Mastering FaceSwap, Pyracanny, CPDS, and More in Fooocus!

Table of Contents

1. Introduction
2. History Log
3. Input Image: Upscale or Variation
4. Image Prompt
5. Control Nets
   • Pyan
   • CPDS
   • Face Swap
6. Settings and Their Effects
   • Weight
   • Stop At
   • Performance
7. Visual Examples
8. Conclusion
9. Resources
10. FAQs

1. Introduction

In this third installment of my series on Focus, we will delve deeper into the input image tab and explore some useful tricks and tips that you might not be aware of. I will explain certain settings and break down the control nets and their respective settings to help you better understand their functionality and determine which ones are most suitable for specific purposes. If you haven't watched the previous videos, feel free to catch up, but you can also jump right in and start exploring the features.

2. History Log

The history log is a Hidden gem that many users may not be aware of. By accessing the history log in the advanced settings, you can view all your previous generations. This log is particularly useful if you want to revisit the settings and inputs used for a previous image. You can simply copy them to the clipboard and load those parameters to recreate the image. However, please note that this feature only loads the settings and not the input images themselves.

3. Input Image: Upscale or Variation

Within the input image tab, there is a subtle setting called "Upscale or Variation." This tool provides an opportunity to introduce slight variations to your image. It is especially handy for changing expressions in an image, as it can produce more natural results compared to the inpaint tool. By adjusting the resolution and experimenting with different settings, such as mouth width or smile intensity, you can subtly alter the expression of a person in the image.

4. Image Prompt

The image Prompt feature allows you to combine up to four images. By checking the "Advanced" checkbox, you gain access to additional settings and customization options. The "image prompt" setting, similar to the mid-journey setting, uses your uploaded image as the base to influence the final image's style, character, composition, and colors. By adding a text prompt, you can further refine the output based on your desired outcome.

5. Control Nets

5.1 Pyan

The Pyan setting is particularly useful when you want to preserve specific details or poses from one image in the final generation. It creates a line art picture that captures intricate details from your image and applies them to the text prompt and other control nets. This can result in compositions that closely Resemble the input image while incorporating the desired style changes.

5.2 CPDS

CPDS (Contrast Preserving Decolorization Structure) is a custom control net developed by the Focus team. It combines depth control and a decolorization method to convert the image into black and white. The structure of the black and white image then serves as a control net to influence the final image. CPDS and Pyan serve similar purposes, but they employ different methods. While you can combine multiple control nets, it's advisable to use either Pyan or CPDS but not both simultaneously.

5.3 Face Swap

Face swap allows you to detect and replace faces and hair from one image with another. This can be achieved using only a text prompt or a combination of images, control nets, and text prompts. The AI will identify the face and hair from the control image and apply them to the final generated image. By adjusting settings like face swap and adding image prompts, you can modify the composition and style of the generated image.

6. Settings and Their Effects

6.1 Weight

The weight setting determines the strength of influence of the selected control net on the final image. A higher weight value increases the control net's impact, while a lower weight value reduces its effect. Experimenting with the weight setting allows you to strike a balance between preserving the original image's structure and allowing the AI to explore more creative interpretations.

6.2 Stop At

The stop at setting controls the number of steps at which the selected control net's influence will be applied during the generation process. By default, each control net stops at 50%, allowing other control nets and text prompts to contribute to the final image. Adjusting the stop at value higher or lower can affect the degree to which the chosen control net adheres to the input image's structure.

6.3 Performance

Performance settings dictate the number of steps the image generation process undergoes before reaching completion. With the performance set to "speed," the image goes through 30 steps. Each step refines the image generated by the AI algorithm, resulting in a final output that satisfies the desired style and composition. Adjusting the performance setting can influence the level of detail and overall quality of the image.

7. Visual Examples

To better understand the impact of different settings, let's explore some side-by-side examples. By comparing images generated with varying stop at and weight values, you can observe how they affect the structure and artistic interpretation of the final output. Additionally, we will showcase examples that demonstrate the capabilities of face swap, highlighting its ability to Blend images and generate compositions that Align with specific prompts.

8. Conclusion

In this article, we have explored various features and settings within the Focus tool. From the history log to control nets and image prompts, there are numerous ways to customize and influence the generation process. By understanding the effects of different settings, you can harness the full potential of Focus and create unique and compelling images that align with your artistic vision.

9. Resources

• Official Focus Website - www.focus.com

10. FAQs

Q: Can I combine multiple control nets simultaneously?
A: While you can combine certain control nets, such as Pyan and CPDS, it is generally recommended to use only one at a time to achieve optimal results and avoid conflicting influences.

Q: Can I use my own images as control nets?
A: Yes, you can upload your own images and use them as a base for generating images with desired style, composition, and character. By combining them with control nets and text prompts, you can further refine the output.

Q: How do I choose the appropriate weight and stop at settings?
A: Experimentation is key when determining the ideal weight and stop at settings. Start with the default values and make adjustments based on your desired outcome. Increasing the weight strengthens the influence of the chosen control net, while adjusting the stop at value determines the duration of its influence during generation.

Q: Can I save my generated images for future reference?
A: Yes, you can save your generated images in your focus outputs folder. Each image is segmented into folders by date, allowing you to easily locate and revisit previously generated images.

Q: Can I mix images with different styles and compositions?
A: Absolutely! By utilizing image prompts and adjusting the weight settings, you can blend multiple images with distinct styles and compositions to create unique and visually appealing combinations.