ComfyUI Node: HighRes-Fix Script

HighRes-Fix Script:

The HighRes-Fix Script is designed to enhance the resolution of images generated by AI models, ensuring that the final output is of high quality and detail. This script is particularly useful for AI artists who want to upscale their images without losing the intricate details and textures. By leveraging advanced upscaling techniques, the HighRes-Fix Script allows you to refine and improve the resolution of your images, making them suitable for various applications, from digital art to professional presentations. The script provides flexibility in choosing different upscaling methods and parameters, allowing you to customize the upscaling process according to your specific needs and preferences.

HighRes-Fix Script Input Parameters:

upscale_type

This parameter defines the type of upscaling method to be used. It determines whether the upscaling will be performed in the latent space or directly on the pixel level. The choice of upscale type can significantly impact the quality and speed of the upscaling process.

hires_ckpt_name

Specifies the name of the checkpoint to be used for high-resolution upscaling. This checkpoint contains the pre-trained model weights that will be utilized during the upscaling process.

latent_upscaler

Indicates the method to be used for latent space upscaling. This parameter allows you to choose from various latent upscaling techniques, each with its own advantages and limitations.

pixel_upscaler

Defines the method to be used for pixel-level upscaling. This parameter allows you to select from different pixel upscaling techniques, which can affect the final image quality and processing time.

upscale_by

Determines the factor by which the image will be upscaled. This value is typically between 1 and 2, and it controls the extent of the upscaling process. The higher the value, the more the image will be enlarged.

use_same_seed

A boolean parameter that specifies whether to use the same seed for the upscaling process. Using the same seed ensures consistency in the generated images, which can be useful for reproducibility.

seed

The seed value to be used for the upscaling process. This value influences the randomness in the upscaling algorithm, affecting the final output.

hires_steps

Specifies the number of steps to be taken during the high-resolution upscaling process. More steps can lead to better quality but will also increase the processing time.

denoise

A parameter that controls the amount of denoising applied during the upscaling process. Denoising helps in reducing artifacts and noise, resulting in a cleaner image.

iterations

Defines the number of iterations to be performed during the upscaling process. More iterations can improve the quality but will also require more computational resources.

use_controlnet

A boolean parameter that indicates whether to use ControlNet for the upscaling process. ControlNet can provide additional control and refinement during the upscaling.

control_net_name

Specifies the name of the ControlNet model to be used if use_controlnet is set to true. This model provides additional guidance during the upscaling process.

strength

Controls the strength of the upscaling effect. Higher values result in more pronounced upscaling, while lower values provide a subtler effect.

preprocessor

Defines the preprocessor to be used before the upscaling process. The preprocessor can perform various tasks such as normalization and augmentation to prepare the image for upscaling.

preprocessor_imgs

Specifies the images to be used by the preprocessor. These images are processed before the upscaling to ensure optimal results.

script

An optional parameter that allows you to provide a custom script for the upscaling process. This script can override the default behavior and provide additional customization.

my_unique_id

A unique identifier for the upscaling process. This ID can be used to track and manage different upscaling tasks.

HighRes-Fix Script Output Parameters:

script

The output script contains the configuration and parameters used during the upscaling process. This script can be used to reproduce the upscaling or to apply the same settings to other images.

HighRes-Fix Script Usage Tips:

• Experiment with different upscale_type and latent_upscaler combinations to find the best quality for your specific images.
• Use the use_same_seed parameter to ensure consistency across multiple upscaling tasks, especially when working on a series of images.
• Adjust the denoise parameter carefully to balance between reducing noise and preserving details in the image.
• Utilize the iterations parameter to improve the quality of the upscaled image, but be mindful of the increased processing time.

HighRes-Fix Script Common Errors and Solutions:

"Chosen latent upscale method not found! defaulting to 'default_method'."

• Explanation: This error occurs when the specified latent upscaler method is not found in the available methods.
• Solution: Ensure that the latent_upscaler parameter is set to a valid method. Refer to the documentation for the list of supported upscaling methods.

"When using 'ttl_nn.<method>', 'upscale_by' must be between 1 and 2. Rounding to the nearest valid value."

• Explanation: This warning indicates that the upscale_by value is outside the acceptable range for the specified method.
• Solution: Adjust the upscale_by parameter to a value between 1 and 2 to avoid this warning and ensure optimal upscaling results.