your daily reminder that you can simulate the exact light transport for a 3d scene and not end up with something that looks realistic. realism is based on modelling, materials, and lighting. you can make something realistic using the blinn-phong model and some clever fakery. all that matters is that it looks good
even realism shouldn't be the end goal of a 3d artist, it should be your expression and development. you can make a hyperreal 3d scene that is boring to look at. you need visual interest. ultimately art is a way to communicate, and it's important to learn how best to get your message across. realism is just a tool for that goal
this is why the fetishism for exact light transport algorithms and photorealistic rendering can be insidious. it doesn't matter if the material for some object in your scene is photorealistic when you were unable to get it looking like it did in your head. it doesn't matter that some beam of light wouldn't realistically fall onto some part of the image if having it there would make the image look 10x better.
this is why when disney made their principled shading model they purposefully added some parameters that would be impossible in the real world. in particular "specular tint"
check out their paper (page 12) https://disney-animation.s3.amazonaws.com/library/s2012_pbs_disney_brdf_notes_v2.pdf
@SuricrasiaOnline the principled bsdf is really neat imo because it _is_ physically accurate if you need it to be (like when you're trying to integrate rendered objects with real footage) but it lets you be not physically correct in useful ways if you want to
Cybrespace is an instance of Mastodon, a social network based on open web protocols and free, open-source software. It is decentralized like e-mail.