Visual anthropologists are aware that we can effectively use cameras (photo or video) as methodological devices. We can document events, create relationships, and present our findings in visual format. But few of us have discussed the conceptual capacity of the images we create in ethnographic research. Taking MacDougall's argument that images produce a different kind of knowledge, I use the example of a film I made about a Brazilian immigrant in Lisbon to explore how the process of filming has helped me to conceptualize some of my ethnography. I contend that the double potential of images (as indexes and icons) can result in a specific mode of analytical approach. In the case I highlight here, filming Eugenia's first visit back home has made us both realize the transformative nature of this event to her perception and practice of homemaking.