ComfyUI Node: High Resolution v2

High Resolution v2:

The SeargeHighResolution node is designed to enhance the resolution and detail of images in AI art generation workflows. It provides a sophisticated mechanism to upscale images while maintaining or even improving the quality and detail. This node is particularly beneficial for artists looking to refine their AI-generated images, ensuring that the final output is of high resolution and visually appealing. By leveraging advanced upscaling models and detail processors, the node can significantly boost the detail and clarity of images, making it an essential tool for high-quality image production.

High Resolution v2 Input Parameters:

hires_mode

This parameter determines the mode of high-resolution processing to be applied. It influences the overall approach the node takes to upscale and detail the image. The specific modes available and their impact on the image processing are not detailed in the context, but typically, such modes could range from simple upscaling to more complex detail enhancement techniques.

image_width

Specifies the width of the input image. This parameter is crucial as it defines the initial dimensions of the image before any upscaling or detailing is applied. The width, along with the height, helps in calculating the new dimensions after processing.

image_height

Specifies the height of the input image. Similar to the image width, this parameter defines the initial height of the image and is used to determine the new dimensions post-processing.

hires_scale

This parameter sets the scale factor for upscaling the image. It can take values like UI.HIRES_SCALE_1_25, UI.HIRES_SCALE_1_5, and UI.HIRES_SCALE_2_0, corresponding to upscale factors of 1.25, 1.5, and 2.0 respectively. The chosen scale factor directly impacts the final resolution of the image.

upscale_factor

Defines the factor by which the image will be upscaled. This is derived from the hires_scale parameter and determines the extent of enlargement applied to the image.

new_width

The width of the image after upscaling. This parameter is calculated based on the original width and the upscale factor, and it defines the target width for the high-resolution image.

new_height

The height of the image after upscaling. Similar to the new width, this parameter is calculated based on the original height and the upscale factor, setting the target height for the high-resolution image.

hires_softness

Controls the softness of the high-resolution image. A value less than 0.999 indicates the need for nearest-neighbor scaling, while a value greater than 0.001 indicates the need for bicubic scaling. This parameter helps in blending different scaling methods to achieve the desired image softness.

hires_detailer_enabled

A boolean parameter that indicates whether the high-resolution detailer is enabled. When enabled, additional detail processing is applied to the upscaled image, enhancing its clarity and detail.

High Resolution v2 Output Parameters:

latent

The latent representation of the high-resolution image. This output is crucial for further processing in the AI art generation pipeline, as it encapsulates the enhanced details and resolution of the image in a format suitable for subsequent stages.

upscaled

The upscaled version of the input image. This output provides the high-resolution image after applying the specified upscaling and detailing processes, ready for final use or further refinement.

High Resolution v2 Usage Tips:

• To achieve the best results, ensure that the hires_scale parameter is set according to the desired level of detail and resolution. Higher values will result in more significant upscaling.
• Enable the hires_detailer_enabled parameter to add extra detail to the upscaled image, especially if the initial image lacks clarity.
• Adjust the hires_softness parameter to blend different scaling methods and achieve the desired image softness, balancing between sharpness and smoothness.

High Resolution v2 Common Errors and Solutions:

Warning: high res upscaler should be a 4x ESRGAN model

• Explanation: This warning indicates that the upscaling model used is not a 4x ESRGAN model, which is expected for high-resolution upscaling.
• Solution: Ensure that the upscaling model provided is a 4x ESRGAN model to achieve optimal results.

Warning: detail processor should be a 1x ESRGAN model

• Explanation: This warning suggests that the detail processor used is not a 1x ESRGAN model, which is required for proper detail enhancement.
• Solution: Verify that the detail processor model is a 1x ESRGAN model to ensure the correct application of detail processing.

Image dimensions mismatch after upscaling

• Explanation: This error occurs when the dimensions of the upscaled image do not match the expected dimensions based on the upscale factor.
• Solution: Check the upscale factor and the initial image dimensions to ensure they are correctly set. Adjust the parameters if necessary to match the expected output dimensions.