Krita AI diffusion, A look at v1.15 update, bugs on older hardware and new cool features

Streamtabulous
9 Mar 202428:18

TLDRThe video discusses the AI diffusion 1.15 update by Supicious Firefly for Krita, a digital painting program. The update introduces new features that enhance the program's capabilities, making it more powerful and comparable to Adobe Photoshop's generative fill. The presenter highlights the ability to edit artwork more effectively, control glitches, and use new core components. They also mention the addition of a discard feature for rendered images, the option to copy prompts, and the potential for future improvements like text file export and import capabilities. The video covers the installation process, new samplers, and customization options. However, the presenter encounters bugs on older hardware, such as the GTX 1070, and shares workarounds for these issues. They also discuss the community's response to the update and the potential for further enhancements. The summary emphasizes the update's benefits for photo restoration and digital artwork, despite the challenges faced with older hardware.

Takeaways

  • 🎨 The AI diffusion 1.15 update by akley for Cryer introduces new features that enhance the editing capabilities of artwork, making it more powerful and on par with Adobe's Photoshop for generative fill.
  • 🔍 Users can now discard rendered elements they don't like by right-clicking and selecting 'discard', providing more control over the final output.
  • 📋 Copying prompts has been made easier, though the user suggests a feature to save prompts as text files for future use.
  • 🔄 The update includes the ability to generate custom content within selected areas of an image, allowing for more precise and controlled generative edits.
  • 🚀 Cryer tools and Auto selection with outlining are highlighted as very handy features for users who don't have them, improving workflow efficiency.
  • 🔧 The user experiences bugs when using the AI diffusion on older hardware like the GTX 1070, suggesting that updates may not fully support older graphics cards.
  • 🛠️ A workaround for the 'expand' feature bug involves adding a dot to the image before using the feature, which tricks the AI into processing the image correctly.
  • 📈 The user appreciates the increased power and usability of the AI diffusion tool but acknowledges the need for bug fixes, especially for older hardware.
  • 📸 For photo restoration and digital art, the update provides new possibilities and makes certain tasks easier, such as removing content or enhancing image areas.
  • ⚙️ The user suggests that the developers might not have access to the same hardware for testing, which could explain why some bugs persist in older graphics cards.
  • 👍 The video concludes with a recommendation to like, subscribe, and share for community support and to help the YouTube algorithm promote the content to a wider audience.

Q & A

  • What is the main focus of the 1.15 update by akley for Cryer?

    -The 1.15 update by akley for Cryer focuses on introducing new features that make the AI diffusion more powerful and controllable, allowing users to edit their artwork in ways that were not previously possible.

  • How does the new update compare to Adobe's Photoshop in terms of generative fill?

    -The new update is on par with Adobe's Photoshop in terms of generative fill, but with the added advantage of more control to prevent certain glitching issues.

  • What is the process for updating AI diffusion to the latest version?

    -To update AI diffusion, users need to download the zip file from the GitHub page, extract it, and overwrite the existing files. Then, upon opening Cryer, the AI configuration will prompt for an update, which should be confirmed to include all new features.

  • What is the significance of the lightning model introduced by Civ AI?

    -The lightning model introduced by Civ AI is designed for generating fantasy creatures but also works well for humans. It allows users to see different types of renders and share prompt information for testing and comparison.

  • How does the new update improve the user's ability to edit specific parts of an image?

    -The new update allows users to discard rendered elements they don't like, copy prompts directly into the workspace, and use the entire image for generating custom content, providing better control and integration with the original artwork.

  • What are some of the new core components and features introduced in the 1.15 update?

    -The 1.15 update introduces new core components such as the ability to auto-select with outlining, various samplers like DPM Plus+, SD, and sgm, and options to adjust the diffusion and paint outside of selections for better blending and customization.

  • What is the workaround for the bug encountered when trying to use the face tool or expand feature?

    -The workaround involves adding a single dot to the image before using the face tool or expand feature, which provides the AI with enough information to generate the rest of the image without errors.

  • What issues are users with older hardware like the GTX 1070 facing with the 1.15 update?

    -Users with older hardware are experiencing bugs such as runtime errors related to data type mismatches and issues with the face tool and expand feature, which may prevent them from fully utilizing the new update's features.

  • How can users provide feedback on bugs and issues they encounter?

    -Users can provide feedback by reaching out to the developers, such as suspicious Firefly on Reddit, and sharing their experiences and error reports to help identify and fix the issues in future updates.

  • What are some of the benefits of the new features for photo restoration and digital painting?

    -The new features allow for more precise control over image editing, making photo restoration easier by enabling the removal of unwanted content and improving the blending of generated elements with the original image. They also help with digital painting by providing better control over brush strokes and selections.

Outlines

00:00

🚀 Introduction to AI Diffusion 1.15 Update

The video begins with an introduction to the AI diffusion 1.15 update by akley for Cryer, highlighting the new features that make it more powerful and versatile than ever before. The update is compared to Adobe's Photoshop in terms of generative fill, but with the added benefit of user control to prevent glitching. The video promises a deeper dive into these features after the intro and mentions a link to a tutorial on how to update the software. It also gives a shoutout to Civ AI for their furnace model, specifically the lightning model that works well with both fantasy creatures and humans.

05:01

🛠️ Exploring and Updating AI Diffusion Tools

The video continues with a walkthrough of the GitHub page for akley, showcasing the AI diffusion and Cryer tools. It emphasizes the usefulness of these tools for those who don't already have them. The AI diffusion is selected to demonstrate the update process, which involves downloading a zip file and extracting it to overwrite the existing files. The video also discusses the addition of new core components and the recommendation to install all available features. A note is made about a missing model, which the presenter already possesses.

10:02

🎨 Enhancing Artwork with New Features

This paragraph delves into the new features of the AI diffusion 1.15 update, such as the ability to discard rendered images and copy prompts. The presenter expresses a desire for a feature to save prompts as text files and to import prompts from external files. The implementations made are praised as fantastic, and a question from a previous video about the lightning model is addressed. The paragraph also discusses the absence of certain samplers in the current version and the hope for their inclusion in future updates.

15:03

🖌️ Utilizing the Pen Tool and Custom Generation

The video demonstrates the use of the pen tool for selecting areas in an image and the custom generation feature. The presenter selects an area of an image and copies the prompt, ready to generate a new version. The importance of selecting the entire image is discussed to ensure the AI can generate content that fits seamlessly with the existing elements. The video also explores the options for generating custom content, such as selection borders and masks, and the presenter shares their preference for certain settings to avoid color distortion.

20:05

🐞 Addressing Bugs and Workarounds in AI Diffusion

The paragraph addresses the bugs encountered while using the AI diffusion 1.15 update. The presenter shares a runtime error related to data type mismatch and discusses a workaround involving altering the bat file, which was previously effective for a different software. The video suggests adding arguments to the boot process to fix the issue and mentions reaching out to the developer, suspicious Firefly, for a solution. The presenter also shares a workaround for an expand feature bug, which involves adding a dot to the image to provide the AI with enough information to generate the correct content.

25:06

🎉 Wrapping Up and Encouraging Community Engagement

In the concluding paragraph, the video wraps up with a discussion on the capabilities of the AI diffusion 1.15 update, despite the bugs encountered. The presenter expresses gratitude for the community's support and encourages viewers to like, subscribe, and share the video. They also invite feedback on the graphics card used, error experiences, and render times to gain insights into real-world usage scenarios. The video ends with a thank you note for watching and an invitation to join the next stream.

Mindmap

Keywords

💡AI diffusion

AI diffusion refers to a technique used in the 1.15 update by akley for Cryer, which involves using artificial intelligence to generate and manipulate images. It is a powerful tool that allows users to edit their artwork in ways that were not previously possible, providing features comparable to Adobe's Photoshop for generative fill. The AI diffusion is controlled by the user, which helps to prevent unwanted glitching or artifacts in the artwork.

💡Cryer tools

Cryer tools are a set of utilities that are used in conjunction with the AI diffusion update. These tools enhance the functionality and user experience of the AI diffusion by providing additional features like auto-selection and outlining. They are integral to the process of using AI to edit and generate artwork, as they streamline and simplify the tasks that artists perform.

💡GitHub

GitHub is a web-based platform that provides version control and collaboration features for software developers. It is used by developers to store, manage, and collaborate on their code projects. In the context of the video, GitHub is where the AI diffusion update and Cryer tools are hosted, allowing users to download and implement these updates for their software.

💡Auto-selection

Auto-selection is a feature within the AI diffusion update that allows the software to automatically identify and select areas of the image for editing or manipulation. This saves time and effort for the user, as it eliminates the need for manual selection and allows for more precise and efficient editing.

💡Generative fill

Generative fill is a feature in image editing software that uses algorithms to fill in selected areas with new content. This content is generated based on the surrounding image data, and it is designed to blend seamlessly with the existing image. In the context of the AI diffusion update, generative fill is a powerful tool that offers users a high level of control over the final result.

💡Lightning model

The Lightning model is a specific type of AI model introduced by Civ AI, which is designed to generate high-quality images, particularly of fantasy creatures. However, it is also effective for rendering humans and other subjects. This model is notable for its speed and efficiency, allowing for quick generation of detailed images.

💡Render

In the context of the video, rendering refers to the process of generating an image using AI algorithms. This process involves taking the input data, such as a prompt or a selection, and using the AI model to create a detailed, finished image. Rendering can involve complex calculations and may take varying amounts of time depending on the complexity of the image and the capabilities of the hardware.

💡Samplers

Samplers are a component of the AI diffusion process that determines the method used to generate the image. Different samplers can produce different results and have various characteristics, such as speed and quality. In the video, the user mentions DPM Plus+, SD, and SGM samplers, which are specific types of algorithms used for rendering.

💡Bugs

In software development, a bug is an error or flaw that causes the program to produce an incorrect or unexpected result, or to behave in unintended ways. In the context of the video, the user encounters bugs while using the AI diffusion update, particularly when working with certain features on older hardware.

💡Hardware

Hardware refers to the physical components of a computer system, such as the graphics card, processor, and memory. In the context of the video, the user discusses the performance of the AI diffusion update on older hardware, like the GTX 1070 graphics card, and the limitations encountered due to these hardware constraints.

Highlights

The AI diffusion 1.15 update by akley for Cryer introduces a plethora of new features, enhancing the software's capabilities.

The update allows for better editing and utilization of artwork, rivaling Adobe's Photoshop in generative fill but with more control to prevent glitching.

GitHub is the platform where users can find and download the AI diffusion and Cryer tools for their creative projects.

The process of updating the AI diffusion is straightforward, involving downloading a zip file and extracting it over the existing files.

Civ AI has released a new lightning model that is not only useful for fantasy creatures but also works well with human renders.

The new core components in the 1.15 version of the plugin offer significant improvements over previous versions.

Users can now discard rendered elements they don't like, a feature that enhances the control over the editing process.

The ability to copy prompts is a handy addition, allowing users to quickly reuse or modify prompts for generating models.

Samplers play a crucial role in the rendering process, and the 1.15 update has made some adjustments to these for better performance.

The lightning sampler is designed to work seamlessly with the lightning model for optimal results.

The auto-selection feature with outlining is a fantastic addition for automatically selecting areas of the image for editing.

The expand feature allows for the generation of content around the selected area, taking into account the entire image for a more cohesive result.

The generate custom option provides users with more control over the generation process, including the ability to select the entire image or specific parts.

The 1.15 update has improved the blending and out-painting of generated content, resulting in more natural and seamless integration with the original image.

Users can adjust the diffusion settings to control how much the generated content fans out and paints outside of the selection.

The update has fixed some bugs from previous versions, such as the error when using the face tool, by altering the bat file and adding arguments.

Despite the new features, the 1.15 update still has some bugs, particularly on older hardware like the GTX 1070.

A workaround for the expand error involves adding a dot to the selection, which tricks the AI into generating the content without issues.

The AI diffusion 1.15 update is a significant step forward, offering more power and control to users for photo restorations and digital art creation.