Talk us through how you created the images.
We decided to take an artistic approach and work from overall composition and colors, before concentrating on the details. This allowed us to focus on the creative side of the project, without worrying too much about technical aspects of shading and rendering.
Using a 3D CAD model from one of our previous jobs, we quickly camera matched the car to the backplates, and selected several different angles we liked.
We then narrowed it down to two images. Ian made a quick grade to the shots to establish an overall look and feel of the images, and I worked on the lighting, painting out unwanted areas on the HDRI, and adding extra lights to the scene to enhance the car shape. I built a quick proxy geometry of the street and the building to get the correct reflections on the car, and that was pretty much it in terms of the scene setup.
Of course, I spent some time prepping the CAD model for rendering: I split the model into elements, assigned shaders, and tested different materials. We relied on the VRscans and almost completely skipped the shading part on this job. The majority of our materials were VRscans, and for the rest we used shaders from our library.