Spectogram of Dag Kargiraa singing
Spectogram of Dag Kargiraa singing

Spectogram of Dag Kargiraa singing
These films show spectograms of various recordings. Spectograms are representations of sound on three axis:

- Time is on the horizontal axis
- Frequency is on the vertical axis
- Loudness is represented by line color and thickness

In effect, spectograms show the spectral composition of the sound and its evolution in time.

One can read various things on a spectogram, like the precise frequency of a sound, or the time it takes for a speaker to utter a given syllable. Spectograms are also useful to help untrained listeners "hear" specific sonic features and stay focused on them. Some people for exampe have trouble hearing the tunes produced with overtone singing. In my experience, spectograms would help them to distinguish the melodic linem, especially in techniques such as Dag Kargiraa.

The films are screen recordigns of Sonic Visualizer displaying the spectrograms (sonic visualizer can also export still images but of course you loose the synchro with the sound). I supply the settings of the analysis in case you want to reproduce it yourself and tweak it further. 

 

Russia (Tuva) : Dag Kargiraa technique

This is one of the five overtone singing styles practiced in the republic of Tuva, and the one that has the lowest fundamental. From the disk Voix du monde/Voices of the World, Chant du monde / CNRS, CD II, track 37. See this link for the complete disk and the pdf booklet.

Sonic Visualizer settings: Default color scheme; Scale: dBV^2; Window: 8192; Overlap: 87.5%; All bins; Linear scale.

South Africa (Xhosa)

While overtone singing is usually done on a stable fundamental tone, here the singer alternates two fundamentals. The result is particularly intricate as the two melodic lines (the one of the fundamentals and the one in the overtones) seem to evolve independently from each other. From the disk Voix du monde/Voices of the World, Chant du monde / CNRS, CD II, track 36. See this link for the complete disk and the pdf booklet.

Sonic Visualizer settings: Sunsetcolor scheme; Scale: dBV^2; Window: 8192; Overlap: 87.5%; All bins; Logarithmic scale.

France (Britanny)

Kan ha diskan, litterally "singing and counter-singing" by Yann-Fänch Kemener and Marcel Guillou. The first singer sings one line, then the second singer joins him on the last sounds and repeats the line. The first singer takes his breath meanwhile and come back in on the last sounds of his companion, continuing with the following verse. Such a performance can last for dozens of minutes. It is meant to accompany dance. One can actually hear on the recording the steps of the dancers which translate as regular vertical "bars" of non-harmonic content.

This recording illustrates how a "familiar" tone of voice looks on the spectogram (useful before looking at overtone singing and the like). In the first seconds, before the rhythm kicks, there are also some nice illustrations of the way vowels sung on the same fundamental differ from each other.

From the excellent disk Celtic Mouth Music, gathered by Matthew Kopka, Ellipsis Arts 1997.

Sonic Visualizer settings: Default color scheme; Scale: dBV^2; Window: 8192; Overlap: 75%; All bins; Logarithmic scale.