I want a picture that resists being placed, not obviously of a particular era or a typical scene. So I work by collision: a Romantic, almost Pre-Raphaelite pose, with its lace ruff and red coat, meets a darkroom-era colour experiment, meets a digital artifact, and the friction between them creates meaning as much as the figure does. The “shifting spectrum” is partly literal — the separated colour channels — and partly that refusal to settle into a conventional path that is instantly recognisable. When it works, the viewer can’t quite date or place what they’re looking at, and that uncertainty is the point.

I’m a great fan of photographic history, and I think huge amounts of interesting, experimental work from the early decades of the twentieth century have been forgotten — often in favour of pure messaging. I like to explore where those earlier experiments might have led if we’d taken them further, or had today’s technology. The channel shift won out because it isn’t really a “glitch” at all: it’s the colour separation that colour photography is built on, pulled apart and made visible. Other distortions that I tried felt more like decoration; this one was closely aligned with the medium itself, which is why I stuck with it.

I always start with a basic idea — a context, an approach, a composition, a fusion — that I want to probe for its possibilities; for where it might take me, the model, and the materials at hand. It’s a collaborative and serendipitous process. I have great faith that if you choose a good starting point and stay open to the play of elements, amazing pictures can result. Often, as in this case, I try to infuse a painter’s feel with a photo’s exactitude—a reassuring sense of classical authority struggling with a modern uncertainty about technology and its risks.

Carla was staying with us and helping on our small olive farm when I was circling this idea, and she carried the initial concept in herself — that copper, Pre-Raphaelite colouring, the kind of face Rossetti might have painted. So she gave me the tradition to begin with, and the distortion was applied to it. Carla has a vivid, particular presence that resists being abstracted, and I wanted that resistance: the more real the person underneath, the more the glitch would read as something done to a living face rather than just a graphic effect.

When we see an image that’s super-clear, hyper-realistic, we assume it must be true. Intensity and sharpness are a form of hidden persuasion. I’ve painted in a hyper-realistic fashion, and I’m not convinced the result says anything more or truer about the subject. I’m starting to think what we actually want is a picture that makes the viewer look longer and harder, where it’s less obvious what’s real. That’s what I attempted here, and you can see it most where the channels pull apart across the eyes. Carla’s skin stays warm and alive, but her gaze doubles and slips, so the one feature we instinctively search to decide whether a face is being honest is exactly the part the image won’t let us trust.

Honestly, it expanded after. I made the piece out of an older preoccupation — that images have always been built to persuade, and that even early photography traded on a “truthfulness” it never really possessed. The deepfake anxiety is just the latest and loudest version of a very old problem. We can say it went back to many portraits of powerful figures in history. The Tudor king, Henry VIII, assumed this about his fourth wife. So I’m not responding to the news cycle so much as finding that the news cycle has caught up with something the medium was doing from the start. If anything, our moment makes the older point easier to see: perhaps we’re safest with an image that admits its own unreliability rather than hiding it.

Specific encounters between things that shouldn’t quite fit. A studio portrait sitting wrong against a colour-separation error. A Renaissance gesture surviving inside a corrupted file. I’m drawn to the moment when several traditions or ideas are forced into the same frame and nothing quite wins — that unresolved collision is where the interesting result lives for me. I trust it more than the clean, almost unconscious execution of a single oft-repeated theme, because the collision stimulates the viewer to do their own looking, their own thinking.