P3D v4 Ground poly base image vs ground poly airport

[Colin Gray]

Good evening.

Not really sure the best way to describe what I’m trying to say. I have put together a ground poly of runways, taxiways, and aprons (airport diagram). I have pretty much figured that out - for now anyway. I also have an image that I want to use that covers the immediate area around the airport. It’s that image that I’m trying to figure out how to use correctly - we’ll call this the base image.

Using MCX,s ground poly wizard, I can set up the airport diagram at the airport elevation. Using z-bias, I separate the taxi lines and surface textures so it looks proper.

Rather than deleting all the asphalt and concrete surfaces on the base image, I was hoping to place the base image under the airport diagram. If I use the ground poly wizard to place the base image (using the airport’s altitude), what altitude should I use for the airport diagram? I’m assuming it should be placed higher than the base image. I tried this, but I was getting some bleed through. I could see the base image through the airport diagram ground poly. I set the altitude at .005 meters above and it didn’t seem to help. I’m trying to avoid having the plane look like it’s sinking into the scenery.

Thank you in advance. Not sure if I made any sense.

Cheers,

Colin

 

[GaryGB]

Hello:

Considering your discussion in this prior thread:

https://www.fsdeveloper.com/forum/t...ayers-for-ground-polygons.458240/#post-922769


...I am curious why you seem to be attempting to use BGLComp XML methods to place XtoMDL type G-Polys at differing assigned Altitudes ?


P3Dv4.x is able to assign Z-Bias Material Functionality 'layer numbers' to control placement of G-Polys relative to co-planar scenery objects.

Thus, although assigned Altitude may be identical for 2 or more objects, Z-Bias enables objects to display at "offsets" from (1) Altitude internally.


The more negative a Z-Bias number, the less of a distance that layer is displayed from user aircraft camera position (top layer = most negative Z-Bias).

https://www.google.com/search?q=site:+www.fsdeveloper.com++P3D+Z-Bias+functionality&sca_esv=a95ded86ae4b46ea&ei=QvaJaa6QHIbUp84Pz_G1qAo&biw=1088&bih=504&ved=0ahUKEwiuuf6O1sySAxUG6skDHc94DaUQ4dUDCBM&uact=5&oq=site:+www.fsdeveloper.com++P3D+Z-Bias+functionality&gs_lp=Egxnd3Mtd2l6LXNlcnAiM3NpdGU6IHd3dy5mc2RldmVsb3Blci5jb20gIFAzRCBaLUJpYXMgZnVuY3Rpb25hbGl0eUihMVDtAVjtAXABeACQAQCYAckBoAHJAaoBAzItMbgBA8gBAPgBAZgCAKACAJgDAIgGAZIHAKAHZ7IHALgHAMIHAMgHAIAIAA&sclient=gws-wiz-serp


Note as well that according to "toprob", Z-Bias 'layering' has been changed further in latter versions of P3Dv4 to accommodate PBR texturing:

https://www.prepared3d.com/forum/viewtopic.php?t=131953


P3Dv5.x adds further complexity to use of this feature via "Enhanced Properties":

https://prepar3d.com/forum/viewtopic.php?t=144488

https://www.prepar3d.com/SDKv5/sdk/modeling/3ds_max/modeling/PBR_Enhanced_Properties



BTW: AFAIK, SCASM / ASM methods for non-MDL-based G-Polys / VTP placement also still works in P3Dv4.x.


Most blending is via Alpha transparency in the diffuse texture image for 1 or more G-Polys, although MCX can assign more esoteric methods.

https://www.prepar3d.com/SDK/Environment Kit/Modeling SDK/FSXMaterial.html


https://www.fsdeveloper.com/forum/threads/pbr-decal-order-disappearing-layers-based-on-zbias.447652/


There are complexities in blending 2 objects that use Alpha transparency for use in FS2Kx versions of FS:

https://www.fsdeveloper.com/forum/t...tain-material-with-alpha-transparency.446555/



Regarding your "Base Image":

IIUC, that is aerial / satellite imagery as a form of SDK Resample custom photo-real land class ?

Or is that imagery applied to another SDK XtoMDL G-Poly as a diffuse texture Material ?


If it is aerial / satellite imagery as a form of Resampled custom photo-real land class, it covers up all other SDK SHP2VEC land class textures.

This cannot be controlled by VTP layering or Z-Bias Functionality.


Airport infrastructure default BGLComp AFD textured objects such as RWYs, Taxiways, Aprons display on top of custom photo-real land class.

This also cannot be controlled by VTP layering or Z-Bias Functionality.


If all your imagery is texture Materials on XtoMDL G-Polys, you will have issues with flickering that may be difficult to eliminate.


P3Dv4.x can display aerial / satellite imagery as a form of SDK Resample custom photo-real land class at very high resolutions.

That type of imagery will drape directly onto local terrain mesh surfaces but display under AFD type airport infrastructure.

But, latter P3D versions have the ability to turn off display of the textures for such AFD infrastructure objects; ADE can toggle this On/ Off for us.


Otherwise, you may instead opt to make your G-Polys as SCASM / ASM legacy objects in order to control VTP layers and eliminate flickering better.


PS: Which airport ICAO is this project for ?


GaryGB

 

[Colin Gray]

@GaryGB thank you for the response. I wish there was a method to my madness. Most of what I do I copied from YouTube tutorials. Bill Womack has a series of videos on how to make airport sceneries. They are a few years old, but easy to follow. They do miss out on some steps, but through trial and error I found a way to make get a rudimentary airport diagram. I’m sure there are better, more efficient ways to do it, but I’m happy how it looks. Needless to say, my standards aren’t all that high. As I get more experience, I’m sure that’ll change.

A little background. Sim of choice: P3D v4.5. I’m working on Honolulu. Probably should’ve tried something smaller. I’ve ‘worked’ on other airports in the past (the link from a previous post was for CYWG), but eventually something distracted me and I’ve moved on to something else. I’ve been at this since mid December. Had to restart a couple times, but I’m kind of on a roll now. I’d say I’m about 60% done with the ground layout. So here is what I have done:

1) I have a very basic AFCAD drawn up with ADE. Just taxi lines, parking, and runways. This was put together using a satellite image from Earth Tiles.

2) using the same FSET image, I switch to Blender to draw the surfaces and surface markings. Everything I’ve drawn is set to an altitude of 0. I don’t separate the layers until I export it to MCX. I export it as a .DAE file which I view with MCX. Using material editor, the base asphalt has a z-bias of -4, concrete sections -8, and taxi lines -12. I’ll probably refine this a little as I go. These are the steps from Bill’s tutorials. I’m not doing any PBR textures or UV baking. Both of those are above my skill level.

3) I’m trying to place a larger ground poly between the AFCAD and airport diagram. I’m taking the same FSET satellite image and using both photoshop and sketch up to make a 2d image and then converting it to a BGL. I’d then use the ground poly wizard (MCX) to place it into the sim.

If I understand what you’re saying, this may not be as simple as I thought. I’m not using anything from the P3D SDK directly. I think MCX use the bgl compiler and maybe Resample as well. I’ve used Resample to view other ground polys, but that’s it. Can I use to to make them as well? I’ll look into that. I tried inserting a plane (back to Blender) with the FSET image, but I got a message saying the texture was the wrong size. I also tried to break it down into 1024x1024 images, but struggled lining each image up in Blender.

 

[GaryGB]

Hi Colin:

So, IIUC, you are working on PHNL ?

And you have AFD infrastructure made in ADE intended for display not via default FS textures, but via aerial / satellite imagery on a G-Poly ?


Arno's MCX does not use SDK Resample (Arno's ScenProc does); so you are working strictly with image textures mapped on flat 3D G-Polys.


IMHO, SDK Resample custom photo-real land class is easier to make than G-Polys.

You can also use just about any imagery via SDK Resample to make custom photo-real land class, rather than just aerial / satellite photos, airport diagrams from FAA or Jeppesen, maps, topos etc..


But working with G-Polys is popular for airports, as the airport infrastructure flat 3D models can display some extra environmental attributes such as wet-in-rain etc.


G-Polys and SDK Resample custom photo-real land class draped directly onto local terrain mesh surfaces both display high resolution imagery.


MCX G-Poly Wizard can cut G-Polys into 100 Meter segments that fit onto the curved Earth surface in FS and places them using BGLComp.


SDK Resample automatically maps and drapes the imagery as MIPMAPs onto multiple LODs of terrain mesh 'tiles' that are also "placed".


I am not too familiar with Blender, but AFAIK it has tools and/or plugins that can cut texture mapped 2D planes / Faces into precise 'tiles'.

In Sketchup, one can "drape" a aerial / satellite or other image onto the flat surface of a plane or Face and then do the cutting into 'tiles.


Photoshop 'could' cut images into ex: 1024x1024 pixel tiles, but to manually map them onto the flat surface of many planes or Faces is very laborious.

For that one can use either Blender or Sketchup directly.


You may have used SDK TMFViewer to view SDK Resample custom photo-real land class BGLs, but not G-Polys.

We could use MCX to output G-Polys, but that process would not use SDK Resample.

IMHO, it would be easier to cover larger scenery areas with aerial / satellite imagery by using SDK Resample to make custom photo-real land class BGLs.

How large of an area around PHNL do you want to render in P3D (for your personal use ?) if you tested making custom photo-real land class ?

GaryGB

 

[Colin Gray]

Yes, PHNL - Honolulu. Forgot that an airforce brace is also part of it. A little more work to do, but I’m getting very good at making taxi lines.

Airport layout has changed a little bit since FSX/P3D, just trying to bring it up to date. Textures are custom - not default.

I don’t know if this is considered big or not, but the ground poly I’m using is 7km x 4km. Just a little bit larger than the airport’s footprint. Presently, I have no intention of environmental effects. That may change later, but I’m just trying to learn how this all works.

Thanks.

 

[GaryGB]

What are the pixel count-by-pixel count dimensions of the FSET aerial / satellite source imagery downloaded for your PHNL - Honolulu ?

7 km x 4 km is a good sized airport area in FS; but the resolution of the imagery mapped over that area would determine its visual quality.

That would be true whether the imagery is mapped onto G-Polys or draped onto terrain tiles via custom photo-real land class BGLs.


It may be that as you have no immediate goal of using environmental effects on G-Polys, custom photo-real land class BGLs is the way to go.

If you have an interest, I can offer additional info on how to do this as an optional alternative to save some work.

GaryGB

 

[Colin Gray]

Pixel count for my satellite image: 8930px x 5080px.

If you have an optional alternative and the time, I’m open to any suggestions.

Thanks Gary.

 

[Tim Collins]

From my experience, an image of that pixel count for an area of that size is pretty low resolution, even though the file size will be quite large (depending on the quality you selected). You could have picked a smaller airfield as a learning project! However, FSET will certainly cope with a large area and will manage the splitting of the image into manageable tiles within its own algorithm. This is different to the tiles that Gary's referring to, but has the same purpose of managing image size. Using FSET and ReSample gives a fairly easy background image management workflow. You'll also need Gimp, Affinity or Photoshop to crop the image to the airfield boundaries and avoid it affecting autogen outside the airfield.

Your GP-based airport should sit above the Resample image, which will (as Gary says) automatically map to the terrain. You can use FSET to generate watermasks if you wish, but I find it easier to simply make a blendmask in Gimp or Affinity that ensures the image generated is only visible within the airfield boundary. This process is described well in the SBX tutorial which I'm sure you've already found.

 

[GaryGB]

Some practical considerations are in order here.

There are already freeware custom photo-real land class BGLs for FSX / P3D for the entire Hawaiian Island chain at LOD-15 (1.2 meters / pixel).

https://flyawaysimulation.com/downloads/files/24780/fsx-hawaii-complete-photoreal-airports-scenery/

https://www.rikoooo.com/downloads/viewdownload/68/730


IMHO, that would make it impractical to reinvent the wheel as resolution for the OP's own scenery calculates to ~LOD-15 (0.78 Meters / pixel).

I don’t know if this is considered big or not, but the ground poly I’m using is 7km x 4km.

Pixel count for my satellite image: 8930px x 5080px.

7,000 Meters / 8,930 Pixels = 0.78387458006718924972004479283315 Meters / Pixel

4,000 Meters / 5,080 Pixels = 0.78740157480314960629921259842520 Meters / Pixel


When compiled from either MS-VE / BING or Google Satellite imagery (currently identical in SBuilderX' tileserver), both will likely yield LOD-15.

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


Photo-real LOD decreases as Altitude AGL increases, so IIUC, a primary focus 'should' be making high resolution airport infrastructure objects.


High resolution airport infrastructure objects will only display ultra-sharp LOD texture Material when the user aircraft is at / near ground level.


One can use LOD-18 or LOD-19 MS-VE / BING or Google Satellite imagery (currently identical in SBuilderX' tileserver) for airport infrastructure.

One can also use LOD-19 or oversampled LOD-21 MS-VE / BING Satellite imagery via SASPlanet in GeoTIFF file format for airport infrastructure.


Thus, IMHO, custom photo-real land class is preferable unless OP wants environmental effects on RWYS, Taxiways, Aprons, Ground Markings.

If so, the workflow requires flat 3D G-Polys with all the headaches of layering and transparency, as well as suppression of flickering / shadows,


The OP still can superimpose opaque ultra-sharp synthetic RWYS, Taxiways, Aprons, Ground Markings on top of Custom photo-real land class.

While one 'could' make one's own ultra-sharp Ground Markings in Photoshop and Sketchup, ADE has tools by Don Grovestine to do this too.


One only needs (1) ultra-sharp layer for Ground Markings RWYS, Taxiways, Aprons, Ground Markings on top of Custom photo-real land class.

That could even be merged into the top of Custom photo-real land class, but some may prefer to keep it a separate and truly ultra-sharp layer.


AFAIK, that ultra-sharp layer for Ground Markings RWYS, Taxiways, Aprons, Ground Markings will not need any transparency attributes.

AFAIK, that ultra-sharp layer for Ground Markings RWYS, Taxiways, Aprons, Ground Markings will not need any flickering / shadow suppression.


I believe the above observations / suggestions could help simplify the OP's task considerably.


GaryGB

 

[Tim Collins]

That’s a great point, Gary - absolutely no point in doing any photo-real outside the airfield.

Don’s GP editor inside ADE is perfectly capable of making the polys needed, but ADE uses vertical separation to try and make all the layers visible. It’s worth using MCX to apply Z-bias so that all of the polies are co-planar. Chat GPT will talk you through the process. It’s probably the only thing that’s read and digested everything Gary’s ever said and can regurgitate it in a way ordinary brains can understand!!

 

[GaryGB]

For those that insist on having every aspect of an airport scenery represented in ADE's GUI, Don's Tool facilitates that goal.

When necessary, I de-compile such ADE-GP BGLs, work around the error messages (if any), and then import them into MCX' G-Poly Wizard.


PS: I like the idea ordinary brains may understand what ChatGPT does not (...I'm not convinced AI is "conscious" or able to 'comprehend').

https://www.google.com/search?q=reg...0DwgcHMC4xLjUuMcgHGoAIAQ&sclient=gws-wiz-serp


A virtual ‘whirlpool’ of information (...and not to be confused with a "torrent" of information).

GaryGB

 

[Colin Gray]

@Tim Collins I definitely could’ve picked an easier airport for my first attempt. I thought using the same image I used in Blender would fit better. It isn’t the best quality. Is there a better source for satellite images?

Two questions for you. First, my GP-based airport should sit above the Resample image. What should I set the altitude at for the Resample image? PHNL is at 4m. Do I set different altitudes for both Resample and my GP-based airport?

Second question - which SBX tutorial do you mean? I’ve watched one by Soarfly, but he uses both FSET and SketchUp.

Thanks for your help, Tim. Not sure why I picked this airport to try. Definitely a steep learning curve, but I’m enjoying the process.

Cheers,

Colin

 

[Colin Gray]

@GaryGB this is going to make me look bad, but I had no idea there were tools inside ADE. I mean I know I can change airport reference data, but I wasn’t aware Don had tools in ADE as well. I know I can import images with the coordinates from FSET and SBX.

Your reply (#9) has given me a lot to think about. To be honest, a lot of it was over my head. I know that’s all part of learning something new, so all good.

The OP still can superimpose opaque ultra-sharp synthetic RWYS, Taxiways, Aprons, Ground Markings on top of Custom photo-real land class.

By this do you mean the work I’m doing with Blender?

 

[GaryGB]

@GaryGB this is going to make me look bad, but I had no idea there were tools inside ADE. I mean I know I can change airport reference data, but I wasn’t aware Don had tools in ADE as well. I know I can import images with the coordinates from FSET and SBX.

Your reply (#9) has given me a lot to think about. To be honest, a lot of it was over my head. I know that’s all part of learning something new, so all good.

Prior to his untimely passing a couple of years ago (RIP), Don was quite prolific in his creation of FS utilities, including some for ADE.

IIUC, ADE author (Jon Masterson) incorporated a few of Don's utilities tailored specifically to work with ADE some years ago, so they are relatively recent.


A review of my numerous posts will show that not only do I care about helping folks, but I do attempt to translate and explain things for others here.

But I confess I have emphasized using the most commonly used terms (which may be technical) by which they are most often known / found at FSDEV.

That terminology (most not originated by me) is well-intended, but probably seen as "obscure" by many that are new to FSDEV and the FS SDKs.


I learn something new- and humbling- nearly every day from my FS endeavors and reviewing the FSDEV forums, so we are all "beginners" IMHO.

By this do you mean the work I’m doing with Blender?

Yes, there are apparently other sources of MS-VE / BING imagery which are higher resolution in some world areas that can be used for your G-Polys.

SASPlanet is one such source that can eliminate some of the tasks performed by parts of the FSET work-flow to get a higher resolution imagery GeoTIFF.

https://sasplanet.geojamal.com/2022/01/archive-of-sasplanet-stable-versions.html

I use SASPlanet version 200606: http://www.sasgis.org/programs/sasplanet/SASPlanet_200606.zip


Output of "ADE-GP" can be used as a stand-alone layer, or can be 'merged' with SASPlanet higher resolution aerial / satellite imagery if desired.

GaryGB