Impossible Paintings #1

For many years I saw my digital craft as completely separate to my art practice, which has always been rooted in painting. Over time I started to recognise ways in which I could explore digital processes as extra layering techniques for my paintings, rather than a space for something new. I already look for new ways to layer different elements into each other, where no single element is distinctly in the foreground or the background.

The translation into the digital space gives us some interesting options for sure, but what put me off at first was the idea of generating new content from scratch with the digital tools, I wanted to focus on image manipulation instead so that the ‘source image’ has come from a physical painting.

There are a few main approaches that work well, one is using advanced area-selection tools - whereby you select all the pixels within a given colour range for example. With a single click you can mask out shapes that could take you days to do on the canvas. Then there’s the collaging approach, mix and matching different images at different scales. One of my favourites in the various Blending Modes. These define how the colour and value in each layer interacts with the layers below. It’s here that we’re stepping well away from digital short-cuts and into impossible colour relationships. Think glazing layers that create an inverse colour, or glazes that only appear on top of values above a certain threshold.

My favourite avenue to explore in this space is when I look at ways to take the digital result and bring it back into the physical world. For me this is an extension of weaving layers into each other, to break apart the linear progression between stages by turning back on myself.

There are 2 ways I’m currently exploring this, one is to generate a displaced 3D mesh (digitally) and then 3D print (physically) the result, creating a distinct divergence in the resulting image from the original one, but in a real tangible form.

A common approach to this would be to use LiDAR scanning (available on the last few iPhones) but the results just aren’t good enough, the ‘translation’ is to lose detail in the mesh in a way that cannot be revised without considerable effort, by which time there’s too much manual change. The idea that we’re taking the exact data, but then choosing to look at only one element (value not colour) is more interesting to me - we are abstracting something about that image rather than distorting it through the limitations of a technology.

Both are valid objectively as ways to translate the physical to digital, and I’ll definitely be looking at other ways to utilise LiDAR in this ongoing research project.

Painting to Greyscale to Displacement

By translating the image into greyscale first, we lose some information, dark blue and dark purple for example may end up indistinguishable in height. Highlights also become harsh peaks and drips become mini peninsulas.

The next is through Projection Mapping. This is a technique for projecting images onto complex atypical surfaces, such as the a building facade, festival stage design, or even a car. I have been using these skills professionaly since 2013 so I have been fortunate enough to not need a learning curve in my research and instead focus on the content.

So to project an image of a painting back onto self, and line it up perfectly, what do I do with that image? Going back to the collage techniques I mentioned above, most of the animated content was slowly cycling colours, or distorting specific areas within the image; usually being a narrow colour range across the whole image. If it retains it’s overall essence but appears also to be shifting form then I feel like I’m heading in the right direction.

I’m pretty happy with the results so far* and am looking forward to making an exhibition full of this stuff!

*I’ve noticed too that the natural next step at first seems to be to then make paintings specifically with projection in mind, but I’m not sure I want to do that! It’s going to change the way I paint and I want them to exist in their own finished state before I move on to the Digital Post-Process stage.