How to Write a Great Prompt for Carveco’s Text to Relief Feature
Creating a great 3D relief model with Carveco's Text to Relief tool starts with one essential ingredient: a well-crafted prompt. Whether you’re a beginner just exploring or a professional maker aiming for precise results, understanding how to guide the AI with clear, specific keywords can make the difference between an average model and a great one.
This article explains why using keywords matters, how different keywords shape your results, and offers categorized lists and examples to help you write better prompts.
Why Keywords Matter
Keywords are the instructions that help Carveco's AI understand what you want. Without them, you might get vague or random results. With the right keywords, you can:
- ✅ Shape the style of the model
- ✅ Control the level of detail
- ✅ Define the background and scene
- ✅ Achieve results that match your project needs
Keywords
Here are the main keyword categories you can mix and match to craft powerful prompts.
1. Model Style Keywords
These tell the AI what kind of model or visual format you want:
- Digital 3D model
- 3D model
- Texture
- Pattern
- CGI
- Photorealistic
- Flat cartoon
- Illustration
- Painting
- Bas relief
- Relief
- Carving / Carved
- Ornate
- Graphic
A digital 3D relief model of a woven rattan pattern, showing crisp detail and depth separation.
2. Enhancement Keywords
These adjust the level of realism, finish, and quality:
- Sharp detail
- Highly detailed
- Crisp detail
- Hyper realistic
- 4K
- Vibrant
- Greyscale
- Depth separation
- Smooth shading
- Close-up
- Facing forward
- Side on / Profile
✅ Example:
A hyper realistic carving relief of a lion's head facing forward, with sharp detail, in 4K, on a flat background.
3. Background & Scene Keywords
These control what goes around the subject:
- Isolated
- Plain background
- Flat background
- Blurred background
- Full scene
A photorealistic model of a knight's helmet isolated on a plain background.
Tip: If you describe a busy environment (like "pirate's cave full of treasure"), the AI will create a scene with background elements. If you say "on a plain background" or "isolated," the AI will focus only on the main subject, producing a single stand-alone model without extra surroundings. This approach keeps the design clean and free of background clutter, making it ideal when you want a cleanly cut, focused relief model.
How to Achieve Different Looks
🔶 Full, Rounded Models
Keyword examples: digital model, 3D model, CGI
🔶 Decorative Models
Keyword examples: Bas Relief, carved, ornate
🔶 Bold Graphic Models
Keyword examples: flat cartoon, illustration, simple block colors
🔶 Photo Realism
Keyword examples: photorealistic, painting
🔶 Textures
Keyword examples: texture, pattern, close-up
🔶 Scene vs. Isolated Object
To get an isolated object: include keywords like isolated, on a plain background, flat background.
To get a scene: describe the environment and let the AI fill in surroundings (e.g. “inside a medieval castle”).
Final Tips
- ✅ Be specific. The more specific you are about the subject, style, and background, the better the AI can deliver.
- ✅ English Language. Using English will produce the best generated images.
- ✅ Mix categories thoughtfully. For example, mixing "flat cartoon" with "highly detailed 3D model" can confuse the output.
- ✅ Note on faces. Faces tend to generate best when they are facing directly forward or directly side-on. For the most accurate and clean face reliefs, we recommend specifying the face to be facing forward, or side-on.
- ✅ Test and adjust. Try variations: run the same subject with different keywords (e.g., photorealistic vs. flat cartoon) to compare.
Comments
Please sign in to leave a comment.