SD Tutorial:How to make a Logo with Stable Diffusion?

Omniinfer
3 min readAug 16, 2023

--

Today in this post, We’ll build a creative logo with SD hand-over-hand.First we’ll start with a comparison image, then we’ll show you step by step how to generate it.

Operation steps

  1. First, type in the text or brand logo you want in your usual design tool, and design it as black on white
  2. Next, open the text to image interface and enter the Positive prompt and Negative prompt
  • Positive prompt

masterpiece, ((clean white silver light metal)), ((Cyberpunk style)), best quality. ((at night, in the futuristic city center1.3, concept art, sci-fi scenes, planet)), (super details, aesthetics, beautiful composition, rich bright colors, voluminous soft light). 4k, unreal Engine, octane render, cuteness render, 3d rendering, Adobe Photoshop, masterpiece, ((clean glass, shine:1.2. (Transparent:1.1), gloss)), best quality. (Pixar — style :1.4), (in a cave, at night, surrounded by glowing plants, flowers, flying fireflies, bonfires), (Ultra detailed, aesthetic, beautiful composition, rich bright colors, volumetric soft light). 4k, reflection in the water, water ripple Inspired by Alice in Wonderland, magic, fairy tales. unreal Engine, octane render, cuteness render, 3d rendering, Adobe Photoshop, awe inspiring, beautiful, <lora: XSarchitectural-38InteriorForBedroom_36660:1>

  • Negative prompt

EasyNegative, ng_deepnegative_v1_75t, low quality, worst quality:1.2,BadDream, Unrealistic

  • Other information

Clip skip: 2

Sampling method: DPM++ 2M SDE Karras

Sampling Steps: 33

Size: 512x512

Batch size:3

CFG scale: 14

Checkpoint: revAnimated_v122

Lora: “XSarchitectural-38InteriorForBedroom_36660

3. Open the ControlNet plugin, upload a black logo image on a white background

4. click Enable, select the Control Type All, select the preprocessor invert(from white bg & black line), the model control_v11p_sd15. Other parameters can be kept by default

5. Finally, click Generate to generate pictures of different scenes

All the images above were generated using Free NO GPU WebUI

--

--

Omniinfer

Fastest generation in just 2s, Pay-As-You-Go, $0.0015 for each standard image, and 15% off for high-volume usage, you can add your own models