A-Rumenoy — DSSA: Scobie's Synthetic Voice Song from a Journal

Journal feature / DiVA No. 57

From an AI voice experiment to a music release.

The experimental artwork behind DSSA—combining synthetic singing, character animation, and music—was featured in DiVA, the journal of the Society for Art and Science. This release carries that research from the printed page into your headphones.

Read the English abstract

Abstract. This project investigates an AI-assisted workflow for creating experimental audiovisual art from diverse source materials. Its central study focuses on synthetic singing: a source vocal is transformed through retrieval-based voice conversion (RVC) into the voice of Scobie, an original Scottish Terrier character. AI-assisted 3D generation, avatar animation, lip-sync, and musical arrangement are integrated with human direction and editing. The resulting song, DSSA: Scobie's Synthetic Voice Song from a Journal, reframes the journal-featured experiment as a standalone music release and explores character identity, authorship, and the expressive potential of synthetic voices.

Keywords: synthetic voice, RVC, AI-assisted music, virtual character, audiovisual art

DiVA No. 57 journal page featuring the experimental AI artwork behind DSSA

Published feature · view full page ↗ Journal of the Society for Art and Science, DiVA No. 57 · p. 23