Let’s say you generate the following image of men standing on the street. Obviously, you need to fix faces. We have discussed multiple approaches to repair faces. ADetailer provides another way to fix faces in an automatic way. If you turn on ADetailer, then face repair is automatically done when an image is generated. Below …
How to use ADetailer to auto-fix damaged face & hands in Automatic1111 Read More »
In our previous tutorial How to use Ultimate SD Upscale extension in Automatic1111, we saw a problem with Stable Diffusion generating a face for each tile when we use Ultimate SD Upscale with Image to Image denoise strength is set to 1. In case you haven’t read the tutorial, here is a brief recap. Let’s …
How to use Ultimate SD Upscale extension with ControlNet Tile in Automatic1111 Read More »
Overview The Ultimate SD Upscale extension is more than just a simple upscaler; it’s akin to a high-resolution fix. This tool first upscales an input image to the desired dimensions. Following the initial upscale, it uses the Img2Img process to enhance image quality. The key distinction between the Hires fix and the Ultimate SD Upscale …
How to use Ultimate SD Upscale extension in Automatic1111 Read More »
Overview Extensions are a game-changing feature for Automatic1111, offering the capability to introduce functionalities not originally included in the main program. Imagine equipping a car with wings and a jet engine to transform it into a flying vehicle – that’s the level of enhancement we’re talking about. Installing an extension opens up a new realm …
Problem to solve – Overview of Hires. fix To understand the need for a high-resolution fix, often referred to as ‘Hires. fix,’ see the photo below where the faces resemble characters from a horror movie, while the rest of the image appears normal. At the default 512×512 resolution, Stable Diffusion can struggle with rendering detailed …
Textual Inversion offers a way to modify Stable Diffusion models with custom data in Automatic1111. You can adjust image generation by specifying the embedding file name in either the positive or negative prompt, or both. For an in-depth understanding of Textual Inversion, please refer to our comprehensive guide How to use Textual Inversion with ComfyUI. …
How to use Textual Inversion or load custom embedding in Automatic1111 Read More »
Welcome to the third installment of our series on using ControlNet in Automatic1111. This tutorial builds upon the concepts introduced in How to use ControlNet in Automatic1111 Part 1: Overview and How to use ControlNet in Automatic1111 Part 2: Installation. If you haven’t read them yet, we recommend doing so. In this segment, we’ll cover …
How to use ControlNet in Automatic1111 Part 3: Lineart & OpenPose Read More »
Welcome to the second installment of our series on using ControlNet in Automatic1111. This tutorial builds upon the concepts introduced in How to use ControlNet in Automatic1111 Part 1: Overview. If you haven’t read it yet, we recommend doing so. In this segment, we’ll continue from where we left off, focusing on installation. Regrettably, ControlNet …
How to use ControlNet in Automatic1111 Part 2: Installation Read More »
What is ControlNet and why is it needed? Welcome to the first in our series of tutorials on ControlNet. This tutorial aims to explain what ControlNet is and its necessity. Understanding its core concept is crucial for both installation and effective usage. ControlNet is a software add-on to provide visual cues to Stable Diffusion, guiding …
How to use ControlNet in Automatic1111 Part 1: Overview Read More »
What is LoRA? LoRA, is one of the popular ways to customize Stable Diffusion models. A LoRA model is a supplementary model that works with a regular Stable Diffusion model to modify the model’s behavior with specific custom data so that the combination of the full model and the LoRA can produce images that mirror …