Simple question:
Is it better to have more 3D objects with less triangles in a single model, or 1 big mesh with more triangles? Basically, if I can save poly count by making a model in different pieces, or just combine all the 3D objects to make 1 mesh. Maybe a happy medium between the 2?
Thanks: James
It is quiet harder to provide the exact answer, you need to profile it as to get the best working solution.
However, for example You have a big Building 3d model with multiple components, Lets say windows, doors etc. The complete mesh will be rendered each time when it is inside the visibility. But if you break the model in to small chunks and if the player see only the front side of the house (Occlusion Culling Visibility) only that part will be rendered, if you have built the model as one Big piece, the complete model will be rendered as it would be harder for the engine to split without having actual broken mesh components. So depending on situation it is better to break the models in to chunks for faster renering.
Also check for optimization articles and Culling best practices.