I wanted to share a way to create a fully customized 3D text render like above by using the following tools:
- Adobe Photoshop – Photoshop is used to create an Adobe Illustrator path.
- Adobe Illustrator – Illustrator is used to convert an .ai file to .svg.
- Blender – Blender is used to convert the path (curve) to mesh, and to extrude.
- Stable Diffusion – Stable Diffusion is used to produce the final scene.
If you don’t have access to Adobe Photoshop, I would recommend you to use Photopea instead, to directly save the image as .svg. If you don’t have access to Adobe Illustrator, please google ‘free .ai to .svg converter’ to find and use a free website that provides the file conversion service.
This is a great way to create the best looking text objects for various projects.
Create a new text object and create a letter “T” text object. [Control + J] to duplicate the layer 3 times. Either use [V] or [Control + T] to separate the text objects.
Change the letters to “TEST”.
Use the [Control + T] shortcut to move and rotate the letters like this. [Right click] each layer to rasterize them. [Shift + Select] all layers and right click to merge all visible layers.
[Control + Click] the layer to make a selection, and go to ‘Path’ to create a new work path from the selection (there’s a button at the bottom to create a new work path). When that’s done, go to [File -> Export -> Path to Illustrator]. Open the saved file in Illustrator, and go to [File -> Save As] to save the file as SVG.
Import the SVG file in Blender. If you don’t see that menu item, please activate an add-on called ‘SVG 1.1. format’.
The idea is to [Right Click] each item to convert to mesh, select all and join them together by using [Control + J], and go to Edit Mode [Tab] to fill [F].
After that, go to Edit Mode [Tab], type [A] to select all, and [E] to extrude like this. Render and save the image.
Import the rendered image as ControlNet’s Depth image (Please refer to this video for Stable Diffusion’s depth setup procedure) and click [Generate] to create a new stable diffusion image like this.
This image was created by Stable Diffusion with the following prompt:
Prompt – 3D text says “test”, isolated in black background, glossy golden metal
Negative prompts – (worst quality, low quality, normal quality:1.8)
This is the end of the tutorial. 🙂