I've got a tricky job.
At the minute all I have to work with is some elevations and a very rough site plan.
It's a central courtyard surrounded by railway arches. Photo attached.
Once the site is fully cleared, I'll be taking a photo from ground level.
The client then wants the arches to be occupied with glass fronts and bar/restaurant interiors. They also want the external courtyard area to be populated with chairs/tables/people/food carts/planters etc etc. Ultimately they want this space to look like a busy saturday afternoon.
How would you go about it? Modeling everything isn't an option because we don't have the drawings/information to do so. It also doesn't make sense because it exists and will be clean, new and empty when I take the photo. So i'm thinking along the lines of shadow catchers or using camera map per pixel. But in all honesty I'm not sure of the best approach. I have a feeling it'll be a combination of the two and maybe even two separate renders. The issue I have is the elevations are rough at best. But that's all I have to work with. That and the the photo i'll be taking once we get site access.
This is how I was going to go about it
Scene A Render A (The Arches)
Model the arch fronts and internals, model some dummy geometry for the arches themselves, move these into position as best i can using rough site plan and the photo as a guide in the viewport background
Model the paving and a basic version of the central stairs based on the rough site plan and the photo as a guide, create dummy geometry for the surrounding walls around the archways, populate the outdoor space - make all of this invisible to camera but cast shadows checked and visible to reflections/refractions checked.
Render.
Take this render and skew/move each archway into the correct position in the photo, they should be roughly correct to start with but won't be an exact match.
Scene B Render B (The Courtyard)
Make everything in the previous scene visible to camera again.
Delete the archways and internals
Cap the holes of the archways so i have a solid piece of dummy geometry represting the walls with the archways.
Apply shadow catcher material to dummy walls, dummy paving, dummy stairs. All other entourage (table/chairs/people/food carts/planters) to be rendered as normal with normal materials.
Render
Drop this render at the top of the photoshop for file for Scene A and mask back in the arches
Does that make sense or is that massively overcomplicating things? Would Camera Map per Pixel be a more useful approach?
