A Multistep Approach

This is a workflow to use an existing idea and to create a really good looking Compositioned Image with RevAnimate
First Image: Generate an idyllic background of your choice.

Second Image: Composition all the elements (Background, Beer-Mug, Willo) in you favourite Image editor.

Third Image: Generate it with Stable Diffusion in img-img using a decent prompt to guide the model.

Fourth Image: Upscale to required resolution

Please Note: No LoRa were used, a LoRa for Willo exists, but not when I did this.