• 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.

FSX SBuilderX, sloping ground polygons.

Messages
96
Country
us-washington
I'm stuck... I almost have all of my roads completed, 3D object modeling is coming to a wrap here in the next couple weeks and now I'm ready to tackle the biggest hurdle of them all - reestablishing the pre-1980 elevation data. To that end, I've downloaded five 15-minute USGS quadrangle topographic maps of my scenery package's coverage area and these topographic maps have an 80-foot contour interval. That roughly equates to a 30-meter mesh resolution. These topographic maps, while done in the 1950s, remain to date the most accurate and highest quality representation of the pre-eruption elevation data in existence of the area.

However, attempts to find a reasonable source for help on creating the DEM from this contour data are falling flat. In light of that, I've loaded the topographic maps into SBuilder, calibrated their XY coordinates for the NW and SE corners, and began experimenting. In the documentation I have read, FSX has support for sloping ground polygons, and it seems as though SBuilder has support for creating them.

Two nights ago, I set out on a test to create a sloping ground polygon matching that of the contours in the Spirit Lake topographic map, in a small area near a cabin I've placed in the sim close to the Spirit Lake Lodge a mile west of Spirit Lake itself (this lodge was not the one owned by Harry Truman).

The purpose of this test is twofold:

1.) To see if I can override the terrain mesh in use in the area (ORBX Sceneries), and,

2.) to see if I can override a AB flatten I've placed under the cabin so as to have the terrain near the cabin show as it did based on reference photos of the actual cabin sent to me by a family member of the former cabin's owner.

To create the sloping polygon, I created a polygon with roughly twenty control points. All of the control points on the north and east side of the polygon were set at 3,200 feet (975.36 meters). Those control points on the south side of the polygon, and the two directly opposite the east side control points were set at an altitude of 3,120 feet, or 950.976 meters. A BGL was then attempted, at which point an error message was shown saying it cannot compile the *.SHP file. However, the BGL did get compiled with the file name CVX_MSH_PRE_1980_Test_Slope.BGL.

To see if it worked, I then checked off the check box in the compile BGL dialog in SBX to automatically load FSX upon completion of BGL compiling. After loading the new scenery database, a test flight was initiated from a test heliport I have worked out in the sim in the parking area of Harry Truman's lodge. After flying a mile and a half west, I noticed the slope did not appear.

Was this polygon outside the limitations of SBX and FSX, or what can I do to make this, and thus subsequent polygons, work in FSX?

Attached in this post, is a screenshot showing the exact polygon in question.
test1.jpg
 
Messages
7,450
Country
us-illinois
Hi Steve:

Sloped flattens are CVX vector polygons that are 'TIN' surfaces compiled by SDK SHP2VEC.



These CVX sloped flatten 'TIN' surface BGLs can modify the terrain shape otherwise formed by terrain Mesh BGLs.

These airport boundary / background (aka Airport Boundary Polygon aka "ABP") land class CVX vector types use a flatten-only GUID:

{47d48287-3ade-4fc5-8bec-b6b36901e612} Flatten ...(only)

http://www.prepar3d.com/SDK/Environment Kit/Terrain SDK/Terrain and Scenery.html#The Shp2Vec Tool


SBuilderX and ADE can output these types of BGLs.

MCX can output these types of BGLs by importing a Geo-located Sketchup Google Earth format *.KMZ 3D model of a terrain surface TIN.


A 3D model uses the Cartesian coordinate system, which as a 'non-warped' Mercator-type of projection format, has limits to precision by Geographic extent, so in GIS, such a 'non-warped' Mercator-type of projection format object is precisely rendered only up to a distance of 1 Kilometer from its central datum.


In FS SDK, the maximum extent of a LOD-13 / QMID-15 terrain grid quad is ~1023 Meters at the equator of FS' (theoretically) oblate 3D world globe.

So. large networks of Faces (whether TINs or quads) must use multiple smaller LOD-13 / QMID-15 sized sub-object parts to form larger 3D (sur-)Faces.

https://www.fsdeveloper.com/forum/threads/flattens.425495/post-633002


GIS applications can process source files of elevation data points via 'triangulation' routines to form 'TINs'.

If a 'triangulated' subset of the source data is output in EPSG:3857 GIS projection as a *.SHP file, the 'TIN' can be Appended to SBuilderX, then assigned the GUID for a CVX vector flatten-only airport boundary / boundary polygon, then output as a sloped flatten when compiled to a BGL by SDK SHP2VEC.

Note that SBuilderX (aka "SBX" for purposes of discussion) internally re-projects CVX data sets to EPSG:4326 prior to output / compilation via SDK SHP2VEC.

This is not true of non-SBX output methods, so source data sets of points and/or 3D models used to form TINs via non-SBX methods ...must be set to a 'warped' EPSG:4326 GIS format prior to triangulation, output and compilation via SDK SHP2VEC if intended for use as 'terrain-type' CVX vector objects.

WHY: FS SDK compilers for "Terrain" scenery require all source data to be submitted in EPSG:4326 GIS format. :pushpin:


Source data sets for non-terrain-type 2D and/or 3D models must be set to EPSG:3857 prior to output and compilation via BGLComp.


I have posted info on this topic in a number of threads over the years:

https://www.google.com/search?q=site:www.fsdeveloper.com+GaryGB+"TIN"&ei=1WWHYYaLPIm8tAbw0IRA&oq=site:www.fsdeveloper.com+GaryGB+"TIN"&gs_lcp=Cgdnd3Mtd2l6EAxKBAhBGAFQ2AhY7BBghyRoAXAAeACAAWGIAYwCkgEBM5gBAKABAcABAQ&sclient=gws-wiz&ved=0ahUKEwjGyZGNxIX0AhUJHs0KHXAoAQgQ4dUDCA0

https://www.google.com/search?q=site:www.fsdeveloper.com+GaryGB+"sloped+flatten"&ei=imWHYci8M8eDtQawpYToDg&oq=site:www.fsdeveloper.com+GaryGB+"sloped+flatten"&gs_lcp=Cgdnd3Mtd2l6EAxKBAhBGAFQygdYygdgkhBoAXAAeACAAY8BiAGPAZIBAzAuMZgBAKABAcABAQ&sclient=gws-wiz&ved=0ahUKEwjIqafpw4X0AhXHQc0KHbASAe0Q4dUDCA0


Tomorrow (Sunday), I shall post some links to such threads, with a worked example for the location cited in your OP above.

GaryGB
 
Last edited:
Messages
7,450
Country
us-illinois
Hi Steve:

In lieu of posting a demo, after taking another look at (2) primary goals of your OP for this thread, I am compelled to inquire: :scratchch


* What are the Geographic extents of a DEM you wish to make for your project area for the Mt. St. Helens 1980 pre-eruption area ?


* Additionally, I am compelled to ask whether you used a CVX vector 'Exclude and Replace' protocol for the Spirit Lodge flatten ?


IIRC, a FS terrain mesh surface should always be pre-empted by a CVX vector flatten placed in a Scenery Library Area layer above.

However, if you intend your own custom terrain mesh to be displayed so it pre-empts the OrbX FTX PNW terrain mesh, it may need to be compiled with Geographic extents that exceed those of:

[FSX install path]\ORBX\FTX_NA\FTX_NA_PNW07_MESH\scenery\4_mesh_multiLOD4-12_FTX-PNW_WA-S_v6.bgl


OrbX' Holger Sandmann stated terrain mesh BGL LOD extent may enable a higher priority load / display over 'smaller' extent mesh BGLs.


IIRC, this same FS display mechanism applies to CVX vector sloped flatten TINs, so gaining display priority over larger LOD extent mesh BGLs, requires an LOD extent of coverage for intended CVX BGLs ...to be larger than that of ex: OrbX' underlying terrain mesh BGL.

This can be done via a "trick" of placing isolated corner vertex data points at the outer extent of the desired "larger" (faux) coverage Geographic coordinates, so that a GIS "NO_DATA" value results for any area between those points and the actual intended smaller extent of vertex points of a data set to be displayed for the local area of one's CVX vector sloped flattens and/or other CVX vector objects.

Thus one need only supply point data for 'fake' larger extent corner vertices along with a 'true' smaller extent CVX object display area. :idea:

Since you seem to enjoy figuring out such GIS scenarios through your own research, I shall leave that hint as stated thus far, with an offer that if you are interested in learning more about how this 'trick' may be achieved, feel free to inquire further here. ;)

GaryGB
 
Last edited:
Messages
96
Country
us-washington
Hi Steve:

In lieu of posting a demo, after taking another look at (2) primary goals of your OP for this thread, I am compelled to inquire: :scratchch


* What are the Geographic extents of a DEM you wish to make for your project area for the Mt. St. Helens 1980 pre-eruption area ?


* Additionally, I am compelled to ask whether you used a CVX vector 'Exclude and Replace' protocol for the Spirit Lodge flatten ?


IIRC, a FS terrain mesh surface should always be pre-empted by a CVX vector flatten placed in a Scenery Library Area layer above.

However, if you intend your own custom terrain mesh to be displayed so it pre-empts the OrbX FTX PNW terrain mesh, it may need to be compiled with Geographic extents that exceed those of:

[FSX install path]\ORBX\FTX_NA\FTX_NA_PNW07_MESH\scenery\4_mesh_multiLOD4-12_FTX-PNW_WA-S_v6.bgl


OrbX' Holger Sandmann stated terrain mesh BGL LOD extent may enable a higher priority load / display over 'smaller' extent mesh BGLs.


IIRC, this same FS display mechanism applies to CVX vector sloped flatten TINs, so gaining display priority over larger LOD extent mesh BGLs, requires an LOD extent of coverage for intended CVX BGLs ...to be larger than that of ex: OrbX' underlying terrain mesh BGL.

This can be done via a "trick" of placing isolated corner vertex data points at the outer extent of the desired "larger" (faux) coverage Geographic coordinates, so that a GIS "NO_DATA" value results for any area between those points and the actual intended smaller extent of vertex points of a data set to be displayed for the local area of one's CVX vector sloped flattens and/or other CVX vector objects.

Thus one need only supply point data for 'fake' larger extent corner vertices along with a 'true' smaller extent CVX object display area. :idea:

Since you seem to enjoy figuring out such GIS scenarios through your own research, I shall leave that hint as stated thus far, with an offer that if you are interested in learning more about how this 'trick' may be achieved, feel free to inquire further here. ;)

GaryGB


I'll try to answer this as best as possible;

1.) The geographic extents of the coverage area of the DEM I am trying to create include Spirit Lake, the area west of Spirit Lake down to the current Sediment Retention Structure (e.g. the entire Toutle Valley down to Highway 504 mile marker 24), and the current area of the present-day locations of Coldwater Lake, South Coldwater Creek (north of Johnston Ridge) and Castle Lake. Coldwater and Castle Lakes did not exist prior to 1980, and the valley floor of north Coldwater Creek, where the present-day Coldwater Lake is, is almost 160 feet below the current surface of the lake. Prior to 1980, the valley the present-day Coldwater Lake sits in, was a classic, glacial-carved U-shaped valley with a pronounced V-shape to portions of the valley floor. This is most notable in areas near a giant lateral glacial moraine that forms the northeast boundary of the valley.

The five topographic quadrangles I have downloaded from the USGS TopoView utility are: Spirit Lake, Mount St. Helens, Elk Rock, Toutle, and Cougar. As there was largely no large-scale elevation changes on the South Fork Toutle River, I am opting to eliminate that as a coverage area.

2.) No. In fact I did not see the need to, merely based on the fact that the terrain in that area varied so wildly owing for current DEM in use. For the Spirit Lake Lodge, I used two separate, but neighboring flattens, to allow the lodge to reside in the terrain as realistically as possible, and have it appear as close to the real lodge as the SIM would allow.

With regards to the "figuring out" aspect... I'd love to say I do, but truly, I do not enjoy it, at this step of the game. I am finding that generating a workable DEM of pre-eruption elevation data from existing contours from raster scans of 1950s topographic maps, to be a task that is far more difficult and time consuming than it should be. Even more irritating is that the help I am receiving elsewhere (Reddit, Facebook GIS groups, and GIS web forums) have so far been disappointing. As a matter of fact, I was told on one Reddit reply that I should just "YouTube it... It's not that hard..."

(As the admin of five astronomy groups on Facebook, responses like that result in a ban, but hey...)

The coverage area I hope to accomplish, approximated by the red outline, is seen below.

CoverageArea.jpg



YouTube tutorials have NEVER worked for me.

At this step, I'd rather just as soon pay someone to do this, because it is turning out to be a giant headache.

With regards to the display of the custom mesh over ORBX's custom mesh... We solved that one when I compiled my pre-eruption DEM of Mount St. Helens itself in this thread: https://www.fsdeveloper.com/forum/t...h-display-after-compiling.444927/#post-818222

My goal in that is to create a mesh of identical LOD.
 

Attachments

  • CoverageArea.jpg
    CoverageArea.jpg
    490.8 KB · Views: 81
Messages
7,450
Country
us-illinois
Hi again:

What was the final resolution (LOD or Meters between elevation data points) of your DEM for the immediate Mt. St. Helens coverage area that you currently want to match in the area demarcated by the Red line above ? :scratchch


PS: USGS' TopoView historical scans seem to be distributed as low resolution so bad contour rasters cannot be converted to vectors. :banghead:

There are still high resolution topo rasters available via U of Washington if you do not have them on disk already. ;)

GaryGB
 
Last edited:
Messages
96
Country
us-washington
Hi again:

What was the final resolution (LOD or Meters between elevation data points) of your DEM for the immediate Mt. St. Helens coverage area that you currently want to match in the area demarcated by the Red line above ? :scratchch


PS: USGS' TopoView historical scans seem to be distributed as low resolution so bad contour rasters cannot be converted to vectors. :banghead:

There are still high resolution topo rasters available via U of Washington if you do not have them on disk already. ;)

GaryGB


The contour elevation interval on the USGS topos I have is 80 feet, which by my calculation is 24.38 meters between contours. With respect to the existing pre-1980 terrain of Mount St. Helens itself, that is a 30-meter DEM.

Do you have a link to the high res topo rasters of the area pre-1980? I have tried locating them on the UW Geospatial Data Archive but have been unsuccessful.
 
Messages
7,450
Country
us-illinois
IIRC, these are the linked hi-res topo rasters in GeoTiff format:




[EDITED]

FYI: I also found some hi-res (~1950's ?) topos here:


[END_EDIT]


BTW: One might fill in poly areas of this 3 M data set with over-sampled / SRTM 30 Meter elevation data to eliminate pits / spikes: :idea:


A wind-fall might be a work-flow to create pre-eruption westward extensions to fit into FSX / P3D default and/or FTX PNW mesh. :stirthepo

GaryGB
 
Last edited:
Messages
96
Country
us-washington
IIRC, these are the linked hi-res topo rasters in GeoTiff format:




BTW: One might fill in poly areas of this 3 M data set with over-sampled / SRTM 30 Meter elevation data to eliminate pits / spikes: :idea:


A wind-fall might be a work-flow to create pre-eruption Toutle westward extensions to fit into FSX / P3D default and/or FTX PNW mesh. :stirthepo

GaryGB


I note the pre-eruption topo DRG I see in one of those links is the same one I downloaded from TopoView at the same resolution. It is presently loaded into my SBuilderX project file as a base map for road reference. That same DRG topo is what I used to lay out Weyerhaeuser's 3500, 4000, 4020 and 4040 lines (each one of those played a significant role in the Mount St. Helens story, which I'll include bits of in the readme).

CoverageArea2.jpg


The contour interval is much greater than 30 meters between contours on that particular topo. In fact it is 50-meter. I suppose it wouldn't be too big a deal to attempt it in QGis, but then comes the issue of the fact that 50 meter mesh softens quite significantly, the pre-1980 contours of the end of the Floating Island Lava Flow as well as Dry Gulch, which was a canyon north of Mount St. Helens that terminated at the Toutle River near the county line boundaries of Skamanaia and Cowlitz County. In fact, today's drainage canyon in the landslide deposit out of the crater follows much of the old Dry Gulch contour. One thing I've thought about is softening the terrain using 30 meter mesh of the landslide deposits and trimming it to allow the higher mesh of Johnston Ridge to show through, but that in itself poses issues since QGis, for some reason, won't allow lasso-based canvas clipping instead of having to select areas by rectangle, which currently seems to be the only way QGis supports any sort of extraction/DEM trimming.

I am going to try attempting a few things tonight, before taking a break on the terrain aspects.

Last night, I achieved a significant milestone by converting all of the unassigned ESRI lines to 10-meter width dirt roads in one fell swoop. This effectively completed a huge chunk of the project by finishing all of Weyerhaeuser's pre-1980 roads. An epiphany came via a test import of ESRI line data and automatically assigning the imported lines to one single GUID call - the Dirt 1 Lane Divided Median road type with a 10-meter width. It required eliminating huge chunks of data and eliminating imported duplicates, then subsequent deletion of roads not matching 1979 Weyerhaeuser road maps in my possession (and the 1979 aerials I have), to complete that part of it.

If only that DEM exists... If only...
 
Messages
7,450
Country
us-illinois
Indeed, if USGS can find the pre-1980 eruption DEM, that might reduce your workload considerably.

As for derivation of DEM point data from raster to vector conversion, my reference to "resolution" was a misnomer, since I should properly have indicated that the TopoView images appeared so badly compressed, that raster conversion to vector data would not be practical.

I anticipated that the U of Washington or other USGS National Map (or Earth Explorer ?) topo imagery might prove to be a better quality image source for the raster to vector conversion process.

BTW: Gridded elevation data is likely always in quad arrays, but one may be able to use irregular Polygon objects to assign a GIS NO_DATA value to points outside the intended visible DEM area that are otherwise within the outer perimeter of the quad array of the gridded point data set, when a 32-Bit elevation data GeoTiff is derived from the point data.



GaryGB
 
Last edited:
Messages
96
Country
us-washington
Indeed, if USGS can find the pre-1980 eruption DEM, that might reduce your workload considerably.

As for derivation of DEM point data from raster to vector conversion, my reference to "resolution" was a misnomer, since I should properly have indicated that the TopoView images appeared so badly compressed, that raster conversion to vector data would not be practical.

I anticipated that the U of Washington or other USGS National Map (or Earth Explorer ?) topo imagery might prove to be a better quality image source for the raster to vector conversion process.

BTW: Gridded elevation data is likely always in quad arrays, but one may be able to use irregular Polygon objects to assign a GIS NO_DATA value to points outside the intended visible DEM area that are otherwise within the outer perimeter of the quad array of the gridded point dara set.

GaryGB
It seems as though I've had some success (?) since the last post.

Over on Reddit, a generous user there gave me very valuable (and to my understanding) instructions on what to do, to generate the DEM I need. I am currently experimenting with several open windows in QGis to save contour data as a vector file, with attachment attributes denoting contour elevation to each line.

In one of those windows, I revisited this grayscale GIF I sourced from Malin Space Science Systems' website.

msh_pre.gif



It is the pre-eruption DEM the USGS authored in 1983, when the implimentation of digital elevation models in geologic studies became standard. In fact, this DEM was produced as the first DEM the USGS had created as part of the new DEM standard. This DEM was also used in a 1986 study in which the founder of Malin Space Science Systems was one of its coauthors. That study highlighted pre-and post-eruption elevation changes brought on by the eruption.

(https://www.msss.com/earth/msh/msh.html)

I am currently in the process of experimenting with it and much to my surprise, I was able to extract all of the contour data out of this.

Even more to my surprise, I am able to edit the elevation points to those contour lines individually.

Since it appears that this was generated from seven 7.5-mintue-x-15-degree quads, I am going to attempt to georeference this first using the coordinates of existing quads in the area, then begin the process of experimenting with DEM creation. Before that begins, and assuming I have properly georeferenced this, I plan on greatly trimming this down to several smaller areas, so as not to create an obscene workload.

We shall find out if I am successful!
 
Messages
7,450
Country
us-illinois
Congratulations on locating the USGS DEM; that looks like a promising work-flow. :)

GaryGB
 
Messages
96
Country
us-washington
Congratulations on locating the USGS DEM; that looks like a promising work-flow. :)

GaryGB
Well, it's not quite a DEM, but a grayscale GIF of one turned into a TIFF, which I am hoping I can georeference, so as to begin contour extraction and editing.

Baby steps here, for sure.
 
Messages
96
Country
us-washington
Congratulations on locating the USGS DEM; that looks like a promising work-flow. :)

GaryGB

Tonight was a night of celebration. Indeed, I actually did pop open a bottle of champagne and had a toast to myself...

For a test, earlier this afternoon I decided to trace contours in a small area going up from Spirit Lake's pre-eruption elevation contour of 3,198 feet. For reference, today's lake elevation is 3,406 feet! Once that was done, I then proceeded to trace every contour on a 1958 topographic map, using a vector layer, with each contour assigned an elevation point corresponding to the 80-foot contour interval of the source reference map. (It was a PAIN IN THE ARSE since I had to convert the numbers to meters every time. The small area I chose was the dividing ridge between the east and west lobes of Spirit Lake.

After tracing the final contour in QGis, I then imported it into a geospatial data processing application called SAGA, to which I then did a triangulating extrapolation of that contour data into an digital elevation model. That elevation model's dataset was then loaded back into QGis as a DEM file, and subsequently exported
as a GeoTIFF.

Once a brief test was initiated in FSX, I was ASTOUNDED at the height difference between present-day Spirit Lake and the former shoreline. It shows up best at the lakeshore (where I had placed an exclusion flatten of the lake to eliminate scenery artifiacting) on the dividing ridge, and where Harry Truman's resort is.
Now the fun part begins... Tracing a crap-ton of contour data, then extracting it into height map data, and then finally, off to the SIM.

Now for a few images showing the sequence of steps I followed, then a series of screenshots in FSX showing the change in topography.

I am beyond over the moon over this achievement.
isthissuccess1.jpg
isthissuccess2.jpg
isthissuccess3.jpg
isthissuccess4.jpg
isthissuccess5.jpg
isthissuccess6.jpg
isthissuccess7.jpg
isthissuccess8.jpg
isthissuccess9.jpg
 
Last edited:
Messages
96
Country
us-washington
Top