Closed jeffreyameyer closed 4 months ago
Ok.... this may be a problem in the underlying import data... and also in something we're doing in our rendering.
The vtiles reflect the same shading differences:
So, it looks like some sort of transparency is being applied to the fills as you zoom out, creating the different intensities, but is that the correct behavior? Shouldn't it be a constant alpha? Otherwise, in areas where we have overlapping landcover, but 1 is supposed to be "over" the other, the areas of overlap won't be visible at higher zooms (good), but will be at lower zooms (bad).
What's weird to me is that the transparency effects disappear at zoom=7
and zoom=6
- see below:
I do not see this happening on my end (copy pasting your image): I only see the below on my end:
I DO see that there are 3 different landuse data layers. One for z5, one for z7, and one for z10, and then broader diversity in z12.
Taking a look at these and will make some changes, at least for the consistency in zoom in/out using the above as example
Made some updates and reviewed across zoom levels, lmk what you think. Demo style: https://openhistoricalmap.github.io/mbgl-timeslider/demo/#3.309/34.68647/-96.42187/2023,-4000-2023
in style ✅
Bug description What is happening? When I look at where I've imported some
natural=wood
over a large area at different zooms, there are some very different effects going on across the areas ofnatural=wood
.See Glacier National Park:
At
zoom=11
, the coloring in thenatural=wood
areas is consistent.At
zoom=10
, the coloration is very inconsistent, often along blocky boundaries. I'm wondering if this is a weird artifact of our terrain layer, but I would think that would be under our data.Here's
zoom=9
:same for
zoom=8
:What should be happening? These areas should all be consistent in appearance and not blotchy.