To get closer to this, let's listen to these two choirs singing the same piece in a drier (less reverberant) room, for example a recording studio of a radio broadcasting station!

Sound Example 3

Sound Example 4

Now some more differences are obvious. Because of less reflections from the walls, the chorus effect decreases. We know this from our experience as singers in choirs, that it is different to sing in different rooms and sometimes more challenging to sing either in rooms with dry acoustics or in rooms with a lot of reverberation.

3. Multitrack analysis
With a special technical setup it is possible to listen to these choirs, as if there was no room around them. And, what is more, we can listen to a single-voice section of sopranos, altos, tenors and basses and besides that also to the single voices themselves.
Please listen first to the amateur choir, singing as if there was no room, then to the tenor section of choir and to a single tenor singing within this group.

Sound Example 5

Sound Example 6

Sound Example 7

If we use this audio material for further analysis, we might find some hints to the quality of the choirs and the differences between them:

Let's have a look at the behavior of the fundamental frequency (so-called F0) of the choirs. If we calculate the mean value of all the singers in the choir (which covered 16 singers per choir in this case), we see differences in their singing behavior. The standard deviation gives a clue for their accuracy in singing.

Professional choir: "Laudate Dominum" in unison, Avg F0 (dark green) and StDev F0 (light green)

Amateur choir: "Laudate Dominum" in unison, Avg F0 (dark blue) and StDev F0 (light blue)

Both choirs have in common:
- the fifth and octave are pitched a lot higher than expected
- increasing standard deviation at the beginning of each changing tone; the accuracy of pitch intonation and synchronization is low at the beginning of tones and reaches a kind of minimum value for intonation
- the bigger an interval to be sung, the larger the failing of the singers
- most of the notes are reached with over- or undershoots
- notes in ascending scales are pitched higher and in descending scales pitched flat

Differences between a professional and amateur choir
- tonal steps can clearly be seen in a professional choir, not so clearly in an amateur choir, even though the ears identify them as steps
- repetition of notes seems to be more difficult for amateur choirs, not so difficult for professional singers
- there seems to be a kind of vibrato synchronization for long notes in the whole choir, a little bit more for an amateur choir

4. Conference Presentation and Future Work

At the ASA meeting conference many more analyses of the audio material are shown in detail:
How the choir behaves, when a piece is sung more slowly, or as a polyphonic canon having a closer look at the fundamental frequency, the timing/synchronization, the loudness (sound pressure level, SPL) and the sound of the choir.
There is a huge amount of audio material which will be evaluated in the next years, concerning different pieces, singing exercises, influence of sitting or standing, vibrato behavior, influence of different vowels, acoustical interaction between neighboring singers, maximum timing precision of choir singers, aural training for conductors concerning these aspects, etc.

---

[ Lay Language Paper Index | Press Room ]