HARP-13235: Fix altitude in mercator #2055
Conversation
atomicsulfate
force-pushed
the
HARP-13235_FixMercatorAltitude
branch
from
c52a937
to
ec590c8
Compare
Codecov Report
@@ Coverage Diff @@
## master #2055 +/- ##
==========================================
+ Coverage 66.65% 66.81% +0.16%
==========================================
Files 297 297
Lines 26369 26378 +9
Branches 5966 5969 +3
==========================================
+ Hits 17575 17624 +49
+ Misses 8794 8754 -40
Continue to review full report at Codecov.
|
| @@ -92,7 +92,7 @@ class MercatorProjection extends Projection { | |||
| result.y = | |||
| (MercatorProjection.latitudeClampProject(geoPoint.latitudeInRadians) * 0.5 + 0.5) * | |||
| this.unitScale; | |||
| result.z = geoPoint.altitude ?? 0; | |||
| result.z = (geoPoint.altitude ?? 0) * this.getScaleFactor(result); | |||
There was a problem hiding this comment.
i don't think this is a good idea. This will enforce applying the scaling factor to Z at projection time, resulting always in different heights for objects with large latitude span. So many objects will have oblique planes and their normal vector will also be different, effecting bounding boxes, shading and tessellation. Also, this change will also conflict with the semantics of extruded-technique's constantHeight.
There was a problem hiding this comment.
As discussed, I now scale the height on decoding depending on the technique parameter constantHeight, which now is available in all techniques.
atomicsulfate
force-pushed
the
HARP-13235_FixMercatorAltitude
branch
3 times, most recently
from
5b3a689
to
8416574
Compare
If datasource geometry is 3D and technique's constantHeight is set to true or defaults to true (for store levels lower than 12), then projected geometry height is scaled by target projection's scale factor.
atomicsulfate
force-pushed
the
HARP-13235_FixMercatorAltitude
branch
from
8416574
to
a339006
Compare
No description provided.