Abstract: Road traffic noise presents a significant environmental challenge as it affects human health. This study examined the timbre attributes of road traffic sounds captured in Nairobi city to highlight tambourine-like zills (jingles). This was done by employing a unique perceptual metaphor, likening them to the blended resonant and shimmering elements of a tambourine. Utilizing audio recordings from 42 sampled sites across Nairobi’s road network, the study processed about 2,185 segments to compute essential acoustic parameters such as spectral centroid, spectral roll-off, Mel-frequency cepstral coefficients (MFCCs), and zero-crossing rate (ZCR). These metrics were linked to equivalent sound levels (Leq) from on-site assessments and passenger car unit (PCU) indicators to reveal timbre differences influenced by vehicle categories and flow patterns. Findings demonstrate that heavy vehicles like trucks and buses feature modest spectral centroids (800–1100 Hz), limited spectral roll-off (1200–1600 Hz), and lower ZCR (0.04–0.07), resembling the sustained, foundational tone of a tambourine’s drum surface, especially in elevated Leq scenarios (e.g., 84.2 dBA along Thika Road). On the other hand, motorcycles and low-capacity passenger service vehicles (PSVs) present higher spectral centroids (approximately 1800 Hz), expanded spectral roll-off (2200–3200 Hz), and increased ZCR (0.09–0.14), akin to the crisp, metallic accents of tambourine cymbals, more common in moderate Leq contexts (e.g., 68.7 dBA near Uhuru Park). The MFCC evaluations highlighted unique spectral shapes, with reduced averages for heavy vehicles denoting subdued timbres and greater deviations for motorcycles, thereby suggesting abrupt changes. A graphical representation of spectral roll-off across vehicle groups illustrated a gradient from deep resonance to sharp percussion. In conclusion, findings provide practical guidance for noise reduction strategies. This is particularly important for guiding the modulation of the acoustic soundscape for inclusive, human-centered noise abatement planning. This resultant "tambourine model" is a tool for noise perception.

Keywords: Road Traffic Noise, acoustic parameters, timbre, tambourine, jingles, modelling, acoustic descriptors

Pages: 833-849