Inspired by the bold methodological experiment on auto-ethnography in Vienna (Rhodes et al. Human Arenas, 1, 151–165, 2018), we conducted several walking sessions in St. Petersburg. During the walk, each participant took photos of objects, which attracted his/her attention, and dictated to the voice recorder what he/she was experiencing at this moment. The data allows to expand and supplement the existing ideas about individual differences in the perception of urban spaces, which are based on both the influence of the subject’s past experience and deeper, possibly neurophysiological factors causing these differences. The results we compared with theoretical models of the structure of the image of the world elaborated in Leontiev’s school in context of the activity theory. Among the main results of our work, the participants noted two effects that, being a sort of by-effect of the investigation, turned out to be extremely important and interesting for them. First, the self-reflection of the impressions of the environment turned out to be a means of self-discovery. It brought some new knowledge about oneself, sometimes surprising, helping to reveal something about oneself, which was not obvious before. The second unexpected result was a kind of “therapeutic effect” of our walks, noticed by the participants. The possibility of “mediated” self-discovery while focusing on an informational and emotionally saturated object of reality may have potential as a therapeutic practice.