It would be great if there was a feature to set a project wide lightmap resolution quality which would then assign a new lightmap resolution to each object based on the object’s physical size.
There would also need to be an exclude list, so outer walls for example that don’t need decent shadows could be left unaltered.
The individual lightmap resolutions could still be modified independently after running the autolightmap function as necessary.
The engine already has a feature that allows you to edit lightmap resolutions for multiple objects (for example, increase 2x)
But, the engine can’t figure out best quality for an object, it depends not only on the size of the object, but also on things like how detailed the mesh is, and how detailed the shadows need to be. The adjustments that need to be made can often only be understood after lighting is built and you can see what the results are so you can make adjustments. That also doesn’t take into account optimization where you might lower the resolution to improve build times or to save memory.
Hi Darthviper. I understand what your saying, to resolve the issues you’ve pointed out, I would suggest that the feature took into consideration polygon/vert count as well as physical size. This would allow it to determine mesh complexity/detail.
I’m not suggesting that it would replace the need to optimize, but the exclude feature would allow to ignore objects that you want to keep low res for optimization reasons, thus addressing your point on optimization.
It would really be a huge timesaver, as when I optimize my lightmap res, i’m basically doing so based on how large/detailed the object is and how important its quality is. The feature would save hours of work, and just leave optimization of a few (excluded) objects to be done manually.
Of course if some of the auto generated lightmaps needed changing afterwards, there’d be nothing stopping the user doing this.
Polygon count isn’t always a determination of detail either. I think over time as you get more experience then you can be pretty sure about the lightmap resolution needed for a mesh and it wouldn’t be an issue.
It’s not that i’m unable to do it manually, my current skill level allows excellent results…it’s more of an efficiency thing 