The bride's contract specified a maximum of 85 dBA. She had underlined it twice and highlighted it in yellow. The DJ — a twenty-year veteran named Marco — read it and sighed the sigh of a man who had been through this before.

During cocktail hour, 78 dBA felt lovely. Guests conversed, the music was tasteful. The bride beamed. Then the dance floor opened.

At 85 dBA, nobody danced. They stood in clusters, chatting politely. At 88 dBA, a few brave souls shuffled. At 92 dBA — where Marco nudged it when the bride was taking photos — the floor finally came alive. Then the bride returned.

'I SAID 85!' she shouted at Marco. She was shouting because the ambient noise of 150 chatting guests was already 82 dBA. The Lombard effect had taken hold — everyone was raising their voices to be heard over everyone else raising their voices.

Marco showed her the meter. Music: 92 dBA. Crowd noise with music off: 84 dBA. The guests were nearly as loud as the DJ. The bride looked at the data, looked at her packed dance floor, and made the only rational decision: 'Fine. 95. But NO higher.'

By midnight it was 101 dBA and the bride was leading a conga line.

The Moral: The Lombard effect means crowd noise often exceeds your target limit. Use SonaVyx's real-time SPL monitor with Leq logging to document actual levels and show clients the data — it's the best mediator you'll ever have.