There isn’t a single “correct” or universally perceptual mapping from RGB+IR to display RGB. Even ordinary photos require interpretation through tone mapping, gamma, and color rendering, so adding infrared makes the result even more subjective.

In practice, you have two main options:

1. Artistic/manual mapping: Choose a false-color rendering that emphasizes the information you want, then refine it with curves, masks, and local adjustments in an editor. Different parts of the image may need different treatment.

2. Data-driven compression: Use a method such as PCA (principal component analysis) to transform the multiple input bands into three new components that retain as much variation/information as possible. This is a common engineering approach for reducing multi-band data to three channels.

This is similar to how astronomy imaging often maps non-visible filter data into RGB for human viewing: the channels are intentionally remixed rather than reproduced “literally.”

So the short answer is: no one-size-fits-all mapping exists. If your goal is visual impact, edit creatively; if your goal is preserving information from multiple bands, try PCA or a similar multi-band reduction method, then tone-map the result for display.