Default Masking Scheme | PhMuseum

This collection begins with my very own images of landscapes and examines how automated recognition transforms them into new pictures. By transforming the masks generated throughout the recognition course of via layering, softening, and recomposition, I create lands

I’ve all the time wished to protect the preliminary attraction of a picture. Before viewing turns into evaluation, these works seem merely as seductive, ambiguous, and flattened landscapes. Yet when traced again to their making, that attraction emerges from a working course of constructed upon automated picture recognition.

In this collection, I discover how automated recognition programs interpret landscapes, and the way that interpretation reshapes photographic pictures. Every work begins with my very own images as supply materials, alongside masks generated via Camera Raw’s computerized panorama recognition throughout the enhancing course of.

My course of begins with these mechanically generated masks. Rather than preserving the clear boundaries produced by the algorithm, I repeatedly overlay, soften, and reorganise them, reworking the traces of recognition into new picture buildings. Instead of presenting the software program’s workflow as an finish in itself, I attempt to retain the pictures’ visible enchantment whereas looking for a stability between intuitive viewing and an consciousness of their constructed nature.

The purple overlay is derived from the default color used for masking, which itself originates from the Rubylith masking movie traditionally utilized in printing and plate-making. Once a brief and largely invisible stage inside picture manufacturing, this color is reintroduced as a part of the ultimate picture. It preserves traces of choice whereas making an in any other case hidden working course of seen once more, connecting modern digital recognition with earlier strategies of picture manufacturing.

I hope these works are encountered first as pictures moderately than understood instantly as technical experiments. As questions on their making progressively emerge, one other layer of which means turns into seen: a strategy of automated recognition that doesn’t originate in nature however constructs its personal model of it. The works document not solely the panorama itself, but in addition the methods machines outline panorama and translate these definitions into visually compelling varieties. Rather than accepting the algorithm’s output as a remaining outcome, I deal with it as new visible materials from which the work continues to develop.