• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

P3D v5 Autogen trees on water

F

FLEX_45

Messages
143
Hi,
I tried to create a photoreal scenery both with FSET and Ortho4P3D, scenproc and OSM data but always run into the same issue. For areas that have vegetation polygons near the coastline, some trees are also placed for some distance over water.
I checked the downloaded OSM data manually in QGIS and also after the build the result in annotator and everything is looking good but in the simulator (P3Dv5) I see a lot of trees in the water. Any ideas? I am attaching a screenshot of the annotator, scenproc basic script I am using for vegetation trying to troubleshoot the issue and the in sim result.

Many thanks

-------------------------------------------
ImportOGR|@0@|*|building;landuse;natural|NOREPROJ

SplitGrid|AGN|*|*

FilterFeatures|*|*|DISTANCE_FEATURE|3

CreateAGNPolyVeg|FTYPE="POLYGON" And landuse="forest"|{0053b63d-b2c0-4bd9-9853-d9d21c9ad1fa}
CreateAGNPolyVeg|FTYPE="POLYGON" And natural="scrub"|{2fde0277-1697-4dab-b481-c3985c80596f}
CreateAGNPolyVeg|FTYPE="POLYGON" And landuse="orchard"|{56a4998d-c1e5-416c-a37b-c35ce16504b6}
CreateAGNPolyVeg|FTYPE="POLYGON" And natural="wood"|{dcf543b7-c0d5-4fd4-b970-83965c2911c9}
CreateAGNPolyVeg|FTYPE="POLYGON" And natural="tree"|{7dc6ef4e-92a5-4d0d-b94b-212dd1fa936d}
CreateAGNPolyVeg|FTYPE="POLYGON" And natural="tree_row"|{a4a30975-075c-49ec-87fb-7e0931cb5004}

EXPORTAGN|FSX|@1@
------------------------------------------

2022-4-16_21-24-15-481_annotator.jpg


2022-4-16_21-24-15-481.jpg
 
Are you sure you don't have any old AGN files in your texture folder that might have more trees in there? This is really a lot of trees in the sea and I think it covers more then one AGN file as well.
 
Yes because once I delete the agn files I create the trees go away.....so it's definitely coming from the newly generated agn files.
 
I also tried the ReplacePolygonByVegetationRectangles step as well, again in annotator everything looks ok but in the simulator I get the same result.
 
It's the PR created from Ortho4P3D and FSET with default settings running the above scenproc script. I tried both NOREPROJ and AUTODETECT without any change on the results.
 
it may be that your veg area poly looks aligned from that zoom distance,
once you zoom in you may see coastline and poly area imperfections,
 
Unfortunately it's not that....poly is inland even when zoomed
Also it's not only on that specific area, it's a general issue when there are large vegetation areas very close to water.
 
in that case it is likely that your native LC coastlines are not defined/refined enough for that area,
PR may look great but not necessarily aligned to sim native landclass coastline properties,
if your lc coastline was defined properly there you shouldn't see trees over water,
you will need a fine granularity coastline shape to work off of,

did you define and rendered coastline/water mask with your PR set or just a flat PR image with no coastline/water blend?
 
Last edited:
In FSET I run the process with create masks = yes and in ortho default settings except use_masks_for_inland which I set to True as it was mentioned in a tutorial I found.

On the other hand even if coastline is not accurate, how does it allow autogen to be drawn for such a large area outside the polygon?

Are there any additional settings I should use in ortho options?
 
activating water mask option in FSET just tells the tool to use a pre provided area.kml file with coast and blend polylines notated accordingly in order for FSET to be able to use the info,
generally speaking you placed an image over existing native LC data, just because you place an image that doesn't mean the native drawn LC data will be aligned or match,
especially with these oddly shaped coastlines native data will not be refined to match a PR coastline,
therefore the data below might show water on your PR set but LC below is defined different, probably as land,

you can try to land and save a flight in the middle of the water area where you see trees, turn off your PR set,
than go back in sim to the same saved area you will likely find that you are on land not water as your PR showing,

FYI split grid AGN will still give you rectangular shaped edges, not refined enough for coastline rendering,
with no defined LC coastlines you will get a full AGN rectangular filled with trees,

if you inspect the edges of the trees over the water where they end, you will notice its rectangular shaped,
that is likely where your coastline is in that area, prob a very low res poly and why you get such rough edges,
 
Last edited:
The mistake is in your scenProc script, you are not splitting the polygons into the AGN tiles (look at the Annotator screenshot, you don't see the polygons align to the cells there).

The line

Code: 
SplitGrid|AGN|*|*

is wrong, you are excluding all polygons from splitting with the last *. That is normally used to prevent that buildings are split in two parts, but should not be used on vegetation polygons. So instead you should use:

Code: 
SplitGrid|AGN|*
 
Many thanks for the input from both - it looks like the issue was indeed on the large polygon and the splitting I was doing. However, removing the last * from the SplitGrid step, indeed fixed the issue with the trees in water but inland trees were drawn into visible squares in a few places. So basically before the SplitGrid step, I split the large poly into smaller rectangles with ReplacePolygonByVegetationRectangles step and then did the SplitGrid with both * and worked perfectly.

Many thanks to both once more for the help!
 
Hi,

I'll repeat, you should not use the second * in SplitGrid for vegetation. It is only supposed to be used for features that should not be split on the tile edges, like buildings. For buildings you will get weird issues if they are split (as you get two buildings). For vegetation it should always be split.
 
Hi,
In cases like this one, that vegetation polygon from OSM is very big (covering many grids), without the second * I get weird results (very noticable squares of vegetation - ie. you can tell the grids in the sim). Is there another way around this?
 
Do you mean some tiles have no output?

How does it look in Annotator when you process the polygons with SplitGrid?
 
Hi,
First screenshot is just with the second * removed and the second is with the ReplacePolygonByVegetationRectangles step before splitGrid and still without the second * (which worked perfectly).

Agn_1.JPG


Agn_2.JPG



Many thanks!
 
Which version from scenproc are you using? A lot of polygons seem to be lost when splitting.
 
I downloaded Scenproc quite some time now, however checking at the site says latest stable version is 3.0 and doing a properties on my scenproc.exe shows version 3.0.0.0
 
Version 3.0 should be OK, you could give the development release version 3.1 a try just to be sure.
 
You must log in or register to reply here.
Back
Top