N3V: Its time to update the low poly threshold to match the TRS19 marketing hype!

Hello Bob. I could be wrong, but your assets on the DLS really don't appear to explain how they were made

They do however provide the following really useful information on you so I can judge the merits of what you are telling me I should do in Trainz:

1. The quality of your work that I can actually see.
2. The breadth of your work that I can actually see.
3. The sophistication of your work that I can actually see.
4. What level of Trainz you are currently designing for that I can actually see.
5. How well you support your older content, or if you do at all that I can actually see.
6. That you have created similar content using what you are suggesting that I do that I can actually see.

Bob

Again, Bob, Making LOD assets doesn't really have to do with Trainz. That's something you would do in your 3d modeling software. I'm not telling you what to do with trainz (As I expect you already know how to script a basic lod asset). My argument wasn't about the trainz level I'm developing at, or the support of older content. The discussion was about creating multiple LOD's for your models and implementing the LOD's into your assets (Something we both know how to do in trainz).

If I am understanding correctly, you are skeptical of my skills because I haven't shown much to the public, and you don't want to use LOD because it takes too much time (I would argue it doesn't have to take much effort, but I could be wrong).

If you are interested in my 3d modeling skills, I'd recommend you look at this thread, where I post all my progress: https://forums.auran.com/trainz/sho...any-other-modeling-program)-screenies-renders

And as I said before, Trainz isn't your regular high-end game, and it can be quite slow compared to some of its competitors that have much higher polys assets, but are running on a more efficient game engine. NV3 recommends your small scenery assets have no more than 4k polys, and the large ones have no more than 30k. This link might be helpful: http://online.ts2009.com/mediaWiki/index.php/Art_Recommendations. You can't expect to run hundreds of 100k poly assets on trainz as you would with a game that uses say Unreal Engine.

And with all respect, I still appreciate all the money and all the assets you have bought and imported into trainz for all of us to enjoy. So thank you for that:)

If anything I said is wrong, I'm completely open to corrections.
I hope this helps.
 
Last edited:
Yes, I am aware of that. Most modern games use some kind of LOD system for performance. However, the close-up (and overall) poly detail of some of N3Vs competitors is much higher than of TRS19, yet I get overall better performance.

That is certainly my experience as well. I don't know how they manage to do it. Better game engines and more skilled 3D artists, I presume.

I've only ever purchased and uploaded one 3rd-party model. It was Frank Lloyd Wright's famous house "Fallingwater". I bought it from TurboSquid to help me fill a request job (from Epoche3bis4) because I simply could not work out the structure from published plans of the house. The model, as purchased, was terrible - insanely and needlessly high in poly and vertex counts, badly UV mapped with awful textures. I could not morally justify uploading it as it was, so it took me months to unpick the mess that I had paid money for, then simplify the mesh, create LOD's and map it with better textures. Never again!
 
Hello Bob. I could be wrong, but your assets on the DLS really don't appear to explain how they were made

You are correct, if you don't download and look at them in Surveyor or open them up in "preview asset", as I would to see a good cross section of the work and how that person did things. In your case there is nothing to do that with on the DLS. If your items are .IM meshes I could even look at them in Blender to see how you do and structure things, should I choose to do that using the Shareware 3D Object Converter application (and no I wouldn't abuse this and never have!).

In my case I started some years ago with GMax and eventually migrated to Blender and have never looked back. With CGTrader assets I have to find out which import file format being made available looks best (ie; Blender, Obj or FBX). There are often a number of issues to fix before they can be used in Trainz, and very few of them work "out of the box". Sometimes I have to make major changes in order for them to work in Trainz. Same goes with their provided textures. I cannot provide the Blender source code for those items to others per their CGTrader royalty free license. For many of the other content items I have done I provide all my Blender mesh source code along with that content item for anyone to use as long as what they make from it is (a) freeware and (b) loaded only to the DLS for everyone else to use. Considering how long I have been around in Trainz I am surprised you are not aware of all this. I am not sure anyone else does this when it comes to the mesh source code but me at the present time. Feel free to make use that Blender source code yourself, although don't expect perfection, I am just a talented amateur at this point and have never claimed otherwise. Also search the forums using the keyword "MSGSapper" and you will also see many contributions in tutorials and other things I have done over the years for Trainz users.

The discussion was about creating multiple LOD's for your models and implementing the LOD's into your assets (Something we both know how to do in trainz).

Actually my original post that started this thread was centered on increasing the LOD error threshold and that was all I intended for it. As with so many other threads we have gone far afield from what this all started about, especially on the topic of the merits of multiple LODs.

If I am understanding correctly, you are skeptical of my skills because I haven't shown much to the public, and you don't want to use LOD because it takes too much time (I would argue it doesn't have to take much effort, but I could be wrong).

Yes on both points. If I came out of the blue and took you to task for the work you have done and told you you needed to do things differently wouldn't you want to know I had the chops to make these statements, or would you take what I said on simple blind faith? You came out of the blue, and I don't remember ever running across you or your work before, and I have been here for many years, so I think wanting to know something about your level of competence in making such statements is not unreasonable at all. I tend to evaluate Trainz content item competence based on hard evidence provided by items uploaded to the DLS, and not just talk. Again, I don't think that is unreasonable at all as most authors provide little hard information on themselves to come to a competence decision with, other then what they have uploaded to the DLS.

As for the subject of multiple LODs I want to put the multiple LOD thing to rest here so we will just have to agree to disagree. I understand the merits completely, but I choose not to implement them at this time for my own reasons. At some point in the future I may revisit these items I have done and make such multiple LODs available for them.

Right now I have a full plate and no time to do it. If I am going to be punished for that by N3V then I will simply stop making content for Trainz, but in fairness you better be willing to punish a fair number of other major content authors out there, assuming they are still around, who have never done or are not doing multiple LODs either.

If this level of complexity and expectations continue to advance with each new major Trainz version eventually the only quality content authors left will be those who are making payware, and that seems to be the general direction N3V is trying to push us into, as they get a slice of the payware action. Don't believe me? When was the last time you saw highlighted freeware content, not routes, in the Trainz newsletters? Freeware authors get little recognition or encouragement from N3V for what we do for Trainz, except in the area of routes, which they do recognize.

BTW I can't blame N3V for that, and I understand it, but I will never be a payware content author, nor want to be. I am in this for the fun, not the money. For those who want to be payware authors more power to you and by all means do that, but that is not the path I choose to walk. I want folks to have a little fun in their lives with this hobby and if in some small way I can provide quality almost-payware level content items at no charge to them to enjoy, then so much the better.

BTW nice Wiki entry!

Bob
 
Last edited:
Fair enough. I would also be curious as to how much effort it would take to switch TRS19 to a game engine like Unreal.
 
Fair enough. I would also be curious as to how much effort it would take to switch TRS19 to a game engine like Unreal.

I think the problem would be in the control N3V that have with their own game engine. Scripting and AI come to mind to start with.

Cheerio John
 
Fair enough. I would also be curious as to how much effort it would take to switch TRS19 to a game engine like Unreal.

That's something that's occurred to me. If you search in Blender Market you will find 1645 entries including export helpers for UE4, meshes tested in UE4, etc. CG Cookie and Allegorithmic (Substance) has tutorials supporting UE4. But I often feel embarrassed when potential new creators ask for courses, and especially video courses, on creating assets in Blender for Trainz. There are plenty of content creators, myself included, who could create such videos and courses but we don't have the time or the inclination. I only know of two such videos and both were made by GDennish. If there are more then we need to advertise better.

But I digress from the original topic. On balance, I agree with most of Bob's comments. Perhaps the minimum poly limit could be a bit higher.

I notice no one has commented on the 10,000 poly warning for scenery assets, even if they have LOD. One wonders if that will be an error in TB 4.7/8/9.

And there is the uniform colour warning which is now an error unless a very small image. I spend more time trying to get around that than I do for low poly models.
 
Fair enough. I would also be curious as to how much effort it would take to switch TRS19 to a game engine like Unreal.
A few years to recode everything plus paying Epic 5% of sales for the privilege, besides N3V developed their own Game Engine over several years and are still improving it. I have more respect for a company that creates it's own Game engine rather than use an off the shelf one as DTG did with TSW and it wouldn't solve the problem with inefficient assets unless they started from scratch and removed backwards compatibility and that would probably loose most of the customers.
 
But I digress from the original topic. On balance, I agree with most of Bob's comments. Perhaps the minimum poly limit could be a bit higher. I notice no one has commented on the 10,000 poly warning for scenery assets, even if they have LOD. One wonders if that will be an error in TB 4.7/8/9. And there is the uniform colour warning which is now an error unless a very small image. I spend more time trying to get around that than I do for low poly models.

Paul:

Each new toughening of the standards results in developers either leaving or not upgrading their older content to meet those new standards. In my case it has been a tough haul to bring my older content items up to TRS19 and PBR standards, so much so that I seriously doubt I will be supporting Trainz content wise past build 4.6. I don't disagree with the need for those standard changes, and perhaps understand why they are doing it, but N3V needs to be aware of the cost and repercussions of such moves.

Over the last couple of years we have sadly lost some really great major content developers due to death, poor health, burn out or just plain frustration. The number of replacements coming in with matching capability just hasn't kept up at all. The DLS, and even TRS19, is currently crammed with poorly aging content that is not being upgraded as most of the original developers have abandoned those items for one reason or another. By upgraded, I don't mean just to be able to function in TRS19. I mean bringing them up to match the look and appearance of the current major Trainz release. As an example a build 1.2 structure looks "toyish" in todays Trainz, yet there are thousands of those items built-in to TRS19. IMHO they should be removed or archived as someone else suggested as they don't make the current platform of TRS19 look good at all.

Also N3V does not provide any real encouragement to freeware content developers beyond routes. The used to provide some stuff to help developers, like Content Creator Plus and Trainz Paint Shed, but those items went away many years ago as useful tools and have not been replaced. Of course why should they, as those freeware items compete with payware items which they get a percentage of. They are a business, so I understand this, but at least they could provide a bit of recognition of our freeware development efforts in their newsletters (not routes, which they do do this for). On the other hand I do give N3V high marks for interacting with us in these forums.

Developing for Trainz involves trying to work around all those quirks you run into, such as the uniform color issue, which seems to makes only sense to N3V. If it is a 32x32 or 64x64 pixel single color swatch who cares and why would you? But that's N3V and fodder for a separate topic thread of its own.....

Bob
 
Rather than wasting dev time (however much) on un-doing something that solidifies the future of the game's graphics/performance, a bigger and more real concern that is quite strongly pushing content creators away is the lack of 4K textures and no mentioned plans to bring them to the game. It's a big turn-off when you can see crisp textures on your content in TS, but in Trainz they look speckled with glitter and sand.
Perhaps the emphasis on the game is becoming more "railway simulator" where you're an overlord that views trains from a distance in a model railway fashion, rather than a driving "train simulator" where you can get up close to your trains and appreciate the handling, sights and sounds of your assigned loco/stock.
 
Back
Top