Three fifteen year old Gmelina arborea trees with straight boles were purposively selected and felled at Onigambari Forest Reserve, Ibadan, Nigeria. From each tree, we collected two bolts of 60cm from the base of each tree to make the hourglass-shell designs.
The Design Of The Wooden Hourglass Shell
The bolts obtained from the G. arborea wood were taken to a local wood carver for the construction of the hourglass-shell needed for making the talking drums, at Origbegi compound, Idi Ikan, Ibadan, Nigeria. The manual carving and shaping was done on the wood samples to form the six different hourglass shell geometries (G1 – G6) measuring 28cm in length, 15cm in diameter, and a thickness of 0.6 cm. The geometric shapes considered were Funnel, Twofoil, Inverted Crescent, Cone, Half-ellipse, and Conic Rectangular shapes. In the carving procedure, both ends of the shell were opened since the talking drum is a membranophic percussion instrument with two skins (Gravrand 2012).
Additionally, we introduced three different finishings to the interior surfaces of the six hourglass-shell geometries to complete the design, thus making a total of 18.
Preparation Of Interior Finishes
The three different interior finishing surfaces that were used in this research for the hour-glass shells were:
- Rough surface: carving instruments such as chisel were used to make the interior surface rough during the carving process
- Smooth surface: the use of sandpaper was adopted to make the interior surface smooth. The sandpaper was applied until adequate smoothening was ensured
- Glazed surface: sandpaper, sanding sealer, and varnish were used to ensure a glazed surface. Sandpaper was first applied on the surface, after which sanding sealer was used prior to application of the varnish to seal the pores and ensure smoothening. This process was repeated few times until an appropriately smooth surface was ensured. The sanding sealer was allowed to dry off the surface for 10mins before the application of the varnish. Hence, a glazed surface.
The Production Of The Talking Drums
Upon completion of the designed hourglass shells, an already prepared cow womb (ole) similar in type and thickness were used as the leather cover for the open ends of the 18 shells. They were firmly laid on both sides of the hourglass shell using an adhesive. Then, the leather was sewed to string of the tension cord and the membrane with a thread. To facilitate the tuning of the talking drum, the tension cord was held together with the shell using adhesive. Hence, the production of the talking drum.
As a regular practice, the talking drums were air-dried for two days in the open air under the sun after which the pegs were detached to align the tension cords. The major materials used in the production of the talking drum are shown in figure I
Sound Properties Of The Talking Drums
The sound experiment of the drums was carried out in a partly closed and silent laboratory room – this is to ensure no noise interference from the surroundings. Also, we ensured that no sound signal on the Fast Fourier Transform (FFT) analyzer before the experiment was started. A professional drummer was hired to make strikes on the talking drums, at no extension on the tension cord (ne), and at the extension limit of the cord (el). All sounds generated were then recorded using Audacity software and carefully analysed by the FFT. The sound properties measured were frequency, amplitude, and sound damping time while timbre harmony was calculated using Eq. 1 (Olaoye 2021).
$$\sum _{\begin{array}{c}i=1\\ j=2\end{array}}^{n} \left|\frac{{f}_{j}-{f}_{i}}{{\sum }_{\begin{array}{c}i=1\\ j=2\end{array}}^{n}\left({f}_{j}-{f}_{i}\right)}-\frac{1}{n-1}\right| ÷{K}_{n} \left(1\right)$$
Where
i = 1, 2… n
j = 2, 3…n
f = natural frequency
n = number of frequency observation (6)
Kn = Kay’s constant = 1.6 (n = 6)
Experimental Design
The experiment design adopted was a 2x3x6 factorial experiment in a split plot design.
2 = factor A - extension of tension cord
3 = factor B - Interior finishing
6 = factor C (hourglass-shell geometry)
$${Y}_{ijk}=\mu +{A}_{i}+{B}_{j}+{{C}_{k}+ABC}_{ijk}+{E}_{ijk} \left(21\right)$$
Yij = Observation, µ = Mean
Ai = Effect of factor A
Bj = Effect of factor B
Ck = Effect of factor C
ABCijk = Effect of interaction between AB, BC and ABC
Eijk = Error term
Descriptive statistics and analysis of variance were used to determine the significance of the factors and variables.