Home › Prompts › Stable Diffusion › Keep a layout with SD 3.5 Depth ControlNet

Keep a layout with SD 3.5 Depth ControlNet

Generate a new scene that follows the depth and composition of a reference photo, for consistent layouts and re-themes.

Stable Diffusion Image generationRuleStack original

Written for Stable Diffusion 3.5 Large + ControlNet Depth · Updated 2026-10-17

Note: Review and test before use. Provided as is, without warranty. See Terms.

Prompt

[New scene description that matches the reference layout, e.g. a cozy reading nook with a green velvet armchair and tall bookshelves], [style], [lighting], [palette].

Replace everything in [square brackets] with your own details.

Settings & parameters

ComfyUI or diffusers with the SD 3.5 Large ControlNet Depth model; depth map from your reference photo
Control strength ~0.6-0.8; lower it for more creative freedom

Variations

Canny ControlNet to follow outlines instead of depth.
Blur ControlNet to upscale while keeping structure.

Tips

Use reference photos you took or have rights to.

More: all Stable Diffusion prompts · all image generation prompts · prompt library

Related prompts

Fast drafts with SD 3.5 Large Turbo

Quick low-step drafts to test composition and wording before a full-quality SD 3.5 Large render.

Stable DiffusionImage generation

Fast concept batch on SD 3.5 Medium

Generate many quick concept thumbnails on a consumer GPU to explore ideas before final renders.

Stable DiffusionImage generation

SD 3.5 Large photoreal with negative prompt

A local or API photoreal render using SD 3.5's natural-language prompts, negative prompt and sampler settings.

Stable DiffusionImage generation

Brand style lock with --sref and --p

Generate on-brand images for a campaign by combining a style reference, style weight and your personalization profile.

MidjourneyImage generation

Cinematic film still with motivated lighting

A frame that looks pulled from a feature film, using lens, film stock and light-motivation language that V8.2 follows closely.

MidjourneyImage generation

Editorial product hero shot

A magazine-grade product hero image with controlled light and a clean set, tuned for V8.2's stronger prompt reading and HD mode.

MidjourneyImage generation

Get the best new AI prompts every week

One short email a week: the best new prompts and rules for ChatGPT, Claude, Gemini, Midjourney, Cursor and more, plus what changed in the tools. Free, unsubscribe anytime.

Finishes on our newsletter page at beehiiv (opens in a new tab). Or subscribe directly.