• 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 Converting flying/ai aircraft to scenery objects, can I reduce poly count?

Messages
52
Country
us-pennsylvania
are there any utilities to reduce the poly/mesh count of a model to make it more suitable for a static scenery object? I'm new to making airports/scenery in FSX, so I have no idea if this is possible. I have searched, but haven't found anything useful, so I'm sorry if this subject has been previously discussed.
 
Are you using Blender? If so, look up information about the "Decimate Modifier". The truth though is that your results will vary wildly.
 
The LOD tool in MCX is a polygon removal algorithm. Suggest you start with "Grid Size=.02" and incrementally work your way up in grid size until it is removing enough polygons without overly distorting geometry. Bear in mind that as LOD values diminish, resolution also diminishes, therefore excessive distortion on lower LOD models is effectively invisible.
 
The LOD tool in MCX is a polygon removal algorithm. Suggest you start with "Grid Size=.02" and incrementally work your way up in grid size until it is removing enough polygons without overly distorting geometry. Bear in mind that as LOD values diminish, resolution also diminishes, therefore excessive distortion on lower LOD models is effectively invisible.
Thank you Sir! Between the LOD and using the Object Hierarchy to remove anything that wasn't visible, the size has been reduced by just about half! I found an LOD of 40 and a grid size of 0.2 worked best for me
 
OK well you're almost there and glad to hear it works well for you. It is actually a pretty powerful tool and I had not encountered better, until I discovered the one in 3ds Max will do these reductions symmetrically, if desired.
Ultimately you would want 3 or more LODS, the numbers roughly correspond to size on screen and 100 is the most common maximum LOD, although I have a spawnable hangar with moving door that works quite well with 400 being the maximum LOD, obviously from inside the model. Right now, your model is at full resolution from LOD size 100 through LOD 40 where it suddenly changes. It seems like as soon as one was not standing directly next to the model, it could be less complex and not have to wait until it was less than half its original size on the screen, yes?

You have full control over the polygon reduction algorithm. The algorithm will present automatic LOD steps, which IMO, are somewhat arbitrary. FYI, I never, ever accept the recommended LOD levels, until the model is below LOD 40. So I start with LOD 100 and .02 on the grid size. If the model is over 10k polygons, I will want my first LOD under 100, to be close to 10K and I will continue to discard LOD passes, until the model gets there. Once set, I visually examine the result to make sure it does not need more polygons and then, whatever MCX has set for that LOD, probably 40, I use Change LOD to make it 80.

For the scenery addons I upload, I interpret where my models will be in relation to population centers and flight paths. If I believe people that use my airports, also fly over them at much higher altitudes, I will try to make all models such that the large visible ones are still visible, but low in poly count and the ones that would be invisible, are just a triangle or two, so they do not burden the render. To be clear, my scenery models usually have 5 or more LOD's, starting at 100, 80, 40, 16 to 6.

Finally, a note about Grid size. Grid size is like the net you cast to catch fish. If the model is an immense hangar constructed with 8 polygons, you will get no reduction whatsoever until you set grid size to 100, or 500, or 1000, the number corresponds to meters, I think and once you do get the "net" big enough to affect an 8 polygon hangar, you will learn there are no polygons that can be removed without significantly distorting the model, which shouldn't matter, it was only 8 polys to start.
Going the other direction, if you have a very detailed model of a cell phone, grid size .2 will obliterate the thing, I hope for obvious reasons Another FYI, .01 is the smallest useable Grid size.

The other settings, Angle weight and Area weight, discriminate the render on a different basis, in that small angle polygons, meaning they are almost sticks, likely do not comprise major portions of a models geometry and the same with area, individual polygons of extremely small area are either composing an elegantly curved surface, which the sim could represent other ways, or they are completely useless and should be swept away.
Floating cell is the alternative to fixed point calculations and while one or the other may have a specific, direct affect on quality, it is to me another setting to experiment with. Ground clamp attempts to keep all iterations of the original model, within reason, the original distance from the x/y plane.
 
Since you are working with FSX, you may wish to do (2) other things before reducing complexity of the overall 3D model by lowering LODs and decimation.

1.) MCX has the ability to load the exterior 3D model without the interior cockpit; some developers omit that A/C interior when making low-poly static A/C.

https://www.google.com/search?q=site:www.fsdeveloper.com+MCX+static+aircraft&ei=oVQoZKDPMoyfptQPspu1MA&ved=0ahUKEwjgo4mrhon-AhWMj4kEHbJNDQYQ4dUDCA8&uact=5&oq=site:www.fsdeveloper.com+MCX+static+aircraft&gs_lcp=Cgxnd3Mtd2l6LXNlcnAQA0oECEEYAVDVCVjPnAFgjaIBaAZwAHgAgAHUAYgBxA6SAQYyMS4wLjGYAQCgAQHAAQE&sclient=gws-wiz-serp

2. SAMM removes A/C cockpit Gauge code while also reducing some other components from A/C 3D models.

https://www.fsdeveloper.com/forum/forums/static-aircraft-model-maker.114/

GaryGB
 
Last edited:
Thanks for the info guys, it's all really helpful and is helping me learn quite a bit.

RK - I'm guessing FSX selects a LOD based on distance from the object? If that is so, is there a 'rule of thumb' for what the LOD should be for a given distance range?
 
It's supposedly based on the number of pixels it displays on the screen. Thus the LOD number depends on the size of the object. So we cannot give out generic number vs distance values. For typical AI aircraft 100-200 is good for the detailed model, then 50, 25, 10, and 5 for further away.
 
Thanks Tom! LOD, as in Level of Detail, is based on "pixels occupied." It is calculated differently between FSX and MSFS, btw, MSFS is percentage based. Anyway, there is an exact formula somewhere, but I don't think it's necessary. Just experiment around, you don't want to see your models change, if you can help it. LOD swapping is nothing new, you can see it in old episodes of Star Trek where they use little models and even colored cubes to represent much more detailed structures and in many video games, where the models seem to "pop" into greater detail as they get bigger.

old_k7_viewscreen.jpg


The goal, I think, is to hide these transitions behind the greater veil of the simulation resolution. If each model switches to a higher detail, just before that (level of) detail becomes visible, the switch should also be invisible, right up until the point of view is directly in front of the model.
 
Back
Top