LCM for Krita - OVERPOWERED!!!!

Discover the power of combining LCM with CR in digital painting, leveraging AI to turn anyone into a master painter without the need for traditional skills. The video showcases innovative workflows for generating avatars with varying expressions, using high-resolution photos with glitch effects, and creating anime-style pets. It guides viewers through installing and using CR and AI image diffusion, demonstrating how to generate images and refine them with personal touches, resulting in stunning, unique creations.

Takeaways

• 🎨 The video introduces a powerful combination of digital painting with AI in Krita using LCM and CR.
• 🤖 Three innovative AI workflows are showcased, including an avatar generator, full-resolution photo manipulation, and a pet creator.
• 🔧 CR (Creta) is required and can be downloaded for free as open-source software for Windows, Linux, and Mac.
• 🔄 The AI image diffusion plugin needs to be installed in the py cre folder and enabled within Krita's settings.
• 🖼️ Multiple control net extensions and models like realistic Vision, dream shaper, and chuga not XL can be installed for enhanced image generation.
• 🎨 Users can create their own styles in Krita by selecting style words and negative prompts, and save them for future use.
• 🖌️ The AI image generation tool in Krita allows for layer usage, where all visible layers contribute to the final image.
• 👩‍🎨 Even with basic drawing skills, users can create detailed images by combining their drawings with AI enhancements.
• 🛠️ The strength of the AI assistance can be adjusted using the D noise slider, allowing for control over the level of detail.
• 📸 The current AI-generated result can be copied as a layer for further editing or used as a new input for continuous creation.
• 😄 An upcoming live stream is teased where the creator will demonstrate more of these AI-powered painting techniques.

Q & A

What is the main theme of the video?

The main theme of the video is the demonstration of how to use AI, specifically LCM and CR, to enhance digital painting and create various art styles with minimal effort.

What are the three workflows showcased in the video?

The three workflows are an avatar generator, a full-size photo with a glitch effect, and a cavipet creator for generating pets in different styles.

What is the role of face detailer in the avatar generator workflow?

The face detailer is used to change the emotion of the face while maintaining character consistency in the avatar generator workflow.

How can randomization be utilized in the avatar generator?

Randomization is used to create an endless amount of automated prompts, resulting in different characters with the same facial expressions each time the CU button is clicked.

What software is required to follow the tutorial?

To follow the tutorial, one needs CR (an open-source software available for Windows, Linux, and Mac) and the AI image diffusion plugin.

What are the steps to install and set up the AI image diffusion plugin?

First, download the plugin and unpack it into the py Cre folder. Then, open Cre, go to settings, configure CR, enable the AI image diffusion plugin, and restart CR.

What types of GPUs are supported by the AI image diffusion plugin?

The plugin supports Nvidia, AMD, Apple M1 and M2 CPUs, and Cloud GPUs.

How can the LCM (Latent Control Model) be utilized in the AI image generation process?

LCM can be selected from the brush icon in the AI image generation section, where it can be used to enhance the generated image based on the strength setting and the layers visible in the document.

What is the purpose of the 'Copy Current Result' button?

The 'Copy Current Result' button allows users to save the current generated image as a new layer, which can be used for further modifications or as a backup.

How does the strength slider affect the generated image?

The strength slider, which controls the D noise, determines how much influence the user's drawing has on the generated image. A lower value means more influence, while a higher value makes the AI-generated aspects more prominent.

What additional features does the video mention for fine-tuning the AI-generated image?

The video mentions the ability to use multiple layers, adjust the prompt and negative prompt in the settings, and select different checkpoints for more control over the final image.

Outlines

00:00

🎨 Introducing AI-Enhanced Digital Painting

The paragraph introduces the concept of leveraging AI to enhance digital painting skills. It explains that viewers don't need any prior painting experience as the power of AI will aid in creating art. The speaker showcases three workflows that integrate AI for interesting applications, such as an avatar generator that produces consistent character expressions, a glitch effect on high-resolution images, and a pet creator that adapts the input image's style to generate various pets. The paragraph emphasizes the innovative use of AI in art and provides a link to the speaker's Patreon for more workflows.

05:01

💻 Setting Up CR and AI Diffusion

This paragraph details the process of setting up the necessary software for AI-enhanced digital painting. It instructs the viewers to download and install CR, which is open-source and available for Windows, Linux, and Mac. The speaker also explains the installation of the AI image diffusion plugin, emphasizing the need for a version above 5.2. The paragraph further guides the audience through the configuration of CR settings, including the Python plugin manager and the AI image generation options. It advises on the selection of different control net extensions and the importance of restarting CR after installations and updates.

10:02

🖌️ Exploring AI Image Generation and Editing

The speaker delves into the features of AI image generation within the software, discussing the various styles and models available for selection. It highlights the use of style words and negative prompts to refine the AI's output. The paragraph also covers the use of the brush icon for selecting between different generation modes, such as appscale and animation life. The speaker demonstrates the use of the strength slider, which adjusts the influence of the user's drawing on the AI's output. It showcases the ability to use multiple layers and the ease of color selection and tool customization. The paragraph concludes with a live demonstration of creating an image using the AI, emphasizing the ease of use and the impressive results that can be achieved even with minimal input.

🌟 Finalizing and Customizing AI Art

In the final paragraph, the speaker discusses additional features for customizing and finalizing the AI-generated art. It explains how to copy the current result to a new layer for further editing or backup. The paragraph also covers the use of the freehand selection tool for targeted adjustments, such as changing the clothing color in the generated image. The speaker then adds a smile to the portrait and demonstrates how the AI can incorporate this change. The paragraph ends with an invitation to join a live stream for more interactive learning and a call to action for viewers to like and follow for more content.

Keywords

LCM

LCM, or Layered Composition Magic, is a feature that works in conjunction with CR (Creta), an open-source digital painting software. It leverages the power of AI to enhance and transform digital artwork by generating detailed layers based on the user's input. In the context of the video, LCM is used to create stunning visual effects and intricate details that would be difficult to achieve manually, showcasing the potential of AI in the realm of digital art and animation.

CR (Creta)

Creta (CR) is an open-source digital painting software that provides artists with a platform to create and edit images. It is mentioned in the video as a crucial component in the process of combining digital painting with AI capabilities. The software supports various plugins, including AI image diffusion, which allows for the integration of AI-generated content. In the tutorial, the creator guides viewers on how to download and install CR, set up the necessary plugins, and utilize its features to enhance their artwork with AI assistance.

AI

Artificial Intelligence (AI) is a broad concept that refers to the simulation of human intelligence in machines that are programmed to think and learn like humans. In the context of the video, AI is used to power the digital painting workflows, such as the avatar generator and the image-to-image style transfer. It allows for the creation of unique characters with consistent expressions and styles, as well as the rendering of high-resolution images with special effects like glitch art. The video demonstrates how AI can be harnessed to streamline and enhance the digital painting process, offering artists new possibilities and creative freedoms.

Avatar Generator

An Avatar Generator is a tool that uses AI to create digital representations of characters or personas, known as avatars. In the video, the avatar generator is showcased as a means to produce avatars with the same details but different facial expressions, maintaining character consistency. This tool is particularly useful for artists looking to create a variety of characters with ease, as it automates the process of generating diverse expressions while keeping the core features of the character intact.

Randomization

Randomization is the process of generating a variety of outcomes from a set of possible elements without any predictable pattern. In the context of the video, randomization is used to create an endless amount of automated prompts for the avatar generator. This allows the artist to produce a wide range of characters with different expressions and styles by simply clicking a button, streamlining the creative process and introducing an element of surprise and novelty to the artwork.

Glitch Effect

A glitch effect refers to a visual distortion or error that is intentionally applied to an image or video to create an aesthetic of digital malfunction or corruption. In the video, the glitch effect is used to enhance a full-size, high-resolution photo by rendering it over the image with the glitch, combining the two to produce a unique and visually striking piece of art. This technique demonstrates how AI can be used to apply complex effects that would be difficult to achieve manually, pushing the boundaries of digital art and offering artists new ways to experiment with their creations.

Image to Image

Image to image is a concept where an input image is used to guide the style or content of a newly generated image. In the video, this concept is demonstrated through the use of an AI tool that takes an input image and applies its style to create a new image, such as a pet creator that generates pets in the style of the input image. This technique allows artists to maintain a consistent style across their work and explore new creative directions by blending the characteristics of different images.

Styles

In the context of the video, 'Styles' refer to the collection of words or phrases that define the desired visual characteristics of the AI-generated image. These styles guide the AI in creating artwork that matches the specified aesthetic or theme. For example, the creator can input 'portrait of a beautiful woman' as a style prompt, and the AI will generate an image based on this description. The video also mentions the use of negative prompts to further refine the output by specifying what elements should not be included in the final image.

D noise

D noise, or 'D' noise level, is a parameter in the AI image generation process that controls the influence of the artist's manual drawing on the final output. A higher D noise level means that the AI's contribution is more prominent, while a lower level allows the artist's drawing to have a greater impact on the result. In the video, the creator adjusts the D noise level to find a balance between their own artistic input and the AI's capabilities, demonstrating how this setting can be fine-tuned to achieve the desired level of detail and control over the artwork.

Layers

Layers in digital painting refer to separate, transparent sheets that can be叠加 to create a composite image. In the video, layers are used to build up the artwork, with each layer contributing to the final composition. The AI image generation process takes into account all visible layers, not just the active one, allowing for complex and detailed creations. The video also mentions the ability to use multiple layers without any issues, highlighting the flexibility and depth of the digital painting process when combined with AI.

Prompt

A prompt in the context of AI-generated art is a description or a set of keywords that guide the AI in creating an image. In the video, the creator uses prompts such as 'portrait of a beautiful woman' to instruct the AI on what to generate. The prompt is a crucial element as it directly influences the output of the AI, determining the subject matter, style, and mood of the resulting artwork. The video demonstrates how prompts can be combined with the artist's manual input to create a unique blend of human creativity and AI-powered image generation.

Highlights

LCM combined with CR is massively powerful, allowing users to become master painters without prior painting skills by leveraging the power of AI.

The presenter has created three amazing workflows that showcase interesting concepts of using AI in digital painting.

An avatar generator is demonstrated, producing avatars with the same details but different facial expressions while maintaining character consistency.

Face detailer is used to change the emotion of the face in the avatar generator, adding a layer of customization.

Randomization is employed to create endless automated prompts, resulting in completely different characters with the same expressions every time the CU button is clicked.

A full-size, high-resolution photo can be used in conjunction with a glitch effect, pushing the boundaries of what was previously possible in stable diffusion.

The third workflow showcases a pet creator that takes any kind of pet input and generates a similar pet in an anime style without using IPA.

The pet creator can produce a variety of animals like tigers, unicorns, bears, and cats, demonstrating the flexibility of the workflow.

Creta (CR) is required for the tutorial, and it is available for Windows, Linux, and Mac for free as open-source software.

Instructions are provided for installing the AI image diffusion plugin into the Creta software, including the necessary steps for different operating systems.

The AI image generation interface within Creta is explored, including the selection of styles, prompts, and the use of GPU for rendering.

The strength slider, representing the D noise, allows users to adjust the influence of their drawings on the AI-generated image.

Multiple layers can be used in the AI image generation, with everything visible being rendered in the LCM or normal generation.

The copy current result to the image feature allows users to backup their work and use it as a new input for further customization.

The freehand selection tool enables users to isolate specific parts of the image for targeted adjustments.

The presenter demonstrates the ease of use and the powerful capabilities of the AI-assisted painting, transforming simple mouse drawings into detailed images.

An upcoming live stream is teased, promising more interactive fun and exploration of the AI painting capabilities.

The video concludes with a call to action for viewers to like, share, and join the live stream for more engaging content.