Nordschleife - laserscanned?

[Simon Kaminski]

Hello, I have seen in the launcher files, that we get teh Nords.
(X:\Games\AC_TechPreview\launcher\themes\default\graphics\64\track-ring_icon.png)

Laserscanned?

Ps: I posted this in the AC Forums but it got closed. Trying my luck here now hehe.
http://www.assettocorsa.net/forum/index.php?threads/nordschleife-laserscanned.506/

Feel free to post presumptions

 

[GamerMuscle]

Bear with me on this ..

On pc there are no other games that look as nice as AC and include multiple ferrari cars to drive which gives ac a huge selling point to more casual mass market users.

AC is also incredibly playable with a keyboard and contains single player again making it a good perchise for casuals.

Look at the numbers of people that pick up games like farm simulator , train simulator despite having no interest in them but just got them because they were in the steam sale.

The point is given that kunos is an awesome company and they legitimately care about the getting the "art" of what it is to drive right and the fact they will likely make a killing from steam sales. If any company is going do the ring properly / push to get it into there game its going to be them.

Not sure if having a huge DLC price would be worth it as they would probably make more selling it for less but to more people , given that I suspect AC will be far more mass market than i - racing there would be no need to charge $15+ for a track.

Regardless from the TP AC for me has revealed a whole new depth to driving simulators after RF2 I'd mostly given up hope , sure RF2 and i racing both have some nice cars that can be quite drivable and fun to compete with , but allot of the time they just don't seem to line up with the reality of what it is to actually drive, especially in terms of the subtleties and the core experiences of the physics. Where as AC does. ( at least with the lotus ) who knows the rest of the cars might be awful !

 

[Stereo]

AC is also incredibly playable with a keyboard and contains single player again making it a good perchise for casuals.

Not sure if having a huge DLC price would be worth it as they would probably make more selling it for less but to more people , given that I suspect AC will be far more mass market than i - racing there would be no need to charge $15+ for a track.

I think the casual group is less likely to be interested in content like the 'ring to start with - it's sort of legendary to racing enthusiasts, but just a long track for more casual drivers. For head to head multiplayer, a 1-2 minute lap means you're never too distant from other drivers so you're more visibly 'racing' even when you're a half lap off the pace. So I'm guessing as DLC goes, a variety of tracks that are easy to learn and be competitive on would be more popular. Although, if they can field the large numbers to make the 'ring feel occupied, it might be a good time.

I was also quite impressed by how it plays on a controller - steering's obviously not as refined but you don't feel disconnected from the car or anything.

 

[GamerMuscle]

I think the casual group is less likely to be interested in content like the 'ring to start with - it's sort of legendary to racing enthusiasts, but just a long track for more casual drivers. For head to head multiplayer, a 1-2 minute lap means you're never too distant from other drivers so you're more visibly 'racing' even when you're a half lap off the pace. So I'm guessing as DLC goes, a variety of tracks that are easy to learn and be competitive on would be more popular. Although, if they can field the large numbers to make the 'ring feel occupied, it might be a good time.

I was also quite impressed by how it plays on a controller - steering's obviously not as refined but you don't feel disconnected from the car or anything.

I agree with what you are saying but I think you might be forgetting that sometimes the more casual racer has less logic as to why they like one track over another or one car over another , most of there liking of one track or car is simply because they have heard or know about its history rather than how it drives.

A Ferrari fanboy/fangirl/casual fan , motor car enthusiast will likely have some over appreciation for the ring even if they cannot race or will likely not be able to make one whole clean lap at close to race pace. ( in the same way people drive to the real track to potter around in some crappy car)

Look how popular the ring is on GTR evo , or with console players simply because "its the ring" allot of more casual users wont even have a good understanding of side by side racing or maximising a racing line / spending 5+ hours on a given track to nail down the last 0.100 of time.

Its exactly the same in how a bog standard Ferrari would sell coppies of AC , its all about the brand and people that want to buy into that brand regardless of the actual factual driving experiences of Y game asset over X game asset.

Fortunetly AC looks like it will be an awesome game with incredible physics , but say it was a run of the mill average game the fact it had Ferrari and the ring would cause it to sell copies alone.

There are Game companies that make most there income by producing well polished but fairly average driving games that sell based almost entirely on the cars in the game rather than the fact the game itself "stands out"

To be fair to you casual might be the wrong word for me to use , I meant casual in the context of driving games not games as a whole. to me a "racing enthusiast" would be a "casual" when compared to a "sim racer"

Also to be clear . I'm not making the argument that people having some general fandom of a track or car is illegitimate , nothing wrong with people liking cars and tracks from history or proxy enjoyment. I can even see how people that don't drive or race could still have an appreciation for the ring as it obviously has an interesting history behind it regardless of if you race or not.

 

[Stereo]

To be fair to you casual might be the wrong word for me to use , I meant casual in the context of driving games not games as a whole. to me a "racing enthusiast" would be a "casual" when compared to a "sim racer"

Yeah, I was thinking more of people who'll buy it on Steam because it was greenlit and has good reviews. No doubt the ring will be quite popular, especially if it has laser accuracy.

 

[GamerMuscle]

Yeah, I was thinking more of people who'll buy it on Steam because it was greenlit and has good reviews. No doubt the ring will be quite popular, especially if it has laser accuracy.

I hope AC storms steam even if its with 100% casual players and people that never play driving games , sure there will be a bunch of noobs and crashers on servers but for proper racing it will give us more people that want to do league races.

The ideal will be having enough users so that if you ran your own server people just jump in then if you have people crashing / greefing you can kick them.

I think the sales on steam will give Kunos a massive boost which will likely translate to more content and features for the competitive simracer.

Of course the guys at kunos might just run away with all the money , I wouldn't blame them !

 

[William Levesque]

I doubt it would take that long.
Laserscan (AC style) would take one day tops.
And building the track i would guess a year at most.

Yea people say it will be heavy but since there are trees covering nearly the whole track there´s not that much that needs to be rendered on screen.

I'd take it between 3 days and a week to properly scan it, then a few months to do the rest.

Building a Ring from scratch takes around two-three years or something (With a handful people working very part-time), so when you already have the track itself with the scanning and people working full-time it'd take less time. They actually said it's 10 times faster with laser-scanning.

 

[Simon Kaminski]

I don't think it takes 2-3 years. They built the track for Project Cars from scratch in some months.

 

[Hampus Andersson]

People are over-estimating the time considerably. They won´t laserscan the worlds road net.

Just a track that´s about three times the size of SPA Franchorchamps.
Where the majority of the track contains nothing but trees as opposed to a place like Monaco etc.

If they average 1km/h per hour it still only take them 20 hours to laserscan the whole track.
Assuming they do some stops but go quicker then 1km/h then that time is probably cut considerably.

iRacing also scanned (with stationary scanners that you manually move around) 4 tracks in 4 days, on different locations.

5.8km - Suzuka
2.0km - Tsukuba
4.8km - Motegi Road
2.4km - Motegi Oval
3.7km - Okayama

18.7km in 4 days with stationary scanners. AC´s method is probably several orders of magnitude quicker.

 

[Bruno Pereira]

I don't think it takes 2-3 years. They built the track for Project Cars from scratch in some months.

Yes but Project CARS at one time had 4 people working at the same time on the track and the Assetto team is a bit smaller so either they scanned it or they have done it all by hand. We'll see when they decide to release the info.

 

[Mr Whippy]

People are over-estimating the time considerably. They won´t laserscan the worlds road net.

Just a track that´s about three times the size of SPA Franchorchamps.
Where the majority of the track contains nothing but trees as opposed to a place like Monaco etc.

If they average 1km/h per hour it still only take them 20 hours to laserscan the whole track.
Assuming they do some stops but go quicker then 1km/h then that time is probably cut considerably.

iRacing also scanned (with stationary scanners that you manually move around) 4 tracks in 4 days, on different locations.

5.8km - Suzuka
2.0km - Tsukuba
4.8km - Motegi Road
2.4km - Motegi Oval
3.7km - Okayama

18.7km in 4 days with stationary scanners. AC´s method is probably several orders of magnitude quicker.

The iRacing method is much better than a mobile scanner IF you are wanting to capture data out to the tree lines, behind armco, nearby buildings, access roads and everything else in between.

Driving around instantly generates problems. Imagine a 2d plane that is perpendicular to the direction of travel. At any instant that is all you can collect the data for. It is also occluded by fences a fair bit (and there are lots of those around the track), armco, concrete barriers etc.

So there is very little beyond the track/kerbs out to the barriers. Buildings, tree lines, portions of buildings that don't point directly at the track driving line.

So from an artists point of view, the static data gives you much more data. Chances are each scan is 50% overlapped with another, so you get both sides of objects so you can get a feeling for the overall form, not just the side of a building that faces the track, but not the sides, top, or back for example.
Looking down on the track from a tall mast also probably gives better quality scanning of the surface, so a bit more accuracy? Though wind can be an issue in these situations.


In any case, the drive along type get you good driving surfaces out to the barriers, but artists have a LOT more work than that to complete. All the stuff beyond the barriers takes time to locate, build accurately etc...


Licensing is probably the biggest thing for commercial tracks these days. There will be a rights holder for the data who pay the track to own it, then they will probably sell it on to those wanting to use it... the end user will also pay to the track owner too no doubt.

I'd not be surprised if a scanning company owned the rights to laser scans for the Ring.

Even if you can avoid that cost and build it using new techniques (like photogrammetry etc in a few years time), you'll still need to pay the Nordschliefe people money to sell your game with their track in it!
And going on their recent history, they probably want a small fortune for it. I can see why iRacing don't bother with it.


I think this is why the Ring for AC would be better as a community mod imo. Avoid all the silly costs associated with it.
I know for certain that a very good track CAN be built by avoiding LiDAR scanning and all that stuff.
A community project team could be an excellent Ring, with surface details good enough for anyone wanting to enjoy driving on the course with some realism... it just takes a LOT of time

Dave

 

[Hampus Andersson]

The iRacing method is much better than a mobile scanner IF you are wanting to capture data out to the tree lines, behind armco, nearby buildings, access roads and everything else in between.

Driving around instantly generates problems. Imagine a 2d plane that is perpendicular to the direction of travel. At any instant that is all you can collect the data for. It is also occluded by fences a fair bit (and there are lots of those around the track), armco, concrete barriers etc.

Yes that´s true, it´s also more exact, even more exact then what F1 teams uses.
But the point was, it´s faster. Much faster.

 

[Xosé Estrada]

The biggest difference (apart from bumps) for me between the Laserscanned and the non-laserscanned tracks are the elevations.

For some reasons the artist that create the non-laserscanned tracks are always too conservative with elevations in track and therefore the result is always way too plain compared with the original track.

While bumps aren't that critical in terms of fun, the elevations really change some tracks and while I don't like SIMBIN brands hatch for instance, I like iRacing one, that is more like a rollercoaster.

 

[Mr Whippy]

Yes that´s true, it´s also more exact, even more exact then what F1 teams uses.
But the point was, it´s faster. Much faster.

Yeah it is a lot faster... but for an artist the big issue is still the stuff beyond the barriers.

Given the choice I'd rather have everything accurate to work from than have the data end at the barriers or so and still have to spend time looking over satellite images and photos and stuff to make it all look accurate.

I can't really see the advantage to the 'fast' version unless you have some other data capture method to make sure the artist can also build the other elements quickly and accurately too. Building locations/sizes and other elements beyond the barriers are also very important for the drivers imo. If your track is laser scanner accurate, having landmarks beyond the barriers for braking markers or guides etc are just as important for me any way... hmmmm...


Xose, again, you can get good elevations without laser scanned. Just it seems before the laser data was widely used studios could make bad tracks.

I'm sure certain courses were very good and accurate in the distant past before laser scanning, it's just that laser scanned ones are more likely to be accurate assuming an artist can drop a mesh over the points nicely


I remember at the time the 'Ring for the Xbox PGR2 (iirc) was a complete joke by all accounts. The version we had in Racer and rFactor at the time (mod from GPL) was amazing in comparison!

There was never any excuse for poor track elevations, it was just cheap quick and relatively crappy track building. Maybe the studio didn't deem it important... but that is no excuse imo for any half simmy kinda game!

Dave

 

[Hampus Andersson]

Yeah it is a lot faster... but for an artist the big issue is still the stuff beyond the barriers.

Given the choice I'd rather have everything accurate to work from than have the data end at the barriers or so and still have to spend time looking over satellite images and photos and stuff to make it all look accurate.

I can't really see the advantage to the 'fast' version unless you have some other data capture method to make sure the artist can also build the other elements quickly and accurately too. Building locations/sizes and other elements beyond the barriers are also very important for the drivers imo. If your track is laser scanner accurate, having landmarks beyond the barriers for braking markers or guides etc are just as important for me any way... hmmmm...

So you think iRacing takes their scanner outside the boundaries of the track?
They use photos like everyone else for visual cues that´s very far away, like the city at Interlagos etc.

If the landmark is close enough to act as a pointer for the racer then it will be noted by the laser scanner.

Take note of the black dots, i.e. where the scanner stood.

Laserscanners will catch quite a lot of things. Even if it´s just an artifact of the real thing you can still have something to work with.
Not sure but i think most scanners have a range of 100-200m.
Obviously the dots will be further apart the further you go but still, should be quite accurate nonetheless.

The advantage of a fast version is obviously the time it takes to scan the place and start actually working on the track.

 

[Hampus Andersson]

from AC,

Notice that despite it being a mobile scanner, it still scans the tower to the left.

I wouldn´t worry about missing details. Seems to me if you can see it, the scanner will pick it up.

 

[Kevin Knorpp]

Horrible waste of resources to laser scan.

 

[Simon Kaminski]

Horrible waste of resources to laser scan.

Not. You can't get a track 100% accurate without laser scanning.

 

[Xosé Estrada]

Horrible waste of resources to laser scan.

Depends on what you focus on... if you want a fun track, yes, you can do that without laser scann. If you want it as real as possible, then is the best choice.

 

[Requiem84]

Horrible waste of resources to laser scan.

Obviously someone who never had the pleasure to race on a good laserscanned track and compare it with personal experience on the same racetrack in the real world. You can't believe how big the difference is.

There is a reason why a smart small team like Kunos Simulazioni use a lot of resources to Laserscan their tracks.

 

[GamerMuscle]

You can get both laser and non laser scanned tracks that are fantastic

But one thing laser scanning gets you is consistency and although it takes time to send a team to the track to scan it , it saves time when building the track V if you were trying to get the same sort of detail in a non laser track.

If you have ever done work as a 3D artist you will instantly realise how much of a speed increase you get when your reference materiel is of high quality and easily accessible.

This is why with fictional games good quality concept artists are important and concept artists often produce specific art and train in how to produce specific reference material so that the people doing the 3D work can get the model right and knock it out to a high quality faster.

In the same way Laser scanning allows the track artist to sit down and blitz out the core track layout , not having to fiddle with trying to get corners bumps and inclines just right an spend ages comparing reference images , as for the most part you can line up your mesh with the point data.

The decisions for the 3d artist becomes more a case of how much detail they should put into any given part of the track verses spending time trying to decide from some photographs / video / track walk , if its right to a basic level.

There are cost advantages and disadvantages to both methods , both methods can also work to a high standard.

In the end when you use point data that you get from a laser scan you will end up with a track that not only matches the real world better at a certain point in time , but it will contain subtle nuances that would not end up in a non laser track.

Given that you can get incredible small scale results from a Kinect sensor which is of very low resolution , I don't doubt the quality of point data possible with more traditional equipment.

Father more the possablits using a stereo camera alone are suprizingly good