L&N Knoxville Division route - roadbed flicker

[ftldave]

Just installed the L&N/CSX Knoxville Sub route. Its roadbed textures along the route have lots of flicker throughout the entire route, running on ORTS 1.5. When I say flicker, grass textures under the roadbed flashing in front of the locomotive as it progresses down the line. Appears between double tracks, too.

Probably the first time I've run a route that uses MTracks. Tried turning off Model Instancing in my OR settings but no change. And haven't yet experimented with my NVidia gpu settings, but all my other many routes run A-OK, beautifully, with no roadbed flicker. I've seen mention of similar problems in OR in the past with UK fine scale track.

Just me or anyone else running the L&N Knoxville Division route seen it?

The route:
MSTS/ORTS Route--L&N/CSX Knoxville Sub
Name: kdsub_wrap.zip
Size: 323,527,740 Date: 12-25-2020 Downloads: 714

[slipperman]

Hi Dave,
The 'flickering' in this route is nothing like that experienced with (unfixed) UK Fine Scale track where it's the sleepers which are flickering. With this route it appears to be some of the textures on each side of the track.

I am completely ignorant of the cause, but could it concern LODs?

Cheers,
Ged

[eric]

This might be an issue with 1.5 -- it's been reported elsewhere on other routes. Possibly not handling mipmaps correctly...

[Rick-MONON]

Do any of flickering shapes have VERY...small texture files????

The L&PS grass would flicker because of the small texture size......

[spoony1999]

I wonder if it’s similar to the roadbed flicker/sparkle I was getting when running Milw RMD East and West earlier in the week? It’s the only route I’ve had the issue with. I was running OR 1.5

[scottb613]

I wonder if it’s similar to the roadbed flicker/sparkle I was getting when running Milw RMD East and West earlier in the week? It’s the only route I’ve had the issue with. I was running OR 1.5

Hi Folks,

Yeah - ORTS seems to have major Z-Fighting issues with everything that has textures in close proximity to parallel faces - we get flickering track, vegetation, buildings, signals, etc etc etc. It's reported that the 32-bit ORTS doesn't have it as bad as the 64-bit ORTS. You could try 32 bit and see if it helps. The quick fix for this specific instance - is - if you plan on using the route - delete all the roadbed texture sheets and paint a new ROW - in TSRE - directly on the TERRTEX. I could probably do 100 miles in an evening with the tools in TSRE - it goes really fast and looks far better than textures placed under the roadbed.

Regards,
Scott

[ftldave]

The quick fix for this specific instance - is - if you plan on using the route - delete all the roadbed texture sheets and paint a new ROW - in TSRE - directly on the TERRTEX.

I think the quick fix I would prefer will be to roll back to ORTS 1.4. Just retested the L&N Knoxville route with ORTS 1.5.1 and no change in the reported scenery problem. ORTS 1.4 was the version that finally renewed my interest in the sim, so I will stick to what works ... for me.

The .1 upgrade changes don't seem to be documented anyplace I can find, unfortunately, after checking the ORTS site and Elvas Tower, except for a brief mention about a 3D cab problem fix that prompted the .1 update. Anyone have any insight as to what else, if anything, the ORTS 1.5.1 update brings? Info about 1.5 persists, not 1.5.1, at least from my searching efforts.

[Kuckoo]

Interesting. I've been seeing for a while now exactly what you describe but on the Monon route. The ballast on either side of the rail flickers like crazy, as does the level crossing textures. Also, the windows on the locomotives behind me as I look back (in 3D cabin), they're flickering as well, unless the locomotive passes through the shadow of a nearby tree or building.

It does look like Z-fighting as parallel textures are constantly trying to overlap each other.

Began seeing this since, I think, shortly before v1.4 released. Using only the weekly testing versions, and the issue has been persistent.

[geepster775]

I created a consist with one of Tyler's 3D cab equipped SD40s and a Trainsimulations recent SD45 with 2D cabs and a rear cabview.

On the IC route, when the IC SD40 is leading, I get flickers ahead (whistle posts, crossbucks, etc) and flickers behind (trailing unit's rain wiper)

If I changed the consist, putting the TS SD45/cabview up front, the flickers in either direction did not exist.

On the MRL route, I ran the MRL activity with the 3 SD40s and did not see any flickering, when looking forward or rearward with the default supplied cabviews.

As for the L&N route, back in 2021 I started a project to optimize it for Open Rails. Terrtexes were changed, track textures were changed, transfers used for ballast were removed to be replaced with painting via TSRE, route-embedded curve squeal sound sources removed (OR does that in the code), and the original 'gantry' trackside bushes (sourced from who knows where?) removed. This effort was put on hold when news came out that the route might be extended again to Etowah.

So when I pulled this half-revised copy out of storage recently and ran it under 1.5, and viewing it through current-style TS cabviews, I was seeing very little flickering. The only remaining flicker-producing elements were secondary backroads and rock walls leading to tunnel portals that were laid using transfers as well (and which would eventually be replaced).


Are we reaching a point where inconsistencies in crowdsourced content (deploying ace vs dds, globally-sourced unknown-author route ingredients, third party cabviews, etc) is proving not as uniform or reliable?