Using multiple photorealistic models, the right prompt, and I upscaled the image.It did also take a long time to get the result I wanted, must’ve been around 50-100 retries while tweaking the prompt and the different LorA’s.