Hello, just making sure I have not misconfigured anything. For me there is a pretty large discrepancy between the surface of CWT and GP3DT. There is no rotation applied to the CWT or GP3DT game objects.
What is this difference down to? Is it different reference ellipsoids, I am genuinely curious.
I believe this is simply a result of Google P3DT being higher resolution than CWT. You can see the information on the resolution of CWT in this table. I don’t have similar information at hand for Google P3DT, but for areas where they have 3D city data, it’s clearly much higher resolution - on the order of a few meters or even tens of centimeters. In an area where CWT has a resolution of 30m, for example, that can be a huge difference - especially in cities.
We’ve also seen cases where GP3DT simply has incorrect heights. And of course we can’t completely exclude an error in Cesium World Terrain, either (though we do have confidence in our process, at least relative to our source data).
If you see something that seems like an error, rather than simply a difference of resolution as Ashley suggested, then the best bet is to check the heights reported by each against a trusted third party. If you believe Google is wrong, you’ll have to report it to them. If you believe CWT is wrong, please let us know.
As an example, GP3DT is 4 meters lower than CWT at the Eiffel Tower. For Manhattan the difference seems to be a few centimeters.
Since the CWT resolution table lists 30 meters for Europe, a 4-meter discrepancy is probably to be expected.
