How is Normal channel interpreted? Tweet

[LewisO]

I notice there is a difference when the output to Normal is from a texture node or OSL node.
When reducing the power of an image(which is theoretically the wrong way to change the strength of normal), the color information in RGB will get closer to 0, but if I check the info tangent normal pass, it looks like Octane will automatically centralize the value to 0.5 so the result is correct.
It's different however if I try to output some normal via OSL, R and G must be around 0.5 as the center value, and it looks like B must be set to 1 or higher than 0.5 at all times.

Any detailed explanation of the rule about how normal data is interpolated under the table?

[bepeg4d]

Which osl script are you using?
Could you share an example scene?

ciao Beppe

[LewisO]

Which osl script are you using?
Could you share an example scene?

ciao Beppe

I think I've made the question clear enough: the interpretation of color between a texture node and the output of an osl node is different when connected to Normal channel, even when "color in = 0, output color c = in". Any details about how the normal data is interpreted in Octane? is it based on the types of nodes?

[LewisO]

Any updates?