Population-Aligned Audio Reproduction With LLM-Based Equalizers

TL;DR

A Large Language Model (LLM)-based alternative that maps natural language text prompts to equalization settings enables a conversational approach to sound system control and indicates that LLMs could function as artificial equalizers, contributing to the development of more accessible, context-aware, and expert-level audio tuning methods.

Abstract

Conventional audio equalization is a static process that requires manual and cumbersome adjustments to adapt to changing listening contexts (e.g., mood, location, or social setting). In this paper, we introduce a Large Language Model (LLM)-based alternative that maps natural language text prompts to equalization settings. This enables a conversational approach to sound system control. By utilizing data collected from a controlled listening experiment, our models exploit in-context learning and parameter-efficient fine-tuning techniques to reliably align with population-preferred equalization settings. Our evaluation methods, which leverage distributional metrics that capture users' varied preferences, show statistically significant improvements in distributional alignment over random sampling and static preset baselines. These results indicate that LLMs could function as "artificial equalizers," contributing to the development of more accessible, context-aware, and expert-level audio tuning methods.

Figures (13)

• Figure 1: Illustration of the interface used for the data collection experiment. Each participant listened to audio through headphones, interact with the interface through a tablet, and attempt to satisfy the presented sentence through audio adjustments.
• Figure 2: Example JSON snippet from the collected questionnaire data.
• Figure 3: The filters that comprise the equalization controller. A horizontal movement in the interface introduces a "smile curve" effect (left) while a vertical movement applies a linear adjustment (right).
• Figure 4: Outline of RAG and RAG-QA recommendation approaches. The sentence similarity is computed as the dot product between the embedding of the input sentence and the embeddings of the database.
• Figure 5: Parameter Efficient Fine-Tuning with regression head. The figure on the left illustrates the details of Prefix Tuning and LoRA methods, while the figure on the right showcases how the final hidden state of the LLM is mapped to Beosonic coordinates.