This article examines how experimental phonetics contributed to – and occasionally subverted – the colonial classification of African languages, focusing on the case of Oshiwambo in northern Namibia and its scientific construction at the Hamburg Colonial Institute at the turn of the 20th century. Based on archival research and sound recordings from 1913 and 1953/54, it analyzes laboratory measurements by Giulio Panconcelli-Calzia and field recordings by Ernst Dammann under apartheid rule. Phonetics, often regarded as a neutral science, was deeply embedded in colonial and missionary infrastructures. The article explores how acoustic inscription, institutional assumptions, and Informant agency shaped both the production and interpretation of linguistic data. While phonetic research aimed to stabilize ethnic-linguistic boundaries, it also revealed the fluidity of orders of speech and belonging – undermining rigid classificatory regimes. By tracing how the Oshiwambo cluster emerged as a phonetic object, the study shows how recording techniques and speaker selection affected what counted as linguistic knowledge. It offers a media-historical perspective on the interplay between sound, technology, and power in colonial knowledge production.