coming from an OpenGL programming background, I´m just setting up multiple camera rendering in TD and wondering if there is something, where I can define the render viewport of my target render buffer. Similar to glViewport in OpenGL.
For now I´m using Render Select with the Camera Index and after that a lot of Composition Nodes.
Has anyone experience regarding performance compared to glViewport in OpenGL, when having high resolution, 8k and above ?
thanks for the link. It already helped me setting up my project.
I´m wondering if there is a solution to define within the camera component a viewport where it should render to in the render component.
So for example, if I have a render component with 4k resolution and 4 camera components.
Then I want to define that camera 1 is rendering top-left part in the render component, camera 2 top-right part and so on.
I’m aware that I can sticht it together later with renderselect (or use multiple render TOPS) and a composition, but this would also mean processing overhead right ? Or do I get anything wrong here ?
I’d suggest using Render Selects.
As a Render Select takes almost no extra cooking time, it’s just a pointer to a buffer which is already rendered, you can compare it to a Select TOP.
If you make the Render TOP a quarter of your target (4k) resolution, the end result of 4 cameras stitched together will be your target resolution, and there is no loss in unnecessary pixels processed.
Here’s a quick example using the new Layout TOP from 099 experimental.
Render TOP is 1024x1024 pixels (so each camera is rendering at this output) , the end composition is 2048x2048. multi_cam_renderselect.tox (2.31 KB)