You already know that you're doing a tremendous job, so I'll focus on the negative . Oh, pardon, on constructive feedback. In the 3rd image there is some visible "lack of smoothness" in the road/kerb radius. If it's not too late (aka a huge job to modify, AND assuming you still have the beziers used to generate the road) a trick that Lilski taught me some time back: - In the turns (from the turn in point to the exit point) select the edges and do an additional subdivision. This will make it look like a proper round corner instead of an N-gon.
Yup everything is set up to be editable. Array and curves modifiers etc … Got a global 1m per face on kerbs. Can definitely do a subdivision on more acute corner sections like that one …
Looks like you got it easy with standard kerbs all around at this track Working on Road Atlanta... it must have about 6-7 different types... recessed saw tooth, sausage style, flat, and this:
As a side note, I've seen that rF2 has licensed this track (n.b. Kyalami), so it'll probably become even more popular (..as much as rF2 can help, that is).
This is the one I really care for tbh (Prototype’s), but would want to see the data they are going to use.
Thanks man I lucked out on the kerbs, but I paid my dues in other areas! TBH, I wouldn't have minded doing additional kerbs, if they were there, its such a definitive thing. That's an intense set of kerbing on RA! Suspension breaker. Aerial laser scanning was done on the layout prior to this one, to carry out the engineering work. I would imagine Reiza would possibly be given that data? So they would have the full set of data for the 1992 track, about 80% of the new track and about 45% of the 1960s track from that scan. https://www.wsp.com/en-GL/projects/kyalami-race-track ^^ " A combination of mobile, terrestrial and aerial laser scanning together with high resolution aerial photography formed the basis of the virtual 3D model of the existing terrain and track. " Its AM2 I think Luchian is referring to. https://www.racedepartment.com/threads/ams-2-new-game-footage-plenty-of-news-in-july-roadmap.171638/
Comparing aerial photogrammetry and 3D laser scanning methods https://www.ee.co.za/article/comparing-aerial-photogrammetry-3d-laser-scanning-methods.html Their conclusion (deviation from the laser scanned data and the photogrammetry data): As shown in the earlier three extracts (see Figs. 2, 3, and 4), 90% of the points are no more than between 0 cm to 3 cm away from the surface. The average error is 1,5 cm, and the root mean square error is about 2,5 cm. If higher resolution imagery had been used, the deviations between the point cloud and pictures would have been much smaller.
UPDATE: The last of the perimeter buildings are modelled, AO baked and low res meshes in scene. Now to do the diffuse 52 unique perimeter buildings in total (including the castle). Another 10 or so track buildings. EDIT: And this is how Ive packed the textures. Atlased 16 buildings into a (currently) 4k texture, giving me 1024x1024 per building. 4 of these for the entire tracks buildings. (64 buildings)
That's definitely the plan! The news of the AMS2 track is exactly what's put a fire under my butt to get this thing done. Thanks man Looking at your random concrete post on @Nericksenna track is giving me fomo on that. I think I did see a post on RD that lilsky shared at some point on how to do it in blender, but Ive got so much ground to still cover just getting the basics done ... wonder how much I'd have to roll back to get something like that implemented …
That's exactly how it should be done! 4 textures for 64 buildings Forza rips are 4 textures per building Are the walls made in prefab sections? Then its definitely something to consider, we are too spoiled nowadays, repetition is so last year
Aaaaand once again I’m godsmacked in the face Perfectly done mate! Beautiful ! I too have around 60 buildings in total, how on earth you went so quick on these buildings I’d like to know your trick.