/sci/ - Science & Math

>>16140745
>Autism a simulated camera into existence
It's not actually as autistic as you might think. Pixar created fully digital simulated cameras to use in their movies ages ago. Renderman has physical based cameras built-in with easy to model lenses through the RixProjection interface. Some people on the Corona forums created presets for them. Cinema 4D had a similar scene camera project. Indigo Renderer could do it in the late 2000s, so could Sketchup. People using LuxRender in Blender way back also managed to do it. Also heard V-Ray has a physical camera but I don't think that's the same from what I remember.
>>16141024
As always, when you're actually simulating something. Whether that'd be a camera or an atmosphere. The fact that we can even do this on our home computers to begin with now is a minor miracle. It also depends on whether or not the rendering code can actually handle it. Cycles for instance despite its prowess and efficiency (especially with K-Cycles) was notoriously fucking bad at doing volumetric scattering (at least as far as planetary atmospheres are concerned) to the point where you were literally better off making your own volumetric engine. And even then from what I remember when you get to stuff like adding haze the render times go up drastically. There are still some tricks you can utilize however to cut down the render times so much to the point where it is basically free. Relatively speaking.