A³ offers the latest and state-of-the-art AI algorithms for 3D generation. The generated 3D models can be downloaded in multiple formats, allowing creators to integrate them into their traditional DCC workflow. Please check out the guidance below to make the most out of this powerful tool.
Text to 3D
You can create 3D models with simple text prompts and choose how many models you want to generate at a time with the same prompt. The average generation time for a single model is about 10-20 seconds.
Refine the preview model
The models generated initially are preview models. You can further refine the model by clicking the 'Refine' button. The refining process takes around 3 to 5 minutes and will enhance the generated texture.
Image to 3D
You can generate 3D models using an input image. Similar to Text to 3D, you can choose how many iterations you want to generate. However, please note that for Image to 3D, the differences between each variation might not be very noticeable. You can find guidance and tips for your input image below.
Tips for uploading the most suitable images for a better 3D result
Get the best chance for the best outputs with choosing the best input images. Any images are not bad, but if you want the best outputs, here are some tips to help!
For Generating Characters
Tips ✅
Clear Separation of Limbs: Ensure arms and legs are not touching the body to improve the ease of modeling and animation.
Standard Poses: Use A-pose or T-pose with the character facing forward for easier rigging. Other poses are acceptable if rigging is not planned.
Avoid Obstructions: Do not include clothing or accessories that hide any part of the character’s body, ensuring all details are visible.
Close-Up Portraits: Artefacts supports close ups generation for higher resolution of facial details.
Avoid❌
Awkward Poses: Avoid using unconventional poses that complicate rigging—unless rigging is not part of your plan.
Obstructed Views: Ensure that clothing or accessories do not hide any parts of the character's body.
Distracting Backgrounds: Use backgrounds that contrast well with the character to prevent blending and to keep the focus on the character itself.
For Generating Game Props
Tips ✅
Front-Facing: Choose either a front or side view to ensure clarity and ease of interpretation.
Background: Utilize clear or solid-colored backgrounds to prevent distractions and enhance focus on the prop.
Isolation: Feature the prop alone, without the inclusion of other items, to maintain the emphasis on the subject.
Avoid❌
Odd Angles: Don't use extreme angles for furniture or props as they can distort the view.
Crowding: Don't crowd the image with multiple props; one prop per image is best.
Busy Backgrounds: Don't use backgrounds that are cluttered or have a lot of detail.
Color Blending: Don't let the prop blend into a similar-colored background; it should stand out clearly.
For Generating Other Stuff In General
Avoid❌
Proper Scaling: Choose an image size that showcases the subject clearly without requiring zooming in, which can lead to a loss of detail.
Simple Contexts: Avoid complex CAD maps or vast landscapes which may overwhelm the focus of the image.
Complete Visibility: Ensure the main object is fully visible in the image, without any part being cropped out.
High Resolution: Ensure the image is not pixelated; it should have a high enough resolution to be clear and detailed.