Add people to an
architectural render.
AI populate is a class of generative image-to-image transformation that places realistic figures into a static architectural render so the building reads as inhabited — scale, life, narrative — without staging a photo shoot or licensing a stock library. In Renovato this lives as preset E.01 (Add people), with sliders for density, posture distribution, and demographic range.
Why an empty render reads as cold
Architectural visualizations almost always omit people from the model — figures take time to model, license, and pose, and they distract from the architecture during design review. The cost shows up later: an empty render reads as a stage set, not a building. Clients struggle to imagine themselves in it. Reviewers can't calibrate scale.
Adding people late in the process is a known trick. The traditional fix is a Photoshop overlay from a stock library — expensive, time-consuming, and the lighting almost never matches. AI populate solves both: figures are generated to match the scene's lighting and perspective, with density and demographic distribution set by sliders.
How E.01 Add people works in Renovato
- Drop the populated source render into the atlas as a source node — typically a finished hero exterior or interior frame.
- Drag the E.01 Add people preset onto the canvas and connect the source.
- In the inspector, dial density (1-12 figures), posture mix (standing / walking / seated), and group size (singles vs. pairs vs. small groups).
- Click Generate. Cost: 3 credits. Wall-clock: 12-18 seconds. Output appears as a downstream node ready for any further preset.
The figures inherit the source render's lighting state, perspective, and depth-of-field. A fully relit-to- dusk source produces dusk-lit figures with warm pendant glow on faces; a midday source produces hard-shadow figures with bright fill.
Density choices, by frame type
- Hero exterior, 1-2 figures — a couple approaching the entrance, a single resident on the terrace. Scale cue without distraction.
- Public-space interior, 4-6 figures— café, lobby, library. Reads as "in use" without feeling staged.
- Plaza / street frontage, 8-12 figures — busy retail, transit hub. Convey throughput and time of day.
What populate can't do
E.01 generates plausible architectural-fitness figures — it doesn't produce specific named people, branded uniforms, or culturally-specific demographic distributions. If your project requires figures wearing a hospital's specific scrubs or a school's specific uniform, treat the AI output as a base layer and finish with traditional staging.
For animation, the static populate output can be fed into V.01 Walkthrough, but the resulting clip animates the camera, not the figures. People-in-motion video requires a different preset (V.04 Pedestrian flow, on the roadmap).
