Social-EQ: Crowdsourcing an equalization descriptor map

We seek to simplify audio production interfaces (such as those for equalization) by letting users communicate their audio production objectives with descriptive language (e.g. “Make the violin sound ‘warmer.’”). To achieve this goal, a system must be able to tell whether the stated goal is appropriate for the selected tool (e.g. making the violin “warmer” with a panning tool does not make sense). If the goal is appropriate for the tool, it must know what actions need to be taken. Further, the tool should not impose a vocabulary on users, but rather understand the vocabulary users prefer. In this work, we describe SocialEQ, a web-based project for learning a vocabulary of actionable audio equalization descriptors. Since deployment, SocialEQ has learned 324 distinct words in 731 learning sessions. Data on these terms is made available for download. We examine terms users have provided, exploring which ones map well to equalization, which ones have broadly-agreed upon meaning, which term have meanings specific small groups, and which terms are synonymous.