ComfyUI Node: LatentUpscaleMultiply

LatentUpscaleMultiply:

The LatentUpscaleMultiply node is designed to upscale the latent code by multiplying the width and height by specified factors. This node is particularly useful for AI artists who want to enhance the resolution of their latent representations without losing the integrity of the original data. By applying different upscale methods, you can achieve various effects and levels of detail in the upscaled output. The node also offers cropping options to manage the resulting dimensions effectively. This functionality is essential for tasks that require higher resolution latent codes, such as generating more detailed images or improving the quality of AI-generated art.

LatentUpscaleMultiply Input Parameters:

samples

This parameter represents the latent code that you want to upscale. It is the core input for the node and contains the data that will be processed.

upscale_method

This parameter determines the method used for upscaling the latent code. The available options are nearest-exact, bilinear, and area. Each method has its own characteristics: nearest-exact is fast and simple, bilinear provides smoother results, and area is good for downscaling but can also be used for upscaling. The choice of method can significantly impact the quality and appearance of the upscaled output.

WidthMul

This parameter specifies the factor by which the width of the latent code will be multiplied. It accepts a floating-point value with a default of 1.25, a minimum of 0.0, and a maximum of 10.0. Adjusting this factor allows you to control the horizontal scaling of the latent code.

HeightMul

This parameter specifies the factor by which the height of the latent code will be multiplied. Similar to WidthMul, it accepts a floating-point value with a default of 1.25, a minimum of 0.0, and a maximum of 10.0. This factor controls the vertical scaling of the latent code.

crop

This parameter determines the cropping method applied to the upscaled latent code. The available options are disabled and center. disabled means no cropping will be applied, while center will crop the image to keep the central part, which can be useful for maintaining focus on the main subject.

LatentUpscaleMultiply Output Parameters:

LATENT

The output parameter is the upscaled latent code. This parameter contains the processed data with the new dimensions as specified by the input parameters. The upscaled latent code can be used in subsequent nodes for further processing or final output generation.

LatentUpscaleMultiply Usage Tips:

• To achieve a smoother upscale, consider using the bilinear method, especially when working with images that require finer details.
• If you want to maintain the aspect ratio of the original latent code, use the same value for both WidthMul and HeightMul.
• Use the center crop method if the main subject of your latent code is in the middle and you want to keep it focused after upscaling.

LatentUpscaleMultiply Common Errors and Solutions:

ValueError: Invalid upscale method

• Explanation: This error occurs when an unsupported upscale method is selected.
• Solution: Ensure that the upscale_method parameter is set to one of the following: nearest-exact, bilinear, or area.

ValueError: WidthMul or HeightMul out of range

• Explanation: This error occurs when the values for WidthMul or HeightMul are outside the allowed range.
• Solution: Ensure that WidthMul and HeightMul are within the range of 0.0 to 10.0.

RuntimeError: Cropping method not supported

• Explanation: This error occurs when an unsupported cropping method is selected.
• Solution: Ensure that the crop parameter is set to either disabled or center.