I've been nosing around the code some and realize that what I'm looking for is at the intersection of performance, culling, the normal use cases, and my own neurosis.
Basically, when using the 3D view, the effective zoom level of tiles near the edges of the whole planet are less than what is shown in the center of the current view. I understand this to be necessary in overall calculations for sanity's sake, but was wondering if there are any tips of manipulating this.
I understand that if I was closer to the ground in 3D view, like say at zoom level 10 or so, and then were to look towards the horizon, having zoom level 10 for every possible sliver of a tile in view would cause way too many tiles to be loaded.
I guess I was just wondering if there was a way to make at least at lower levels (like say 3) that the whole planet would be consistently shown at tile level 3, rather than the edges gracefully degrading to 2 and 1, or at least some way to make less of the edges go back to the lower levels.
I see this most with say Open Street Map layers. You start to see the outlines of the US States, but you don't see North Dakota, for instance. By the time you get some highway and road detail on the east coast, only then do you get detail for say Canadian provinces, if they are still in view.
I may be in what the Perl people call an "A/B" problem, perhaps this is manipulated with camera frustrum and such? Thank you!