A proposed technique for measuring and comparing crowd noise from sporting events

Crowd noise from sporting events is of interest to casual fans and noise control engineers. Fans of sports teams strive to create a noisy environment to distract opponents. Many ad-hoc noise measurements made at sporting events and widely reported by the media and use misleading sound metrics, such as peak level at a single location. The ear is an exponential averaging filter, thus peak levels are not representative of human response to noise. In addition, levels measured at a single location can be significantly skewed by very few crowd members in the measurement's near field. When comparing noise between stadia, these effects are amplified. This paper proposes methods for appropriately measuring and comparing crowd noise from stadium-to-stadium. The method focuses on spatially averaging sound pressure levels over a large area near the playing surface, computing exposure levels to compare crowds over differing time durations, and using ear-appropriate metrics such as C-weighted maximum levels using exponential averaging. Subjective loudness and percentile levels are also explored.