The paper presents a project that focuses on activities of interpretation of visual data. Not only in sociology but also in other disciplines and also in occupations and vernacular fields the interpretation and analysis of videos is a everyday practice. E.g. policemen try to understand footage from crime scenes or sports trainers analyse the mistakes of their teams. This paper shows how those video analysts generate meaning based on video data, how they use specific technology and how those practices can be compared to video analysis within social science. (By comparing it, the video-ethnography becomes self reflexive, this is a starting point where science studies can start to think about its own interpretive practices.) The research shows how the participants produce meaning by reconstructing the video material using their body movements as devices for reenactments, how they highlight and frame certain details on screen and how they generate meaning by bringing it into narrative order. Interpretation is understood as a communicative process, that can be understood by an ethnography that focusses on communicative actions. Methodologically the project combines a "focused ethnography" that will be discussed with the analysis of video recordings of certain key events such as joint data sessions.