Ok NV3 i'll ask again, any plans regarding vegetation ? .

If it was easy, we would have done it already. :)

Things to consider - engine support, animation, software license fees, redistribution license (many tree packages won't allow distribution in games), finding which package is best, figuring out how many people would actually make the trees, how to avoid duplication (like we have with 1,000 "ballast" textures.

If someone wants to evaluate all the options and licenses out there, we're happy to consider providing some resources to assist.
 
If you will just take a look at PlantFactory through the link above all those questions are answered.

Yes, commercial license covers distribution in games and DLC.
Yes, some animation is supported and more in development.
Yes, it supports many game engines.
The software also generates 5 levels of LOD which are fully customizable.
The software can produce three levels of detail in trees and bushes. Hero, which is high detail, HD which is very detailed for the CGI industry and RT which is for real time for games. Max of 50,000 polys and variants can be produced for seasons.

Free trial download available and E-on Software is a small company in France and not some huge corporation.

And best of all for just $3 a day. Less than a double latte mocha frappe. :)
 
Last edited:
From my reading of the PlantFactory information web pages, the less than $3 per day would be for the Enterprise Solution ($999 for an annual subscription). It has a slightly more expensive monthly subscription of $99 per month with a minimum commitment of 1 year.

That would really put it out of my price comfort zone unless I decided to produce all my plants as payware (and my skills are not that good).
 
That is true and that would be the one that N3V would need to own in order to distribute the plants on a commercial basis. However the enterprise level includes 25 "rendernodes" and I don't know what they mean by that term. Many 3D programs allow the purchase of "seats" which expand the number of users that can use the software under your license. New users simply activate the software using the same license number. As long as the total number of users remain within what is allowed by the license everything is fine. Does rendernodes equal the same system? The main program that does the rendering is Vue and it can spread the rendering load out to multiple PCs using a local network and a small program called "Render Cow" and you are allowed an unlimited number of Render Cow nodes. That whole system is for producing CGI animations with Vue. In this case, Vue would only be used to export the plants out into the FBX format after they are created using PlantFactory.

Here are my questions. Can the Creator level which is for personal use only be used to create plants that can be uploaded to the DLS where they can be distributed under N3V's enterprise license? Can the Rendernodes be shared with select individuals under the enterprise license? The software is exactly the same among all 3 license levels so there is not a compatibility problem.

Now Vue is far more sophisticated software than the E2 engine. It has been using procedural terrains and textures for about the last 15 years. This allows it to do some amazing things. Yesterday I rendered a scene which contained roughly 18 billion polygons. However, it took about an hour to complete. Hence the need to spread the rendering work across a render farm. Thankfully, this software allows the native Vue format to be reduced down to the regular FBX format of less than 50,000 polys for real time usage.

I just want someone that knows the E2 engine to see if this is a possibility for the future. The answer may be no but I think it is worth looking at to see.
 
If it was easy, we would have done it already. :)

Things to consider - engine support, animation, software license fees, redistribution license (many tree packages won't allow distribution in games), finding which package is best, figuring out how many people would actually make the trees, how to avoid duplication (like we have with 1,000 "ballast" textures.

If someone wants to evaluate all the options and licenses out there, we're happy to consider providing some resources to assist.

Hmm thats a bit of a throwaway line as it ignores the use of Blender , which is pretty much cost free , at the cost of compiling a tutorial I can't see any other objections in its way.

I hope I don't have to have to make this same post again in another three years if I'm still here but i have a nasty feeling that I will .

Is there anyone out there who would evaluate the plug-in ? All we need is someone who if familiar with blender to download the plugin and play with it for a few hours to see if its possible to use it for trainz and if it will make assets with relatively low polys.
 
Last edited:
Hmm thats a bit of a throwaway line as it ignores the use of Blender , which is pretty much cost free , at the cost of compiling a tutorial I can't see any other objections in its way.

I hope I don't have to have to make this same post again in another three years if I'm still here but i have a nasty feeling that I will .

Is there anyone out there who would evaluate the plug-in ? All we need is someone who if familiar with blender to download the plugin and play with it for a few hours to see if its possible to use it for trainz and if it will make assets with relatively low polys.

If it was simple then someone would have done it. There is a reason that speedtrees are widely used and it's probably the same reason that Microsoft gave up on train simulators. This stuff is complex and best left to niche specialists.

Cheerio John
 
Is there anyone out there who would evaluate the plug-in ? All we need is someone who if familiar with blender to download the plugin and play with it for a few hours to see if its possible to use it for trainz and if it will make assets with relatively low polys.

I just downloaded 'modular_tree_4_0_2_windows' and created a test tree. No leaves or textures for this test. In Blender 236316 triangles. Decimated down to 47263 triangles before it becomes too distorted. Exported as fbx and imported into Trainz Ok.

Too many triangles just for 1 tree.

Regards,

John


MTree-Test-1.jpg
 
The Plant Catalog is made up of add-on plant species available for sale. Both PlantFactory and Vue come with builtin base models of more common plants. My Vue Complete 2016 has several species of maple and ash trees.

The problem with XFrog trees is that they are not designed for use in real time as in a game. They are designed for 3D software that produces animations for CGI. The license terms do not allow you to distribute the tree models themselves, just the images of them you generate.

That's what a thought about the XFrog trees. They do look nice though. The problem I see it if the trees are images, then we're back to square one with flipboards or billboards.

An alternative to this is Digital Element's Verdant Plants and Trees.

Verdant « Digital Element, Inc. (digi-element.com)

These can be exported and used in other programs. I've never tried the demo, but my gut feeling is that these are probably billboards as well.

Vue Esprit is/was a competitor to World Builder from Digital Element. World Builder has the Verdant tree system built-in and the trees are created using an L-system program module. The trees and plants can be "genetically" altered and can produce children based on the parents. World Builder was not inexpensive when it came out in the mid-1990s and is still available today. The problem is that it renders really, really, really slow even on modern processors and graphics cards. With that said, World Builder can use slaves, i.e. nodes to do the frame-by-frame (individual tiff images) to generate animations. The program used to connect to 3ds Max and other 3d programs and allow direct rendering and manipulation within those programs. An object generated in 3ds Max can then become an object that picks up the rendering and effects from World Builder using this connection. I used World Builder for some hobby projects and for some graphics used in some advertisements done for my family's graphics business.
 
I looked at World Builder when it first came out and was amazed by the realistic images it produced. If it had supported Lightwave 3D I would have bought it. I chose Vue Esprit 6 instead as it did support export to LW. With version 7, I upgraded to Vue 7 Complete that included all the optional add-on packages. Around 2015 E-on's website got hacked and the thieves got away with all the company's customer records as well as emptying out the store. It was a huge mess. It took about a year for them issue new serial numbers and get everything working again.

It sounds like I used Vue pretty much as you used World Builder. I generated background plates and reflection maps for CGI projects for local video production companies. For some unknown reason Birmingham was a hot bed for video production back then. Big companies doing national ad campaigns. I first got involved with them in 1990 when the NewTek Video Toaster came out and I got one. All of them wanted a demo so I would pack my monster Amiga 2500 up and take it to their locations. I think sold about 20 of them for the local dealer.

I started in 3D in the mid 80s using programs like Imagine, Cinema 4D and then Lightwave. I got a bit burned out working 100 hour weeks and jumped into computers and database design. But I still love 3D so when I retired I picked up LW and began to relearn it all over again. Trying to wrap my head around using nodes to do textures these days.

I think the big problem with trees in Trainz now is that none of us know what the E2 engine can do or how to build something for it. Really need an up to date CCG for it.
 
It sounds like we both got into this at the same time! The old Amiga machines were quite powerful for their time. I worked for someone that used one for video production for his training videos for injection molding. The early version of World Builder didn't support Lightwave but later on it did. I think this option came about in version 2.0 and up. After that, they used Rosetta for that which I think supports multiple formats. I never did much with that other than use it with the old 3ds R4 for DOS and later 3ds Max and an early version of that too.

I was looking at other world and landscape-generating programs including Vue Esprit when I I learned about World Builder through a friend of mine. She did a video interview in Russia with the lead developer at Animatek before they sold the program to Digital Element in California. Animatek is famous for their computer fish tank program El-fish that allowed people to raise their own fish, crossbreed them, and maintain them as living creatures. I had that program early in the PC days. What caught my eye was the realism and the collision detection. Imagine placing objects and have them squish grass down instead of clipping it. Their demo of dinosaurs walking through a grassy field was amazing and a great seller.

The problem I had and still have with World Builder is it's so slow Running the program on an i9-12600K with 64 GB of RAM and a RTX3080 still renders so slowly. Back then, a simple animation could take days at high resolution. To get around that, the ability to run a render farm was a necessity rather than a luxury. But still, even with the render farm, that program was and still is so slow.

What's good about it is the use of areas for texturing and asset placement. Draw a polygon over the region, for example and place the texture in that region and set parameters such as scale, angle, and other parameters, and the rocks will only appear on the face of the cliff and nowhere else. The same can be done with plants, grasses, and trees. These assets too can also be restricted from growing in water or on other paths such as roads. I wish we had that capability in Trainz. That would make placing grass and trees so much easier.

I agree with you. We don't want to invest in anything without knowing what the E2 engine is capable of. This software is way too expensive to invest in only to have it not work due to some incompatibility with the E2 engine. Having that CCG would work wonders for that.
 
If it was simple then someone would have done it. There is a reason that speedtrees are widely used and it's probably the same reason that Microsoft gave up on train simulators. This stuff is complex and best left to niche specialists.

Cheerio John
Well John, unless some of us actually test to see if it's as complex as you think it is, we'll never know will we? As there haven't been any niche specialists creating trees for the past three years and there are no signs of anyone else announcing they are going to make vegetation just sitting on our hands isn't going to achieve anything. Using blender may well be a dead end , but it also may prove fruitful, using libraries for trees on the DLS has resulted in trees being under a mgb each, there are also some trees on the DLS that are 50mg each , we will only be able to achieve anything on this front if we experiment , not give up before we've even started because it might prove difficult to achieve workable results.
 
Blender has a built in Tree generating script will that work? You have to enable it in pefrences. I was able to add a tree in blender and customize it. The customization options box pops up after add > curve > Sapling Tree Gen is clicked. I'm not sure how you would get the poly count down for something as complex as a tree.
 
Last edited:
Wow 4 pages of debate on this. I thought this was a train sim, :wave:not a vegetation sim.

Well you better remove ALL vegetation from any route you use then and just keep the tracks and rolling stock, in fact why not just have bare baseboards ? , then we can remove all those other superfluous items we don't need and simplify everything.Why didn't we all think of that before ? , possibly we don't have your intelligence level.

You apparently have no interest in the subject we are discussing and have no constructive comments to make , so I presume you thought you would inject some levity into the proceedings, well you failed as far as I'm concerned. :)

Please let me know which routes you have been using that have no shrubs and trees so I can avoid using them.
 
Well you better remove ALL vegetation from any route you use then and just keep the tracks and rolling stock, in fact why not just have bare baseboards ? , then we can remove all those other superfluous items we don't need and simplify everything.Why didn't we all think of that before ? , possibly we don't have your intelligence level.

You apparently have no interest in the subject we are discussing and have no constructive comments to make , so I presume you thought you would inject some levity into the proceedings, well you failed as far as I'm concerned. :)

Please let me know which routes you have been using that have no shrubs and trees so I can avoid using them.

Middleton for laptops base system has no trees or vegetation. Adding them would need the speedtree engine and with a target of being able to run on a system with a 3D score of 350 above 20 fps it simply wasn't possible. Even none speedtree vegetation would have added too much demand. It does have the station layouts from RMWeb though so you can do fairly interesting things operationally.

I think with canals some vegetation was added.

I get the impression that for many the range of trees available in speedtree 6 is sufficient but for some reason the ones you'd like for your particular layout are not available?

I'm not even sure why the version of trainz the assets are made in matters. For scenery objects surely the objective is to make them in the lowest version compatible so they are available the largest number of users?

I have seen a beta of some new spline vegetation with some randomness but I suspect the content creator will be targetting content that is useful to a wider audience.

TurfX not available to Mac users, or AMD GPU or Intel GPU users either. There is a tradeoff between using standards and performance and how long the standards bodies take to incorporate something. nVidia would prefer you to use nVidia cards hence their extension into TurfX, N3V didn't have to make use of it but it was relatively cheap to implement and adds value to many clients.

In the same way there was discussion in the forum about what to do with trees before speedtree was introduced about what to do. It was felt that the 2D style trees didn't look too wonderful. When asked what did other games use Windwalkr mentioned speedtrees and that is the option that the community thought worth perusing to get a reasonable number of reasonable trees quickly given that it was recognised we didn't have the resources to create them from scratch. That was speedtree 5.

speedtree 6 came out and that meant all the 5 trees would no longer work and that wasn't nice. speedtree is now up to 8.4 and the licensing costs are substantial so I don't think that is an option these days besides too many layouts would need to be updated.

I think given the resources available N3V aren't doing too bad a job. Should they have diversified into the Mac environment dunno, I assume that the return on investment is equal to that of the PC in terms of numbers of copies sold per dollar invested but it keeps windwalkr happy which is a consideration.

At the end of the day it comes down to resources available. In general people specialise in their niches and are more productive as they gain experience. However it is open ended to an extent and you always have the option to create the content you'd like to see. Blender.org has the software and tutorials. In software development we used to think it took two years to become proficient in a language if you were working in it full time. I suspect the learning curve for Blender is similar. Whether or not you'll be able to create content that equals speedtree 6 trees at the end of two years is a matter of debate but if you have the time available you might like to try.

N3V I think commissioned some speedtree 6 trees, if there is something specific you'd like to see in speedtree 6 and you have the money available I feel sure some contacts could be found but we are probably talking $50 an hour and several hundred hours work.

Have fun

Cheerio John
 
As some sort of fiscal incentive for this project we could set up a hedge fund.

Sorry .... not helping :)

That might well help the English hedges but how would you do it?

Step one would be gather funds, using something like wise.com that wouldn't be difficult.

Step two would be which hedges do we want to fund? and finding a method to decide that was agreeable to everyone might be more difficult. Also would these be payware of freeware, on the DLS or not?

Cheerio John
 
Back
Top