You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Well, I don't think the IECore::Renderer subclasses should have to worry about it, so it wouldn't be a renderer option. I was thinking of a render:overscan option set by the StandardOptions node, and then in outputCamera() in RendererAlgo.h we'd adjust the resolution and screen window appropriately just before passing to the renderer. I don't have a strong opinion on how overscan should be specified - should it be % values or pixels? And should it be specified individually for each edge or just once overall? I guess % vs pixels is partly down to how we want to handle specifying half-res renders and the like - do we want to just change render:resolution explicitly, or would we prefer a render:resolutionMultiplier option? The latter seems to imply that we might want to use % rather pixels for overscan.
We haven't done anything about film fitting yet either - what to do if the aspect ratio from the resolution and from the camera frustum don't match. Now's probably a good time to do that too.
As discussed in ImageEngine/cortex#330.
The text was updated successfully, but these errors were encountered: