In Mei-Jia & Ting-Ting & Chih-Fu & Sin-Ji, C. Spencer Yeh pushes speech synthesizers trained to represent various dialects of Chinese to the threshold of intelligibility and recognition. Through this creative misuse of such tools, Yeh questions the processes underlying codified representations of specific sociocultural identities in contemporary digital sound technologies.
Mei-Jia & Ting-Ting & Chih-Fu & Sin-Ji (2018) is an experimental composition by C Spencer Yeh for human and synthetic speakers. It presents four voices from which the work derives its name. Mei-Jia, Ting-Ting, and Sin-Ji are artificial voice templates native to the Adobe Audition program — a program that has been dubbed the "Photoshop of sound." Created to portray spoken Chinese language from Taiwan, Mainland China, and Hong Kong, respectively, these speech synthesizers are trained to represent idealizations of real-life speakers. For each template, Yeh composed scores to generate unconventional vocalizations at the threshold of Auditions's capacities. The final voice, Chih-Fu, is sung by Yeh in a similar unconventional style, imitating the warped vocalizations of the synthetic voices.
Mei-Jia & Ting-Ting & Chih-Fu & Sin-Ji is inspired by the traditional Chinese performance form of Kouji, which emphasizes vocal mimicry. By attempting to imitate the uncanny byproducts of the synthetic voices, Yeh creates his own form of Kouji to explore the capabilities and limitations of computer generated speech. In contrast to narratives of ever-improving accuracy and higher fidelity that often accompany contemporary digital technologies, C Spencer Yeh examines the abject byproducts of Audition pushed to the threshold of intelligibility and recognition. Through this idiosyncratic use of synthetic speech, Mei-Jia & Ting-Ting & Chih-Fu & Sin-Ji questions the processes that are used by composers and technologists to represent specific identities in the digital domain.