{"id":886,"date":"2023-10-03T12:23:06","date_gmt":"2023-10-03T12:23:06","guid":{"rendered":"http:\/\/liipetti.net\/visual\/?page_id=886"},"modified":"2025-04-10T13:46:45","modified_gmt":"2025-04-10T13:46:45","slug":"confused-cityscapes","status":"publish","type":"page","link":"https:\/\/liipetti.net\/visual\/confused-cityscapes\/","title":{"rendered":"Confused cityscapes"},"content":{"rendered":"\n

Confused cityscapes<\/h2>\n\n\n

\"\"<\/a>\"\"<\/a><\/p>\n\n\n

In 2021-2022 I spent much time creating unstable cityscapes, slice wise mixed of actual places, which were constantly changing. There were recognizable features from actual places, but the place in the image as such did not exist anywhere. Furthermore, watching the image as it changed and evolved there came moments when the your perception of the view had to change completely, even if the image had changed just so slightly. An open field might become a lake, or a wall would turn into a street.
<\/p>\n\n\n\n

Here, a recent interactive work showing a such a slowly evolving image. Various city views appear mixed with each other in an odd way, continually changing in minimal steps. You can recognize places but something is still not quite right. The image keeps changing slowly, but the spectator can also grab a rotary dial to move back and forth, looking for the best mix. In this exhibition, I had included a selection of local views to make the work more meaningful in the context of the exhibition.
<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

How do I do it? It is actually quite simple. Similar to my Scratches<\/a>, I spoil digital photos with virtual scratches or noise and train a model to remove the defects. I then noticed that doing crossfades of the spoiled images and then using the model to “restore” the image produces interesting blends of the scenes. The actual implementation then is more complex, but basically built on this phenomenon.<\/p>\n\n\n\n