Laser scanned tracks

Hey all, just wondering what tracks in AMS2 have been laser scanned, or is there a list somewhere?

 

Does it matter?
But I guess that at least the Northsleife is scanned - because I read somewhere that the company behind this track got it scanned themselves and then sold the raw scan data with some sort of licence (for money that is ).

 

You can continue from ....... Which tracks are laser scanned vs not?

 

Yes.,
When I drive acc for example, I'm driving on the actual tracks. I feel every bump and kerb that the real driver feels, and if I watch on TV, I feel I have driven the actual track.
Without laser scans, the character of the tracks is not captured.

 

OK - if you say so.
But for me there is no sim where I drive "on the actual tracks".
They all feels that its obvious that its only video games.
But when Im lucky there is maybe some resemblance to something we could call "the real world".
OK?

 

It will be drfferent for everyone. There is just more detail in laser scanned tracks. If we are racing simulators, then it would be nice to simulate the tracks too as much as possible, with laser scans.
Not only will we feel more bumps and so on, we will be driving the actual bumps and kerbs etc from real life.

 

After reading this thread as well as the earlier thread you linked, it would appear that no one has offered anything close to a satisfactory answer to a very relevant question. I have been operating under the--perhaps faulty--assumption that many, if not most, of the tracks in AMS2 had been built using either laserscanning or LiDAR data, with the exception of historic variants. It would be nice to have an authoritative and definitive answer in this matter. Simracing may be a niche enterprise, but all the more reason to acknowledge how AMS2 compares with its competitors.

 

All the Premium tracks are developed with the aid of point cloud data (from either laser/lidar scan). Most of the MSV tracks follow the same method, but I haven't found any statement from Reiza regarding Cadwell Park. Azure (Monaco) was originally pretty much the same as in Project Cars 2 but later got an updated track mesh based on laser scan data. Jerez was mentioned in the Oct 2019 dev update as being based on lidar scan data. So the list of confirmed tracks is at least the following:

• Azure
• Brands Hatch
• Buenos Aires
• Córdoba
  
• Donington Park
• Hockenheimring
• Jerez
• Monza
• Nürburgring
• Oulton Park
• Silverstone
• Snetterton
• Spa-Francorchamps

For the Argentinian tracks, it's unclear whether Renatos statement below is referring to all Argentinian tracks being modeled after laser scan data, or if it's only the 2 tracks where they scrapped the AMS1 mesh and started over:

There could be more tracks to include on the list, but these are just the results I found by searching for forum posts by Renato Simioni that included either "laser" or "lidar", in addition to reading Steam Store DLC description texts.

 

Not necessarily true. While laser scan is a powerful technique for surveying, it is not a be all end all solution to capture a terrain. There are other methods of track modelling that involve 3D models, CAD drawings, photogrammetry or even traditional theodolite surveys that still can produce good results.

And there are ways that laser scans can utter screw up that you get complete nonsense because of the data collections, software issues or mistakes in modelling. Or laser scan tracks that are poorly cleaned up that resulted in a rougher surface than it actually is which Kunos is known to do quite a fair bit. FM7 tracks are laser scanned apparently, their tracks are one of the worst iterations because how overly smooth are their tracks. Try their Sebring or Spa then compare to rf2 Sebring, ACC/AMS2 Spa, the difference is insane.

 

Been a track creator and used lidar and photogrammetry point clouds, I can say that the amount of post processing is paramount; been able to generate a clean cloud that preserve the geometries and surface features to the cm or mm is challenging.
And what make it even harder is the massive amount of data you're processing 500million points easily for a car race track, so compromises must be made.

 

Is this list still valid? Was there any update on any other tracks?

 

Is ACC a reference? This is not a real car on a real track and therefore are missing lot's of parameters, what about the brakes on ACC feel they natural to you? In my opinion laserscanned track or not I have not the experience how it feels in real life in a real racing class car, do you? Just curious

 

I can't help feeling that if you can't tell whether it's scanned or not then it's good enough. The only track in the game I've driven at pace is Oulton Park, and I can vouch for the fact that every bump and dip and detail is spot on.

 

The difference at Interlagos between AMS2 and LMU is huge. LMU one is laser scanned and it feels amazing. You can abuse the track's limits on AMS2 but not on LMU. It's punishing. I hope they update some non-laser-scanned tracks with laser-scanned ones.

 

I would be surprised if the AMS2 version isn’t laser scanned

 

Bear in mind there are multiple ways to get laser scan data. The easiest (if available for the location) is the Lidar data from a plane flying with 50cm point mesh. It's accurate enough to show level changes and exact position of the corners, buildings etc, but is not detailed enough for minor details.
This data is often available from the respective local authority, some of it (lower resolution mesh) can be available for free.

It is possible to get a more detailed scan by running a device from a truck driving on a track, which is contains a lot more, but such mesh won't have points that are not visible from the camera pov.

Both are completely different in quality and both are laser scans. So even with laser scanned tracks there may be differences based on the scan type, quality or even date the surface was taken (resurfacing etc).

So just because they differ doesn't mean one isn't laser scanned.

 

Isn't laser scanning 'old fashioned' now, replaced by drone based photogrammetry, which is quicker and cheaper? Could be wrong as I am on many things, but from what I've read it seems this way.

 

So the drones take pictures? Sounds like you lose a lot of detail

 

The opposite I think, photogrammetry allows a greater resolution than laser point scanning. I'm certainly no expert, just an enthusiastic amateur, but I think this is the sort of thing they are using these days:
3D Photogrammetry and the future of gaming - Aerometrex

 

Photogrammetry is exactly what I thought and it doesn’t really give the same details as LiDar