N3V: Its time to update the low poly threshold to match the TRS19 marketing hype!
- Thread starter MSGSapper
- Start date
Tony_Hilliam
Trainz Plus - enjoy Trainz from just 20 cents a da
Put another way, LOD improves performance. Any work around to avoid using LOD means performance suffers.
If performance suffers, users complain.
If users complain, we sell less units, so Trainz is less profitable.
Less profits means longer time to update and less features to add.
So ultimately, effective LODs determines (to some degree) the future success of Trainz.
If performance suffers, users complain.
If users complain, we sell less units, so Trainz is less profitable.
Less profits means longer time to update and less features to add.
So ultimately, effective LODs determines (to some degree) the future success of Trainz.
My possible wrong assumption: High poly counts directly require more memory on the GPU to hold the stuff when the program finishes processing. Thus you can run into GPU memory errors. I have had that with the Lockheed Martin flight sim I use. The product is free with the financial support coming from commercial airlines and flight schools. With the latest version if I rapidly cross terrain at low altitude (max detail) I hit the memory limit on my miserable 3gb 1050 card. I never saw that error before, despite years of use. Lockheed has upped the detail over the years for commercial customers thus the users of the no cost product see no warnings or attention from Lockheed.
N3V does not have the luxury of designing to the commercial market's wishes. Protection of the revenue stream is of utmost importance. The complicating factor is the "non-standard" market that is populated with a wide range of hardware, levels of knowledge and patience.
Putting poly counts forward to customers as "visual performance options" is one way to approach the problem. BUT, then you really upset those who cannot afford to use the available ($$$) graphic detail. They bought the program and once again people with money are happy and they are stuck!
N3V does not have the luxury of designing to the commercial market's wishes. Protection of the revenue stream is of utmost importance. The complicating factor is the "non-standard" market that is populated with a wide range of hardware, levels of knowledge and patience.
Putting poly counts forward to customers as "visual performance options" is one way to approach the problem. BUT, then you really upset those who cannot afford to use the available ($$$) graphic detail. They bought the program and once again people with money are happy and they are stuck!
Last edited:
But, if you use lod, those with the lesser capable hardware can use Trainz more effectively, more consideration needs to be made of those without deep pockets and expensive high end hardware, I'm using lod on anything I create that's over 500 polys and have updated a load of my stuff that didn't have lod or not a good implementation of it. With the mesh-table-lod-transition-distances in TRS19 as well as having control over when the lods cut in, I am culling the items as well if out of view. Baseline is the minimum spec required to run TANE or TRS19, routes should be usable on that equipment.
I'm also in the process of hunting down the high poly assets in my WIP that are killing frame rates, i7 6700K + GTX1080TI. Mostly these are UK houses which look exceedingly good but well over 1000 triangles with no lod and if viewed in preview assets performance, kill the FPS to 10 or lower, in worst case locked up the preview in performance. Downside is having to make suitable replacements myself, route requires a lot of houses however it also needs to be usable on lesser capable hardware than I have.
I'm also in the process of hunting down the high poly assets in my WIP that are killing frame rates, i7 6700K + GTX1080TI. Mostly these are UK houses which look exceedingly good but well over 1000 triangles with no lod and if viewed in preview assets performance, kill the FPS to 10 or lower, in worst case locked up the preview in performance. Downside is having to make suitable replacements myself, route requires a lot of houses however it also needs to be usable on lesser capable hardware than I have.
I agree that LOD does indeed improve performance, but it is not the only way to do that. You provide a ton of tweaks in the Performance tab that a user can do to improve FPS. Also they can further tweak the Trainz Settings once the game is launched. Examples are Maximum Draw Distance, Scenery Detail and Tree Detail. Your comments here seem to put the burden too much on the 3rd part content creator and not enough on the user. It has to be a shared responsibility here. Also users can preview a content item in TRS19 and see if they can handle it or not, and if not, don't use it. I admit I primarily develop for the Ultra Graphics user setting in mind when it comes to build 4.6 items. Realism is what I am after in Trainz and that is why I am a Trainz Plus subscriber. I believe in this so much that I have spent close to $1400 out of my own pocket over the last few months purchasing professional level models from CGTrader and adapting them for use in TRS19. None of these came with LOD and I haven't the skill needed to produce LOD versions of those models as they are often very intricate. I did this for better looking content for my 14 routes that more closely matched the appearance and capabilities of TRS19, as much of the content on the DLS is not up to that level. Your simulation, and every other user out there, benefits from this as I had to upload those items to the DLS.
The primary reason most people, including me, continue to buy and use your Trainz simulator is because of the amount of freeware content available for it compared to rival products such as Train Simulator 2021 or Train Sim World 2, which charge for additional content in a manner similar to the old Microsoft Train Simulator. If quality content isn't available to match the realism capabilities of the current product (ie; TRS2019) users aren't going to be happy and will look elsewhere. In case you haven't noticed the number of quality freeware content developers has diminished as of late because of various reasons such as mortality. I might also mention that some of the people who have spoken here in favor of the status quo are not currently creating build 4.6 PBR quality items for your platform, but continuing to develop for T:ANE or earlier products. Creating content that is both realistic and uses PBR textures is much more difficult then building a simple box which paints house features on it. Add LOD, night and winter modes on top of that and it becomes extremely difficult and time consuming to do with very detailed models. Maybe that is one of the reasons that so little of the older content you continue to carry in your platform and the DLS is brought up to current standards. As an example I count 7954 build 2.0 or earlier built-in items in TRS19. The only reason many of those items continue to function at all is that you have exempted those items from current standards. A build 1.2 item in TRS19 doesn't make your platform shine in the way you intended.
As many game companies have found out, trying to please all levels of gamers is a losing proposition and holds back the game in the end. Your attempt to resolve this by producing too many different versions of your platform, all of which you have to support, acts as a drag on the development of the flagship product because scarce programming time must be divided between them. Add into this continued support for older versions and I can't imagine how you are going to remain profitable over the long haul. Your game company, and I know many, is the only one I know out there who does this to the degree you do. It will kill your business in the end if you continue with that "trying to please everyone" strategy.
So ultimately, effective LODs determines (to some degree) the future success of Trainz.
On this point I disagree almost entirely. The continued development of quality freeware content that supports the degree of realism achieved by the current platform will be the factor that drives customers to you. Without all that quality freeware content, your Trainz platform would have little to offer that sets it apart from your rivals out there who mostly match you in the graphics department - and without the "stuttering" I might add. I know as I own more then one Train simulator product, but I stay with your primarily because of the freeware. It is the freeware content on your DLS that makes your simulation platform so inviting and sets you apart from the others.
Now other then making it tougher on us freeware developers, what have you done to encourage us freeware content developers to continue making better quality build 4.6 or later stuff for you considering all the effort it takes these days to do it? Little, would be my answer. Highlighting new quality build 4.6 or later content recently uploaded to the DLS in your newsletter, in addition to routes, would be a good start. If you want stuff that matches TRS19 in appearance and realism a little recognition would help......
Bob
Last edited:
johnwhelan
Well-known member
I think you're missing an important point. There are more people with a machine uses Intel integrated graphics than any other GPU. These machines can run Trainz, TANE certainly. Intel HD graphics 620 which is the most common one on new machines has a benchmark score of 922 far higher than the old series 4000 but still well under the 10,000 normally considered optimal for Trainz. https://www.videocardbenchmark.net/gpu_list.php
These machines are capable of running TANE with the sliders set to minimum and if you select the content carefully. Which means they are a target market for N3V. They aren't an ideal platform but there are a lot of them and if they get into the SIM then they can upgrade their hardware. Asking people to spend a $1,000+ on a machine that has an RTX 2070 or greater to try out a Sim that costs $5 (recent N3V sale price for TANE.) doesn't make logical sense.
Users with an RTX2070 or better probably make up ten percent of the buyers of TANE.
Like it or not the majority of end users will notice better performance with assets that have LOD. Yes at one time we had content creators who took the view that if it runs on my machine then it's efficient enough but in these days of gobal warming making content that needs a little less electricity is surely a good thing. Perhaps I should qualify that with if you accept global warming as a fact, some don't.
It also has another impact, by allowing trainz to run on lower end machines it widens the market place considerably. Think it terms of lower income countries, to make trainz run needs a certain amount of computer expertise, creating content more so. However this then gives you a pool of expertise in computers. Even doing things like using the forum is a skill set that can be applied to other things and since the world seems to be short of programmers and technical people running Trainz could be a good introduction to the frustration of using a computer.
Cheerio John
Bob, I can see where you are coming from. The newer features of Trainz - PBR textures, LOD, seasonal scenery, built-in collision avoidance, etc makes it much more difficult, both in time and effort, to create quality assets. The software required to do the job properly is also more sophisticated (and sometimes no longer based on using freeware).
Put this in perspective. The first versions of Trainz came on CDs (remember them) which included a copy of GMax. You could create "state of the art" scenery assets for Trainz that "looked good" according to the standards of that time. Fast forward to today - there have been a number of threads along the lines that the old GMax created assets, i.e. those with low build numbers, simply do not "cut it" anymore and are causing performance issues. Mention the name "SketchUp" in these forums and watch the reactions.
When TANE was released there were quite a few posters who bravely suggested that N3V should "break the ties" with all the old asset standards and start completely afresh.
I was until recently running TANE on a low end machine with built-in Intel graphics and limited RAM. This was to test how a TANE route I was developing would go on a lower end machine. I had all the settings sliders set to their minimum values to get any sort of performance out of the program. My TANE route and sessions still ran sluggishly on that machine but by experimentation I also found that there was an improvement when I replaced the older significant (trackside) assets with newer versions that used LOD. A classic example of a user that, as John mentioned in his post, would benefit from the addition of LOD to assets.
You are correct, N3V cannot hope to please all their users all the time and, as you stated, any such attempt would hold back the game. I would put it that insisting users to continue to accept "parts made from rubber bands and chewing gum covered with a thin coat of whitewash" and put up with the resulting performance issues is far more likely to hold the game back.
My opinions.
Put this in perspective. The first versions of Trainz came on CDs (remember them) which included a copy of GMax. You could create "state of the art" scenery assets for Trainz that "looked good" according to the standards of that time. Fast forward to today - there have been a number of threads along the lines that the old GMax created assets, i.e. those with low build numbers, simply do not "cut it" anymore and are causing performance issues. Mention the name "SketchUp" in these forums and watch the reactions.
When TANE was released there were quite a few posters who bravely suggested that N3V should "break the ties" with all the old asset standards and start completely afresh.
I was until recently running TANE on a low end machine with built-in Intel graphics and limited RAM. This was to test how a TANE route I was developing would go on a lower end machine. I had all the settings sliders set to their minimum values to get any sort of performance out of the program. My TANE route and sessions still ran sluggishly on that machine but by experimentation I also found that there was an improvement when I replaced the older significant (trackside) assets with newer versions that used LOD. A classic example of a user that, as John mentioned in his post, would benefit from the addition of LOD to assets.
You are correct, N3V cannot hope to please all their users all the time and, as you stated, any such attempt would hold back the game. I would put it that insisting users to continue to accept "parts made from rubber bands and chewing gum covered with a thin coat of whitewash" and put up with the resulting performance issues is far more likely to hold the game back.
My opinions.
Last edited:
John:
Two questions:
1. Have you actually created any content, PBR or otherwise, specifically for TRS19?
2. Have you created multiple LODs of a content item yourself for TRS19?
Bob
Actually I have....3 stream locos with another one on the way which comes in at a whopping 200k polys with 5 levels of LOD and full PBR.
I have been creating with LOD since T:ANE came out and do not find it anymore difficult than doing without LOD. Why do I use LOD.....simple....performance.
My system is what I call a mid range setup (AMD Ryzen, 16GB RAM, 1660 GTX) and I have had simple house models built with no LOD at build level 2.5 bring my system to it's knees. My current WIP progress route which will only run in TRS19 and uses PBR almost exclusively where I can in both textures and models gives me acceptable performance with most sliders set to high or ultra od an avg 38 FPS.
I class myself as one of those avg users you mention and I can say that without LOD being used......my system would stall completely.
I have been creating with LOD since T:ANE came out and do not find it anymore difficult than doing without LOD. Why do I use LOD.....simple....performance.
My system is what I call a mid range setup (AMD Ryzen, 16GB RAM, 1660 GTX) and I have had simple house models built with no LOD at build level 2.5 bring my system to it's knees. My current WIP progress route which will only run in TRS19 and uses PBR almost exclusively where I can in both textures and models gives me acceptable performance with most sliders set to high or ultra od an avg 38 FPS.
I class myself as one of those avg users you mention and I can say that without LOD being used......my system would stall completely.
johnwhelan
Well-known member
I have 1,671 assets on the DLS. There might be one or two that do not have LOD but if they are more than 500 polys it will be in the order of one or two. Many of my assets have been reskinned and the reskins all have LOD where appropriate. As far as I am aware these assets all run in TS19. There was some problems with some assets created for Middleton for laptops but I seem to recall that was a script problem and was resolved by an N3V update. I
use TS19 to create assets but normally give them a 3.5 version number or TS12 equivalent as I know many users still use TS12 and don't have the funds to upgrade their GPU so they can run TANE.
Scripts in different versions have always been a problem.
Yes I am aware of the problems of creating content using PBR but to be honest I'm quite content to stay with Blender 2.79b and the older .im formats. I don't really see there is that much improvement using the PBR side of things and if there is a difference then you probably would need to use PBR on every asset in the screenshot.
Basically my LODs are take the original finished asset,copy and rename it. Then work my way through the LOD version, I sometimes use two levels, throwing away bits that I think cannot be seen at a distance. Typically I will use the same material on both the asset and the LOD versions. I am aware of N3V's idea of taking a screenshot of an asset then using that as a texture on the LOD but I'm too lazy to do that. I don't see that using a PBR material would mean you couldn't use the same material on the main asset and the LOD versions.
My background is both software including more than ten years writing assembler language code and hardware so I'm probably more sensitive than many to things that impact performance. Before I retired I worked in federal government and one of the things I did was cost reduction by matching the hardware purchased against the requirements. On the database servers making the memory on the server and selecting the correct hard drives often meant I could get away with a simpler software license which saved the odd $50,000 occasionally.
Cheerio John
ZecMurphy
Community Assistant
These settings only function if an asset has LOD. That is what LOD does, reduce detail as you move away. If you don't have LOD, these functions do not work, as there is no way for Trainz to reduce detail.
LOD is not going away any time soon. If anything, it will become more strict as we introduce more graphical features. There are finite hardware resources, and this means that you need to use them efficiently. LOD is a big part of using these resources efficiently.
LOD is one of the most important parts of making an efficient, and usable, model for Trainz. It is ESSENTIAL to making Trainz perform well.
We absolutely agree that we can't please everyone. But we do need to ensure that Trainz works on all computers that meet the minimum requirement. LOD is essential to this.
Realism in content comes at a cost, one of those costs is LOD. Again, the are very few games that you buy that will not use LOD in some form or another. It is an essential part of game art. Very little stock art actually comes with LOD, as each games specific requirements are different. So this means that if you are purchasing stock 3D models to convert, you will need to generate LOD yourself.
There are tools that help with this. The 'optimize' tool in 3DSMax is an example. Same as, in 3DSMax, the 'editable poly' mode allows you to remove edges and vertices from a surface with minimal effect to mapping (unless you change an edge vertex on the mapping), allowing you to fairly quickly reduce polygons for each LOD.
The simple fact, again, is that LOD is not going away. If anything, if people continue to work around it, we will need to make it more strict. LOD is an ESSENTIAL part of making Trainz perform well, as it is essential to several of the performance options actually functioning.
Of course the alternative would be to go back to 'progressive mesh' type systems where the game engine just starts removing polygons for you, with limited control over what is removed (if anyone remembers the SAR Centenary Cars in TRS2004, which turned into an odd triangular bar when you were about 50-100m away...). But I'm not sure people would be overly happy with this type of system these days
Regards
My questions here are not meant to pick on fellow developers, all who I respect, but to see what experience they have with LOD and PBR in a build 4.6 or later TRS19 environment that gives them insights to the problems I have brought up. A few people that have posted here have done no build 4.6 or later TRS19 specific items, nor used PBR. I don't want to call them out by name, as I have worked with some of them before, and respect them.
There has never been as sharp a dividing line in Trainz development as there has been going from T:ANE to TRS19. I have developed stuff for Trainz for almost as long as Trainz has been around. Each new iteration has made things more complex and TRS19 is no exception. PBR and parallax has changed almost everything, and Trainz has yet to resolve some of the issues around PBR and parallax that currently still exists (mixing of PBR with non-PBR textures, sharp angles and parallax, etc).
BTW I do have hard personal experience with multiple LOD in a build 4.6 TRS19 environment. As an example, my experience developing my <kuid2:439337:103187:1> TRS19 SAP Track U.S. 132LB SG Mesh Library-Revised was a nightmare! I have developed track since before TRS2009. Then it was simple and didn't take long. Along came procedural track in T:ANE that upped the development effort required considerably. I thought I was prepared to deal with TRS19 when it came out but the added feature of PBR changed things drastically. It took over two months of six day a week development effort, and long hours, to finally get it right. When it was finally done I had a library totaling 88 meshes to deal with both LOD, track variants and snow. That wasn't the end however. The Moiré effect continued to plague me, because of PBR interaction with ties, and I eventually had to revise everything once again to finally get that issue under control, or at least not as noticeable as it was before, and that took a fair amount of time. As a result this will be the last track I ever develop for Trainz as I don't want to go through that ever again. I do this for fun, and this effort wasn't fun at all. I only did it this time because N3V did not make available a comparable U.S. track with TRS19 when it was released. It is also interesting to note that while I also provided all the Blender source code for all the meshes very few, if any, developers have used it to make their own track types variants. Developing for full build 4.6 PBR standards is not easy for any project of any complexity.
While we have talked here about TRS19 a lot of the points I have raised would apply to almost any of the high end games out there like Call of Duty, Civilization 6, etc, if you want to enjoy the full glory of all their graphics and effects. Civilization 6, which I thoroughly enjoy, even pushes my system at times when on Ultra or High settings. Those game companies don't hold back those games because folks can't afford the graphics cards to run them at those settings. Trainz, assuming it wants to be the ultimate and most realistic Train Simulation out there, shouldn't either in my opinion. The gaming and simulation industry has always pushed the hardware envelope and I expect that will continue. As a result I have to replace my graphics cards every 2-3 years to keep up with current games and will be replacing my EVGA GeForce GTX 1080, which was bought in 2017, with a new Nvidia RTX 3080 in Jan or Feb of 2021. I am a serious computer gamer and have over 230 Steam games at present so I really push the graphics at times.
Bob
John:
First of all I like your stuff and have used many of your items on the routes that I have done. I did a search of the DLS white pages and this is the content item breakout I show for you:
641 are for TRS2004.
30 are for TRS2006.
232 are for Trainz Classic
152 are for TRS2009
543 are for TRS2012
21 are for T:ANE
BTW I hope you will take the time to update those 1055 pre-TRS2012 items to something more current. It would be a real shame to see them lost at some point.
PBR does make a difference in the final realism and appearance of things in Trainz. As one example just compare the old 2d ground textures with the new 3D ones to see that for yourself, if you haven't already. The same goes with buildings and structures as well, although parallax textures can cause issues. Developing build 4.6 stuff that uses PBR is very different in many ways then anything we have done before TRS19 came on the scene. Updating an old .im mesh item that uses .onetex materials to simply work with TRS19 isn't necessarily the same thing.
You are indeed correct that for TRS19 to live up to its hype everything should eventually use PBR texturing in order to look right. As an example; mixing PBR grounds textures with non-PBR ones doesn't work very well and makes things look off.
Bob
Last edited:
Zec:
We seem to have a major misunderstanding here.
I am not against the use of multiple LODs. I fully understand why they are used and what they do for performance. Execution can be very difficult at times however with complex models that have a lot of features, especially when you expect each LOD level to be 20% different from the next higher level. I am also fully aware of Blenders capability to remove triangles using the Decimate modifier, but have found in practice that it can sometimes lead to bizarre appearance issues, especially when you are trying for a 20% reduction, based on my having used it a number of times.
In my OP all I addressed and wanted was updating the LOD error message threshold to something more consistent with current graphic standards. The level of 500, which generates and error if you exceed it, should be raised to something more current and realistic. Alternately go back to making it a warning instead, as it used to be.
Bob
johnwhelan
Well-known member
Quite a few of the very old stuff are scenery objects that look exactly the same with a 2004 build number or a TS12 build number so they might as well stay as they are.
The mesh libraries I intend to leave alone and I suspect I'll do the same with PBR textures. I prefer photo textures anyway. I am well aware of the efforts needed to create content to the new standards. I work with a very loose group creating content and at least one member who was creating coaches has just decided to give up creating content. The demands of TS19 are just too high.
N3V saying use the whatever command in 3DS doesn't cut it for me. Very few people have the correctly licensed of 3DS available. Blender moving to 2.8+ is another show stopper and a few in our group are still using GMAX and creating content. Part of that is the bit box effect. They have sub components that work in GMAX so just modify them slightly to create a new something or other.
Cheerio John
John:
Yes, they are.
As for me, I'll keep up with TRS19 PBR content development for now, or until N3V takes makes the content creation development process so complex and difficult that it no longer is fun anymore. After all, I am a hobbyist, not a N3V employee doing in-house development. Unfortunately that time may not be far off.
Bob
I sympathise. The issue of "creeping complexity" has been raised before in these forums. I see it an inevitable consequence of better technology leading to higher user expectations which leads to better technology and so on. It is not just hobby gaming that has become a victim of this cycle but virtually all aspects of information creativity.
It is becoming harder (and often more expensive) for us amateur hobbyists to keep up.
It is becoming harder (and often more expensive) for us amateur hobbyists to keep up.
Hi Bob, I totally agree with your original post. The 500 poly limit is just too small. You even need to hit this very small poly count for LOD1, which seems ridiculous, or at least I do unless I'm doing it wrong. It should be raised.
I am now only creating content for 4.6 using PBR, but spend a lot of time trying to get the textures correct. I use Blender 2.90 (well worth moving to from 2.79), GIMP, Substance Painter and Substance Alchemist. Alchemist by the way is an excellent tool for creating normals, height and ambient occlusion maps. Drop the albedo image in to B2M and export. The default settings have given me really good results. Not a cheap product, but given me far better results than Materialize. I tend not use parallax as I don't get the results I expected. All in all it has been a steep learning curve.
I also agree with John, regarding 3DS. The vast majority of content on the DLS has been developed by the fan base. We can't afford the so called 'Pro Tools' and tend to use the freebies. So give us a break Auran, where would you be without us. We need decent tutorials based around the tools we can afford. It's just not fair on people like Bob who has put so much effort into providing tutorials for us.
It would be wise to take on board some of the suggested comments and again I'm with Bob on the 500 poly limit. This would seem such an easy fix compared to the broken AI which needs serious work.
While I'm at it I think it's about time there was a DLS v2 with content from 3.5 upwards, after all you can't create new content for earlier builds because Auran no longer support it. You get errors messages and faulty assets if you try. Maybe the end user could choose which version of the DLS to use. I know you can filter content but the existing DLS is just way too big. Time to archive the older stuff, but that is just my opinion.
John
I am now only creating content for 4.6 using PBR, but spend a lot of time trying to get the textures correct. I use Blender 2.90 (well worth moving to from 2.79), GIMP, Substance Painter and Substance Alchemist. Alchemist by the way is an excellent tool for creating normals, height and ambient occlusion maps. Drop the albedo image in to B2M and export. The default settings have given me really good results. Not a cheap product, but given me far better results than Materialize. I tend not use parallax as I don't get the results I expected. All in all it has been a steep learning curve.
I also agree with John, regarding 3DS. The vast majority of content on the DLS has been developed by the fan base. We can't afford the so called 'Pro Tools' and tend to use the freebies. So give us a break Auran, where would you be without us. We need decent tutorials based around the tools we can afford. It's just not fair on people like Bob who has put so much effort into providing tutorials for us.
It would be wise to take on board some of the suggested comments and again I'm with Bob on the 500 poly limit. This would seem such an easy fix compared to the broken AI which needs serious work.
While I'm at it I think it's about time there was a DLS v2 with content from 3.5 upwards, after all you can't create new content for earlier builds because Auran no longer support it. You get errors messages and faulty assets if you try. Maybe the end user could choose which version of the DLS to use. I know you can filter content but the existing DLS is just way too big. Time to archive the older stuff, but that is just my opinion.
John
KotangaGirl
Pre-Grouping Railways Nut
I'm not an especially big player with content creation with only just over 300 assets on the DLS. I work in TS2012 3.5 and 3.7 and I know very well that when N3V finally shuts down accepting 3.5 assets I will be calling a halt on making anything except for my own personal use. I wonder how many other content creators will be doing exactly the same. That will leave content creation in the hands of an increasingly small number of hobbyists who have the time, resources, brain power and energy to work with and understand the new materials.
And then there is the question of how many Trainz hobbyists are using assets that use the new materials in TS2019. A good 95% or more of the assets used by the routes in my TS2019 install are 3.7 or older and to tell the truth I prefer it that way. Ultra realism is all very well, but not at the expense of having a railway simulator that actually works reliably at the nuts and bolts level.
And then there is the question of how many Trainz hobbyists are using assets that use the new materials in TS2019. A good 95% or more of the assets used by the routes in my TS2019 install are 3.7 or older and to tell the truth I prefer it that way. Ultra realism is all very well, but not at the expense of having a railway simulator that actually works reliably at the nuts and bolts level.
johnwhelan
Well-known member
Of course the interesting thing is 4k UHD monitors are now available for around $300. The RTX 2070 and above can handle them, today it might be necessary to select the content carefully but if N3Vdon't support it then someone else will. I hate to even think what that means for content creation.
Cheerio John
Cheerio John