Looking for old gems
Last week, I started working on a new piece using some photographs I captured almost 6 years ago, the pictures were taken on a footpath close to Wimbledon train station, very close to my former home. At that time, I was experimenting with Photogrammetry and I wanted to test how much of the scene I could capture. Needless to say, these photographs just got lost under thousands of files piled up on top to only reappear recently.
I was going through my collection of old hard drives looking for lost treasures and I stumbled upon a folder full of image sets taken in places around my old neighbourhood. This particular set triggered some memories about a footpath behind the house, this particular segment was on the less-visited area, very close to a bridge that was in bad shape and was hardly used by residents.
I was browsing through my old hard drives looking for lost gems when I came upon a folder full of image sets taken in different locations around my former neighbourhood. This particular set brought back memories of a trail behind the house; this particular segment was in a less-travelled region, quite close to a bridge that was in poor condition and hardly used by locals.
Nowadays I am using photogrammetry software that is a lot better than the one I was using at that time, my computer is also a lot better than the one I used to have at that time so I was curious to see if those photos were good enough to produce an interesting model.
Glitchy models
I was expecting a “glitchy” model as I knew the number of images was far from optimal for this particular scene. I gave it a go and the result was a lot more interesting than I expected; the resulting model was actually divided into two pieces, one for the foreground and one for the background.
“Glitchy” models are quite common when doing photogrammetry, over the years, I have collected a bunch of them, they all share a certain aesthetic; a crisp and accurate core surrounded by weird fragmented low-res areas and gaps, triangles and quads become bigger as the model disintegrates and textures lose detail until becoming simple gradients. I have to admit that this model wasn’t the exception.
For a long time, I have asked myself if “crisp” photogrammetry models are the only ones worth keeping and using on projects, my answer has always been no, however, I always felt I needed to work on these models “glitchy” models as I wanted to make them my own instead of conforming to the default aesthetic, this edit work might also present new opportunities.
Flowing matter
I do like the idea of 3D models disintegrating, this opens the possibility for them to perhaps mutate and become something new, in this particular case, I wanted to play around with the materiality, sculpting the model to make it more 3D and dynamic, also wanted to explore creating new versions of the original material.
I’ve decided to really push the materiality by adding some sort of bulky/foamy finish to the mesh, then I started experimenting with the materials by re-projecting/tweaking the UV maps, I also worked on the material properties to add an oil paint finish and this is the result: