Yeah I was thinking that too, but the camera/computer doesn’t have any information about what wavelengths were stimulating, e.g., the red pixels; it only knows how much those pixels were stimulated. It is certainly possible that the input color matrix is somehow able to distinguish this (maybe when the red pixel is really truly the only thing being stimulated it assumes the color is near-infrared, but that could just as easily be a nearly pure tone in the normal range, e.g. from a screen), but I feel like the default color matrix wouldn’t be that extreme? Idk that’s speculation on my part though.