Directional Tiles

What It Is

Directional Prompt Mode lets you assign different prompts to different viewing directions instead of one prompt for the whole image. It uses Qwen2.5-VL, a local vision model, to drive geometry-based prompting across polyhedra face-center directions.

Each polyhedron face points to a specific direction on the sphere. You assign a prompt per face, and the AI generates content matching each direction -- different sky, ground, and everything between.

How It Works

The equirectangular image is divided into directional zones based on polyhedra geometry:

1. Pick a polyhedron -- 14 shapes available, from Cube (6 faces) to Goldberg GP(3,0) (92 faces). More faces = finer directional control, slower generation.
2. Assign prompts per face -- each face maps to a {yaw, pitch, prompt} point. Editable, deletable, freely reorderable.
3. Prompts blend across the sphere -- between face centers, prompts blend smoothly. No hard seams.

Available shapes and their face counts:

PolyhedraTileControls

The PolyhedraTileControls panel shows an interactive list of all face-direction tiles. Each tile card displays its direction (yaw/pitch), assigned prompt, and Describe status.

• Clickable face targets -- click any tile to select it and jump the viewer to that direction.
• AimOverlay -- a 3D overlay on the input preview that highlights which polyhedron face corresponds to your current view direction.
• PolyhedraPreview -- a rotatable 3D solid rendering showing all face center points. Hover or click faces to see their prompts.
• Editable point list -- every generated point is editable and deletable. Add or remove directions freely.

Qwen2.5-VL Describe

Click Describe on a tile to have Qwen2.5-VL analyze what is currently visible in that direction and suggest a prompt. No API key required -- everything runs locally on your GPU.

• How it works -- the app perspective-crops your equirectangular input at the face-center direction, runs Qwen2.5-VL on the crop, and writes a prompt describing the content.
• Batch describe -- use Describe All to run Qwen2.5-VL on every face in one batched session. All directional prompts populate at once.
• Describe cache -- results are content-addressed by SHA-256 hash of the input image and cached in an LRU cache of 256 entries. Describe the same image with the same polyhedron again and results return instantly -- no Qwen reload.
• Qwen2.5-VL is the only describe provider. Requires ~11 GB VRAM. Install it from the Pipelines page.

Cosine-Weighted Prompt Interpolation

Between tile centers, directional prompts blend using cosine-weighted interpolation. Each point on the sphere gets a weighted mix of nearby prompts based on angular distance, with influence falling off smoothly by cosine. No hard seams between directional zones.

When Semantic Blend is enabled (available on supporting pipelines), directional prompts blend in CLIP embedding space via spherical_mean for a spatially coherent baked latent field across the entire sphere. On pipelines without semantic blend support, it degrades automatically to weighted text concatenation.

Image Offset Interaction

Directional prompts stay anchored to absolute sphere directions. When you offset (pan/rotate) the image, the image moves under the fixed prompt grid:

• Your "sky" prompt always points up, regardless of where the image content sits.
• Your "ground" prompt always points down.
• Changing your view angle or panning the image never invalidates your directional prompt set.
• Regenerate prompts only when you load a different input image.

The image offset (yaw, pitch, roll) is applied live at render time. The prompt grid is world-space absolute; the image is camera-space relative.

Generate Flow

1. Pick a polyhedron -- select the shape that matches your desired directional granularity.
2. Place directional prompts -- type prompts per face, or leave some blank to inherit the global prompt.
3. Optional: Describe -- click Describe or Describe All to let Qwen2.5-VL analyze each direction and suggest prompts.
4. Click Generate -- the AI fills each direction according to its assigned prompt, blended smoothly across the sphere.
5. Review -- check the result in the viewer. Edit prompts and regenerate specific directions as needed.

Tile Ordering Strategies

14 strategies control the order in which tiles are generated. The order affects how adjacent tiles influence each other during processing:

Works with all diffusion pipelines (SDXL, SD3.5, SD2.1, SD1.5, FLUX, Z-Image Turbo, Qwen-Image).

Navigator FOV & neighbour influence

When placing or inspecting directional points in the viewer/navigator:

• FOV footprint shows which part of the sphere a direction “owns” for blending — widen FOV to soften borders; narrow to keep a direction local.
• Neighbour influence (cosine-weighted blend) mixes adjacent direction prompts between centers so you do not get hard walls.
• Place-and-describe workflows: drop a direction, write its prompt, then let auto-fill or AI transform expand the set — product language is Directional.

Upscaling vocabulary (when chaining after directional work)

Do not treat “upscale” as one button:

See Upscaling.

Next Steps

• Prompt System -- Interpolation modes and prompt tools
• Interface Overview -- Where prompt tools live in the app shell
• Templates -- Choose the right template
• Pipelines -- Understand different AI models