Skip to content
Module
Module 3 of 6Lesson 2 of 2~15 min

Renders and post-production with AI

Rendering engines now include AI treatments, and overzealous retouching can change what the image commits to with the client. This lesson helps you choose an engine according to the deliverable and the license, and use AI post-production without distorting the project.

Lesson objective

By the end of this lesson, you will be able to choose a rendering engine and its AI treatments according to the deliverable and the license, and apply AI post-production without altering what the image commits to.

Topics covered

  • AI architectural rendering
  • AI post-production
  • rendering engines
  • architecture images

Where it fits

Produce variations and renders

How do you quickly produce variations and realistic images without misrepresenting the 3D model?

Lessons in this module

  1. Generate variations from the 3D model
  2. Renders and post-production with AI (this lesson)

What you will learn in the course

This lesson is part of the course Bring AI into architectural visualization

  • Identify, for a given project, the steps where AI speeds up visual production or analysis, and those where the 3D model and the architect's judgment remain the reference, including for regulatory compliance.
  • Generate façade or mood variations from a 3D view with a built-in tool or an image model, while controlling fidelity to the project (structured prompt, influence, masks, references), then choose a rendering engine and AI post-production according to the deliverable and the license.
  • Run a workflow of mesh export, AI material creation and re-import into a rendering engine, and diagnose a faulty PBR material.
  • Assess the risks of an AI-assisted visual (usage rights, copyright protection, confidentiality of 3D models, transparency towards the client and the public) and write the corresponding usage rules.
  • Select a tool chain suited to a deliverable (competition, design stages, marketing, renovation, planning application) and justify it by fidelity, lead time, quality, license cost and risk.