• 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.

Ortho has ruined my scenery

Messages
539
Country
us-missouri
Once I had P3D looking good with high-resolution terrain mesh, accurate water and road vectors, millions of buildings, and even vegetation shapefiles, I made the mistake of adding golf courses and quarries from ortho scenery.

Now everything else looks ridiculous.

I wonder if there is a more efficient process for importing ortho scenery on a mass scale. Currently my workflow is:

1) download 3.75-minute GEOJSON tiles from USGS server (20-30 megabytes)
2) import into QGIS, convert to WGS84, export to GEOTIF
3) convert GEOTIF to INF
4) bump up contrast of GEOTIF in PaintShopPro, save (now about 250 megabytes)
5) resample INF file to BGL (example code below)

[Source]
Type=TIFF
Layer=Imagery
SourceDir="."
SourceFile="3708952 NW.tif"
PixelIsPoint=0
ulxMap=-89.625026804
ulyMap=37.250027561
xDim=8.98315278575456E-06
yDim=7.15357097106259E-06
Variation=Day
[Destination]
DestDir="."
DestBaseFileName="3708952 NW"
LOD=Auto

I haven't begun dealing with watermasks yet, which would complicate the process, but I think I can somehow import my water shapefiles into ScenProc, so I am not worrying about that right now.

Granted, there are better sources out there, but USGS data belongs to us taxpayers and (unlike Google) I can do whatever I want with it. It's gorgeous to me. But great day, this process will take forever if I want to cover much ground (literally). And one degree of scenery would take many gigabytes. Is there a better way?

Thanks,
 
Once I release the photo real features of scenProc you might be able to use it. But I don't think it will be much quicker, as processing lots of image data is just a complex task that takes a lot of diskspace and time.
 
Arno, I think it won't be long before you'll need to compile a list of things that ScenProc DOESN'T do. Great program, and I am currently learning how to add trees to go with the autogen buildings that look incredible on P3D.

By the way, is it possible for ScenProc to generate complex shapes like those found in dormitories, shopping malls and nursing homes? Starting small, reading the manual, and helpful posts has got me to where I can generate square or rectangular buildings, but it will not follow the more complicated footprints. It would make sense since this is autogen that it won't, but I'd like to know if I can accomplish more. I would include the config I wrote, but apparently I wiped out all my saved files when I updated from the stable release to the latest development release.

Best regards,
 
I recently registered for an account at USGS Earth Explorer and found a high-resolution option for the 3.75-minute ortho tiles (approximately ten times larger than those available at the USGS National Map). When I load the new, larger file into QGIS for conversion to WGS84 and resample using the same steps I did in my original post above, I end up with a significantly smaller BGL filesize than the BGL file generated from the smaller files.

Also, now when I go through the process of adding photoreal scenery like I did in the original post, it does not show up in the sim. I double-checked the INF file, verified the BGL in TMFViewer, and examined the scenery library hierarchy and can't find any issues; Ortho scenery worked for me before but not now. Does anyone have any ideas?
 
Hi,

Do the images maybe have a 4th band that is interpreted as a blend mask?
 
Good catch; According to the TXT file that accompanied the 159 megabyte file and it appears so:
"32-bit pixels, 4 band color(RGBIR) values 0 - 255"

But I thought QGIS exported as a three-channel RGB GeoTIFF. Verifying that as well.

Another variable is that I generate an INF file from the exported GeoTIFF. Then I load the TIF file into PaintShop Pro where I crank up the vibrancy and contrast and lower the brightness. PaintShop then saves as a merged image; would that mean that the channels are combined as well? I then load the INF file into Resample and in the past this has given me pretty decent photoreal scenery. Perhaps the Missouri and Florida 3.75 tiles don't have the fourth IR channel while California and Arizona tiles do. Trying to find out more.
 
Last edited:
I also noticed that the much larger TIF file from Earth Explorer, while it looks more sharp than the smaller JP2 file from the USGS National Map site, actually has less pixels. 6390 x 7670, 187 mb TIF compared to 6964 x 8603, 23 mb JP2.

I was going along nicely, now I feel like I've drifted into the deep end of the FS design pool and am in over my head.
 
Hi,

Those jp2 images have a lot of compression, so comparing only file size does not say a lot.

The fourth band is infrared, so you would have to drop that one.
 
Hello:

As Arno has pointed out previously, having a 4th NIR band is good for Autogen vegetation detection with ScenProc. ;)

But it is best to remove that band for SDK Resample imagery BGLs.

A discussion is linked below on doing that with Jim Robinson's method using GDAL via command line strings; IIUC, QGIS should be able to do this via a GUI.

Also, I posted a work-flow using the Global Mapper GUI.

And Dick posted an interesting command line string batch method using GDAL as well.


Hope this helps. :)


GaryGB
 
Last edited:
Your posts are generally clear as mud to me --at first -- until I start digging into each element of what you bring up. Then as I tackle things one at a time, your helpful posts become crystal clear sir. Just takes a bit to learn and then grasp the various programs, codes, and what-not.

As Arno has pointed out previously, having a 4th NIR band is good for Autogen vegetation detection with ScenProc.

Yes, I was excited to see that IR band for that very reason as I am ready for good autogen trees once I can get my ortho scenery showing up again.

Thanks!
 
That said, the clear-as-mud posts make the ones that have something in them that I already know make me want to do backflips of joy down the hallway.

And Arno's program does so much that I am wondering if the next development release will help me understand women. But seriously, I am really looking forward to ScenProc's ortho features!
 
After reading everything Gary sent and whatever else I could find on the subject I remembered that I had originally installed P3D in the Program Files folder by accident. I uninstalled and reinstalled everything and my ortho scenery showed up! Not sure what was wrong, but "unplugging it and plugging it back in" fixed the issue.
 
Back
Top