5. Results and Discussion

5.1 Analysis of reverberation times

Sound waves from sources created in some spaces within the broch may travel into the courtyard or be redirected by voids, openings or entrances. The accuracy of the reverberation results may be affected by the body mass of two sound recordists and their respective locations within the broch spaces. The nature of the source of the impulse (wood), and the choice of starting point and noise tail (during analysis) for each clack sample implies that the accuracy of the results should be regarded as approximate with high error factors (±30%).

Table 4: Clack-generated reverberation times for spaces inside Mousa broch

Position in broch Clack-generated (approx) reverb times in milliseconds Comments on structure Sound file
Cell A 613 (see Figure 6) Contains apertures
Cell B 914 (see Figure 7) Contains apertures
Cell C 343 (see Figure 8) Contains apertures
Gallery 2 329 (see Figure 9) Circular shape
Gallery 3 396 (see Figure 10) Circular shape
Gallery 4 264 (see Figure 11) Circular shape
Gallery 5 209 (see Figure 12) Circular shape
Broch entrance 327 (see Figure 13) Openings to west and courtyard
Centre courtyard 332 (see Figure 14) Open to sky
Halfway up stairs 594 (see Figure 15) Close to void set

Analysis of clack samples presented in Table 4 show that the reverberation times for all the spaces in the broch are unusually short, with 70% of the spaces calculated at under 0.5 secs and none over 1 second. These times confirmed both researchers' aural perception of the broch as a still, quiet place full of 'dead' space. Short reverberation times can be attributed to the dry stone wall construction. Stacked layers of unfinished sedimentary rocks create cavities and gaps with rough, uneven surfaces. The stone construction leads to lateral displacement and diffuse reflection of sound, in which sound energy is scattered in all directions. The thickness of the walls in every part of the broch (up to 12 feet (3.6m) at the wall's base) will increase these effects. Diffuse reflection dictates that sound waves will not be reflected unless the size of the surface elements on the stone walls are more than 4 wavelengths in size. The average size of the gaps in the dry stone material used in broch construction is difficult to assess, but the majority of stones penetrate the wall to a depth of 4 inches or more. The effect on aural space of gaps in the stone walls is best understood by imagining a 1000 Hz sound wave encountering the wall surface. The 1000 Hz sound wave is approximately 13.4 inches long. Any gaps/cavities in the wall surface 3.6 inches deep or less will effectively diffuse/scatter the energy of the sound wave. The irregular surface of the dry stone walls means that sound waves with frequencies around 1000 Hz or more will be diffused. The thickness of the walls and the irregular dry stone construction may increase this effect (influencing sound waves below 1000 Hz).

The results of the clack analysis indicate that very short reverberation times throughout the broch reflect an acoustically dry environment. Iron bar and drum samples indicate consistently fast decay times, suggesting high levels of direct sound and a lack of reflected sound wthin the broch spaces. The amplitude envelopes of sound samples (see Table 5) recorded in cell spaces imply that direct sound created in the courtyard travels into all the cells and may be audible inside cell spaces and around the wall head. The irregular surface of the dry stone walls suggests that sound waves with frequencies around 1000 Hz or more will be diffused. The thickness of the walls (up to 12ft: 3.6m), and the irregular dry stone construction may increase this effect (influencing sound waves below 1000 Hz).

Table 5: Displacement time and frequency graphs for iron bars, drum and whistle sorted by location in broch with comments

Sample type showing source and recording location for each graph Comments Sound file
Iron bars outside broch recorded in entrance (see Figure 16)
Iron bars were struck and immediately damped. Frequencies between 1440 Hz–2640 Hz peak are prominent showing some pitch content, the sound is short and decays quickly within 50 msecs, implying that little or none of the direct sound is reflected inside the open tunnel shape of the entrance (5–10ft high).
Iron bars in courtyard recorded in cell B (see Figure 17) The time amplitude wave response shows a longer vibration time, 150 ms, than the previous example, which may be due to variation in strike. Direct sound from the courtyard enters cell B (beehive-shape cell 12ft high), and creates a clear signal response.
1 hit drum centre recorded in cell A (see Figure 18) The drum head was damped after the strike. Sound lasts for 80 msecs. The time amplitude wave response for the drum sound shows the sound wave enters cell A (beehive shape 12ft high) from the courtyard and creates a typical amplitude envelope for a drum strike with swift attack and rapid decay time.
Whistle centre recorded in cell C (see Figure 19) Sound lasts for 1520 msecs. The transient rise (attack) and solid-shaped curves of the amplitude envelope show the distinctive periodic (pitch producing) nature of the whistle reproduced inside the beehive shape cell.


© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions | File last updated: Thu May 26 2011