nofaceinbook / muxp

Mesh Updater X-Plane
GNU Lesser General Public License v3.0
18 stars 3 forks source link

Flickering edges (often for ortho mesh) #53

Closed nofaceinbook closed 3 years ago

nofaceinbook commented 3 years ago

FlickeringEdges

Flickering edges are often seen on ortho mesh. Issue is probably coming from different scalings for vertices in original mesh and scaling assigned by MUXP. Issue should have been solved before.

--> seems to be scaling for s/t coordinates. When cut vertices are also border vertices for draping the ortho photo than calculations might not exactly fit and going out of bound. Create_dsf_vertices functions creates than also a vertex out of bound and going out of bound of the photo which causes slightly visible / flickering edge. ===> Was also issue that is solved now.

==> But again it is coming from different elevation scalings. ==> Elevation scaling for new vertex pools has to match existing scalings.

nofaceinbook commented 3 years ago

Partially solved. When existing scaler is better then required elevation_step this is used. However, still issue that a new better elevation_step does not exactly hit the values of previous scaling. Step must be some kind of multiple. Also required for default mesh to hit the exact meter values.

nofaceinbook commented 3 years ago

Should be solved. To be tested. But also set the base value of new pool to the lowest elevation in area. To be calculated e.g. when checking mesh. --> NO not changed. Setting different bases for always lowest vertex might lead to too many pools. Now starting base at -10m

nofaceinbook commented 3 years ago

Setting as elevation step default value of 0.0196078431372549 meters ( 1/51) which matches full meters and ranges for 1285 meters. Starting default base at -10m. So first range would cover [-10 m, 1275m] which should fit in most cases and if not accorindg lower range or higher ranges are used in 1285m steps.

DONE. Will be included in next release.

nofaceinbook commented 3 years ago

Solved with version 0.3.5