*Razif Mohamed

Abstract: SonicLens is an artificial intelligence application designed to assist design students as they encounter the complexities of sound design. Within design education, critique is widely recognized as the signature pedagogy that shapes how students reflect, iterate, and refine their creative work. However, many novice learners find the technical vocabulary of audio engineering difficult to interpret, while instructors often face challenges in providing the frequent, high-resolution feedback required in large studio settings. This logistical gap limits the ability of students to refine their work through iteration, which is fundamental to the creative learning process. SonicLens addresses this by merging automated audio analysis with explainable guidance, translating abstract audio concepts into visual design metaphors grounded in cross-modal perception research. These metaphors connect unfamiliar auditory principles such as dynamic range, equalization, and compression to familiar visual practices like contrast, color balance, and exposure control, making technical ideas more approachable and encouraging students to reflect with greater confidence. The system utilizes Explainable AI (XAI) and the XAIxArts framework to ensure that machine-generated feedback remains transparent and pedagogically sound. A modular six-stage prompt-chain architecture governs the AI’s feedback generation, ensuring consistency, transparency, and alignment with the pedagogical objectives of the design studio. By providing AI-guided critique alongside human input from instructors and peers, SonicLens extends the reach of formative feedback beyond the physical classroom while preserving the relational nature of design critique. This paper presents the pedagogical grounding, system design, and technical architecture of SonicLens, situating it within broader developments in AI-augmented critique. Formal empirical evaluation is planned as future work.

Keywords: Soniclens, explainable AI, sound design education, visual metaphors, formative feedback, XAIxArts