Like previous Stable Diffusion models, you can bash photos together and then with SDX1.0 and a low denoising strength render one consistent image. Here is a funny example of this technique.
Two separate pictures are combined with a tool like Photoshop:
Then it’s rendered (in this case) through the Img2Img tab in Automatic1111 WebUI with a low denoising strength, using the SD XL refiner model, merging the two photos together and creating something consistent and unison.
Parameters:
Positive Prompt: realistic photo of a mushroom man, pareidolia, with a forest background
Negative prompt: drawing, painting, blurry, 3d, crayon, sketch, graphite, impressionist, noisy, canvas frame, cartoon, ((b&w)), Photoshop, video game, 3d render
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 2251206210, Size: 1300×865, Model hash: 7440042bbd, Model: sd_xl_refiner_1.0, Denoising strength: 0.5, Version: v1.5.1