.
Page 3 of 3 FirstFirst 123
Results 31 to 39 of 39

Thread: A hurdle that I can't jump!

  1. #31
    Join Date
    Nov 2006
    Location
    United States of America, Massachusetts, Haverhill
    Posts
    26,702
     

    Default

    Quote Originally Posted by Robert704 View Post
    Bump narrowgauge's original post to this thread. My shortline, basically logging centric, railroad uses a lot of content from narrowgauge and slugsmasher. A lot of the (prototypical based) mainline is in either logged, 2nd generation growth or old growth forest., and it looked MUCH better in TRS2010. Among the surprises in migrating to TRS2019 was a loss of a sense of density in these forests, probably due to resolution changes and lighting effects. I have played with lighting and environmental effects for several weeks now, but I not been able to replicate the "close in" effect of forests in mountain mist that I was able to attain on the previous route. No photos with this post, but I will try to fire up TRS2010 again and make a few comparison screenshots.

    Robert704
    Tacoma & Eastern RR (v)
    I've run into the same issue on my more modern New England routes. Like you have in the west, we have very thick undergrowth and forests, albeit, hardwoods because our pines and spruce disappeared long ago.

    In TRS2019 (TRS19), you can use older splines quite effectively along with the Speed Trees. For bushes and whatnot as splines, take a look at those made by JVC. For Speed Trees, Roystrainz, RMM, and mcguirel work depending upon the purpose. Roy and RMM made some beautiful spruce and pine trees. Roy also has some aspens as well that look nice too being a typical type of the poplar tree, they fit the bill there for me. Ettore48 has recently uploaded some grouped Speed Tree objects. These are made up from bushes and some trees by RMM. Intermixing these various kinds will produce some nice forests.

    More recently, user JVC is coming out with a line of billboard type bare trees, which are effective for dead wood and for bare winter-type scenes. These, however, will be payware and I know of them only through his announcement.

    The issue I do have with Speed Trees is they are too big, and too perfect most of the time, but used in conjunction with shrub splines, and Speed Tree bushes, (Kust in Russian), they can be quite nice. Here's something still a WIP without the shrub splines.

    2018-08-27 220046.jpg 2019-04-29 234541.jpg
    John
    Trainz User Since: 12-2003
    Trainz User ID: 124863
    T:ANE Build: 94829
    TRS2019/Trainz-PLUS: 105100

  2. #32
    Join Date
    Nov 2006
    Location
    Netherlands, for now in, Montreal, CA
    Posts
    5,205
    Blog Entries
    1
     

    Default

    Quote Originally Posted by narrowgauge View Post
    T...

    I have seen Trainz develop over 19 years from that Christmas in 2001 and each version has become more technically demanding than the one before. I have always regarded Trainz as a model railway hobby and enjoyed the experience of turning code into visible working models. Gradually, the restrictions and requirements have taken away the enjoyment of creating and I have now reached a point where creation for my 'hobby' has become a technical nightmare.

    This old monkey is too old to learn new tricks.

    TRS2019 has become the break-point at which I will enjoy other peoples work in 2019 but T:ane will be my construction target version cut-off. ...
    Peter
    I feel your pain. I've always been able to make small but useful objects without having to have a master's degree in graphics. It was fun and a joy to make and share. Luckily my major route never got past the track laying stage. In fact, now in TRS2019, with the new UDS, fixing the track and signal logic is so much easier - run a bunch of AI trains, see where they have an issue, jump into Surveyor, fix and then back to Driver. Repeat as necessary. With the size of the route, it will take a while to get it all to work but it is doable. As for textures and lighting, oh boy, am I glad that hasn't been done yet. Procrastination has its good points.

    The one major piece of work I was involved in that would suffer is luckily stuck back in TRS2006. I was recently talking to one of the other developers and the idea of updating it was considered for about 2 seconds until the magnitude of the work was mentioned. The thought was quietly put away and hopefully no one noticed.

    That made me wonder if trying to up version some of the classic routes is really something that needs to be done. "That was then, this is now" should be considered. It's a hobby and we should be in it for fun. With the new capabilities, perhaps new routes with the new objects and techniques should be done and explored rather than redoing yesterdays work, which if nothing else will just invite comparison.


    70337:
    TRS19 Platinum, build 105100, Win10 Pro 64 bit, i7-7700 3.6GHz 16 GB, GTX 1070 Ti

  3. #33
    Join Date
    Nov 2006
    Location
    Burnside, Adelaide, Australia
    Posts
    5,433
     

    Default

    Martin

    That made me wonder if trying to up version some of the classic routes is really something that needs to be done.
    It depends on so many factors.

    I can understand why N3V have to update, it is their bread and butter, but in doing so they have left so much behind. The insistence on cutting off earlier content is akin to burning all books before a set date. There are probably thousands of man-years in creation of content that is no longer accessible or usable.

    In the case of the DHR, it is still has no replacement, it involved 10+ people for about a year, and Bill and I respect that. Whether or not it is logical, the DHR is our baby and we will nurture it. It has 1600+ content items, to update and upload those is a major task.

    We are contemplating reverting to non-DLS hosting for the updated content, It seems to be our only recourse.

    Peter
    I5-6 Core CPU. 1070-Ti video card, 16GB memory

  4. #34
    Join Date
    Nov 2008
    Location
    Crewe, Cheshire, UK
    Posts
    15,822
     

    Default

    Quote Originally Posted by narrowgauge View Post

    We are contemplating reverting to non-DLS hosting for the updated content, It seems to be our only recourse.

    Peter
    If this is related to getting round new standards, that won't stop things appearing as faulty if installed from a third party site, TANE and TRS19 will pick it up on importing if say a TANE build, TS12 3.5 and can still be uploaded at present so long as they are not faulty at that build.

    Trees, won't help much with TRS19 but have a look at https://forums.auran.com/trainz/show...wn-vegetation) Jan has created drop in meshes to update Billboards so they work properly in TANE, not so good in TRS19 but better than doing nothing plus Harold (hminky) has applied the fix with permissions where required to a lot of billboard stuff, not sure but may be some of Dave Drakes?

    IM's can still be used for content in TANE and TRS19 even though regarded as legacy assets, many older assets actually look better in TRS19, if I wasn't tied up with my own Lynton and Barnstaple project and the CRG, I'd offer to help in some way. I'm staying clear of PBR until I've finished the route otherwise it will never get finished, building it TRS19 was probably not the best idea but I'm stuck with it! My Ffestiniog Route took 8 Years, 7 to create and 1 to update all the earlier stuff with lod where required and normal mapping for my ground textures so it worked in TANE ended up with two versions one for TS12 and one for TANE so I feel your pain!
    Malc


  5. #35
    Join Date
    Nov 2006
    Location
    Netherlands, for now in, Montreal, CA
    Posts
    5,205
    Blog Entries
    1
     

    Default

    Quote Originally Posted by narrowgauge View Post
    Martin
    ...The insistence on cutting off earlier content is akin to burning all books before a set date.
    hmmm, I think it is more like putting the old books in a separate room that requires a special code to open the door. In a Trainz context, that code would be a previous version.

    Naturally updating is your call.
    I'll have to dig up my copy of your DHR to see what it looks like in my TRS2019. I generally pay attention more at the operational aspect of a route. While the look of the route is not unimportant, I usually use my imagination to fill-in any gaps, especially if those gaps are acceptable compromises because of technical limitations.

    As for those embankments, they are rather visually jarring. I've made a few tracks with built-in embankments to fill-in the gaps under the track due to uneven terrain. I gave up on trying to have them blend-in and instead made them look like extended ballast. A lot easier to manage.


    70337:
    TRS19 Platinum, build 105100, Win10 Pro 64 bit, i7-7700 3.6GHz 16 GB, GTX 1070 Ti

  6. #36
    Join Date
    Apr 2010
    Location
    United States of America, Oregon, Bend
    Posts
    67
     

    Default

    While on the subject of meshes, I am totally surprised that someone with a few more skills than my own has not come up with an embankment mesh for roadbeds that dynamically assigns the underlying ground texture as a texture for the fill in the embankment. This would certainly improve the appearance and merge of the embankment into the ground where native fill material was used for the embankment. It would also eliminate the need to preconform hillsides to the 20m grid in a lot of cases.

    Robert704
    Tacoma & Eastern RR (v)

  7. #37
    Join Date
    Nov 2006
    Location
    Netherlands, for now in, Montreal, CA
    Posts
    5,205
    Blog Entries
    1
     

    Default

    Quote Originally Posted by Robert704 View Post
    ... dynamically assigns the underlying ground texture as a texture for the fill in the embankment. ...
    Robert704
    Tacoma & Eastern RR (v)
    Is that even possible? It sure would solve a lot of other visual anomalies. Which mesh should be assigned? The first, last, the majority if it crosses over many different meshes?


    70337:
    TRS19 Platinum, build 105100, Win10 Pro 64 bit, i7-7700 3.6GHz 16 GB, GTX 1070 Ti

  8. #38
    Join Date
    Nov 2006
    Location
    Newcastle NSW Australia
    Posts
    6,384
     

    Default

    Quote Originally Posted by Robert704 View Post
    While on the subject of meshes, I am totally surprised that someone with a few more skills than my own has not come up with an embankment mesh for roadbeds that dynamically assigns the underlying ground texture as a texture for the fill in the embankment. This would certainly improve the appearance and merge of the embankment into the ground where native fill material was used for the embankment. It would also eliminate the need to preconform hillsides to the 20m grid in a lot of cases.

    Robert704
    Tacoma & Eastern RR (v)
    Even if code or script for that was possible, and leaving aside the copyright issues of using other creators’ textures, getting a 3D object to match ground textures is really hard. Trainz seems to render the 2 types of asset very differently even if they use the same texture image. It requires a lot of careful tweaking of UV scale, colour tone, brightness and specular reflection on the 3D object to compensate for the differences. Then there is the scale and rotational blending effects in the ground texture which are variable and therefore even harder to match with a 3D object material. I would say it’s practically impossible to make it work automatically.
    Last edited by Dinorius_Redundicus; January 20th, 2020 at 04:22 PM.
    T:ANE SP4 build 105946 and TRS2019 SP1 build 105096
    Win 10, i7 5820K, 3.3 GHz, 32GB ram, GTX 980Ti, 2x512GB SSD

  9. #39
    Join Date
    Nov 2006
    Location
    Burnside, Adelaide, Australia
    Posts
    5,433
     

    Default

    Martin

    At the momemt the DHR runs without errors using the content on the DLS with build numbers around 2.8. We need to change the texturing of many of Dave's items to compensate for T:ane's brightening effect. To do this, and upload, requires what are effectively unnecessary changes. If it works in T:ane, why is it required to do anything other than add the version to the existing KUID. That is all that is required to differentiate it from the original.

    If you feel we may be pirating his content, we have permission to do this and are acting as his repair agents.

    Peter
    I5-6 Core CPU. 1070-Ti video card, 16GB memory

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •