ComfyUI Node: 🎯 AI → Camera‑Accurate Log Simulator

🎯 AI → Camera‑Accurate Log Simulator:

The LogReconstructionNode is designed to simulate camera-accurate log encoding, a process that is crucial for achieving realistic color grading and dynamic range compression in digital imaging. This node is particularly beneficial for AI artists who wish to emulate the look and feel of footage captured by professional cameras, which often use log encoding to preserve details in both shadows and highlights. By converting images from a linear color space to a logarithmic one, the node allows for more nuanced control over exposure and color grading, making it easier to achieve cinematic results. The node also offers features like inverse tone mapping and local gain adjustments, which enhance the flexibility and precision of the log simulation process. Overall, the LogReconstructionNode is an essential tool for artists aiming to create visually compelling and technically accurate digital art.

🎯 AI → Camera‑Accurate Log Simulator Input Parameters:

image

The image parameter is the input image that you want to process through the log reconstruction pipeline. It should be provided as a tensor, typically representing an RGB image. This parameter is crucial as it serves as the base for all subsequent transformations and adjustments.

input_space

The input_space parameter specifies the color space of the input image. Options include "sRGB", "Rec.709", and "ACEScg". This parameter determines how the image will be linearized before applying log encoding, affecting the final appearance of the processed image.

log_transform

The log_transform parameter defines the type of logarithmic transformation to apply. Options include "Alexa_LogC3", "Canon_Log2", "Sony_SLog3", and "ProRes_Log". This choice impacts the dynamic range and tonal characteristics of the output image, simulating different camera log profiles.

use_inverse_tonemap

The use_inverse_tonemap parameter is a boolean that, when enabled, applies a soft shoulder boost to the image, simulating an inverse tone mapping effect. This can enhance the highlight details and overall contrast of the image.

exposure_index

The exposure_index parameter adjusts the exposure level of the image. It is a numerical value that scales the image's brightness, allowing for compensation of underexposed or overexposed images. The default value is typically set to 800, but it can be adjusted based on the desired exposure level.

auto_analyze

The auto_analyze parameter is a boolean that, when enabled, automatically analyzes the image to adjust exposure based on mean luminance. This feature helps in achieving a balanced exposure without manual intervention.

local_gain_strength

The local_gain_strength parameter controls the strength of local gain adjustments, which are applied based on depth and normal maps. This parameter allows for localized contrast enhancements, adding depth and dimension to the image.

depth_map

The depth_map parameter is an optional input that provides depth information for the image. When used, it helps in applying local gain adjustments, enhancing areas that are closer to the viewer.

normal_map

The normal_map parameter is an optional input that provides normal vector information for the image. It is used in conjunction with the depth map to apply more precise local gain adjustments.

save_exr

The save_exr parameter is a boolean that, when enabled, saves the processed image as an EXR file. This format is useful for high dynamic range imaging and further post-processing.

exr_filename

The exr_filename parameter specifies the filename for the saved EXR file. It is used only if save_exr is enabled.

save_dpx

The save_dpx parameter is a boolean that, when enabled, saves the processed image as a DPX file, a format commonly used in the film industry for high-quality image sequences.

dpx_filename

The dpx_filename parameter specifies the filename for the saved DPX file. It is used only if save_dpx is enabled.

🎯 AI → Camera‑Accurate Log Simulator Output Parameters:

tensor

The tensor output is the processed image in tensor format, ready for further use or analysis. It represents the image after all transformations, including log encoding and any optional adjustments, have been applied.

metadata

The metadata output provides additional information about the processed image, including details about the input space, log transform, exposure index, and any automatic gain adjustments. This metadata is useful for understanding the processing steps and parameters applied to the image.

🎯 AI → Camera‑Accurate Log Simulator Usage Tips:

• To achieve the most realistic camera simulation, choose the log_transform that matches the camera profile you are trying to emulate.
• Use the auto_analyze feature to automatically balance exposure, especially when working with images of varying brightness levels.
• Experiment with local_gain_strength in combination with depth and normal maps to add depth and contrast to specific areas of your image.
• Save your work in EXR format if you plan to perform further high dynamic range processing or need to maintain maximum image quality.

🎯 AI → Camera‑Accurate Log Simulator Common Errors and Solutions:

"EXR support not available"

• Explanation: This error occurs if the OpenEXR library is not installed, which is required for saving images in EXR format.
• Solution: Install the OpenEXR library using your package manager or pip to enable EXR support.

"Invalid input space"

• Explanation: This error indicates that the specified input_space is not recognized or supported.
• Solution: Ensure that the input_space parameter is set to one of the supported options: "sRGB", "Rec.709", or "ACEScg".

"Invalid log transform"

• Explanation: This error occurs when the specified log_transform is not recognized or supported.
• Solution: Verify that the log_transform parameter is set to a valid option such as "Alexa_LogC3", "Canon_Log2", "Sony_SLog3", or "ProRes_Log".