The official "NEWS FROM THE SOCIAL NETWORKS" thread

Important point about the brakes...it helps to own a Ricomotech pedal set though Owning a Fanatec v3 set, braking in non-ABS cars has become shall I say easier to anticipate...brake lock isn't so sensitive anymore, there's a bit of play before real hard lockup. That said, Nicki clearly proved the physics engine...a car class he's familiar with worked for him because it was made right.

 

Comparing to RSR is dumb indeed

- Default setup
- Was talking all the time
- Racing with ton of AI cars in the way
- RSR times are done on empty, optimal track
- Damaged car
- Or dunno, maybe he didn't give a f? And wanted to have fun, unlike RSR types who were probably tryharding with sweaty underpants and restarting their laps for hours, until they finally hit record

But he did iRacing, so familiar with sims, and steering with visual feedback/not feeling G-force and all that.

But I wonder why he says laserscan is not as accurate as iRacing, how can it be not accurate? Laser is laser. Or did the real track change after it was scanned in AC? As it's older track

 

If either title is using the actual full point cloud from the scan, I'll be surprised as ****. They tend to take up quite a lot of space, I've dealt with a rather small and inaccurate point cloud and even that was several gigabytes. That was ages ago though, so maybe working with full clouds is more practical these days.

If they're not using the full data, then it becomes a question of how is the data interpreted and simplified for practical use in the sim. Also like you mentioned there is the possibility of track changes between scans. Or the driver imagining things.

 

Rsr times are done in more then optimal conditions especially in gt3 times at min temps for example. Could be any weather setting and grip levels used, also as any good racer tends to prefer, he wanted a bit of a challenge getting through a field rather then solo hot lapping out front. He clearly shows he has solid driving ability and real world skills that clearly work in AC as they should.

Im sure with a bit of time and sime tuning he would be able to compete with the quickest guys in AC very well. Race wise and consistancy as any good pro driver there would be very few that match his level and very few of the "rsr specialists" would be properly competitive in a race even if they can fluke one quick lap from 100 attempts.

 

Well that must have been weird document format, or perhaps crazy dense point resolution. Also how do you know that it was inaccurate ? And what is inaccurate for you ?

Pointclouds definitely can be differently processed. Roughness could be augumented (I really doubt actually if any scan ever would be "too smooth"), or decreased. Because there would be a few artifacts in a point cloud anyway. But I bet those would be nearly insignificant for proper ground level scan. I also don't know how high is the resolution for those scans. It is fact, that in game this mesh doesn't have to be insanely highly detailed, because physics refresh rate wouldn't pick them all anyway, AFAIK AC has one of the higher physics engine refresh rates. The differences also could come from the way how suspension is simulated in the car, or in AC in general. It is most likely that the end results simply becomes slightly different during the optimisation.

There could be many reasons. But I suppose only racing driver or mad enthusiastic person about that track could know.

In general overall AC phylosophy, seems to me, is more about softening rough edges, rather than making them harder, when it comes to choice. Less annoying things.

 

I think he was referring more to differences to the current real life track probably due to age of the track data. As in real track has had some changes over the years since the AC track was made

 

I think people also somehow get the impression that 'laserscan' automatically means 100% accurate. Not the case. Laserscan provides a lot of data that still has to be translated into a 3d model in a game/graphics engine. And that translation is important.

 

Thats for sure, but what is inaccurate ? If we go full phylosophy mode, and will keep zooming and zooming - it is never accurate.

 

Could have sworn I typed relatively not rather... Scumbag brain.

Well I know because I worked with it to create our virtual scene.
But I should clarify that I meant inaccurate in the sense of not very dense. Positioning of the points was probably on point ( ), so it was accurate in that sense. And inaccurate to me is that I think could have punched a hole in a wall the size of my fist and it might not have actually showed up on the scan at all.

I seem to remember seeing posts about Kunos' scans being in the 1cm range, but could just be pulling that number out of my ***, or remembering numbers someone else pulled out of theirs.

The data certainly could have been compressed down a lot, quite easily too, but storage is cheap and we were plowing through hard drives with footage anyway so nobody batted an eye for a few gigs of scan data when you have several hundred gigabytes of raw video files Bigger issue for us was actually working with the data, when changing perspective of the cloud happened at <1fps, and that was on the two heavier rendering machines we had, which should have been churning out proxies out of the raw footage and not have one of them stuck creating virtual scenes for post. ^_^

The whole thing was a bit off the cuff anyway, we were sort of piggybacking someone elses scans... a case of "hey since you guys are doing that thing for those guys... how much would it cost if you did this other thing for us, I mean since you're already in the area with the gear and all?" x)

 

"Sure, it might not have as many cars or tracks as most games on this list – and the menus and multiplayer are rough around the edges – but for pure driving enjoyment, Assetto Corsa is hard to beat."

^This guy gets it.

 

Those bumps and jumps with the 405 are different to what happened in the video with the crash.

In the Peugeot testing video the jumps are rather more gradual where front and rear are pretty much loaded and unloaded rather equally. The dangerous ones are where the front tyres are already off the ground before the rear tyres even hit the hump/jump.


In the crash video you have a single force at the front that pushes the front up already while the rear is still traveling parallel to the surface ahead of the jump, yet at the point the rear tyres hit the hump the front tyres are already in the air so there is only the inertia of the car that acts against the rear that gets pushed upwards.

To not crash there you either have to drive slower or you need to manage to rotate the car (= shift the weight distribution) before the hump by unloading the front tyres before hitting the hump because then the front tyres hit the ground earlier behind the jump. The car will still hit the ground first with the front wheels, but the whole car will be at a flatter angle so it doesn't frontflip. Although this probably is more a theoretical technique since it looks like the car is already going very fast in a high gear. Unloading the fronts tyres is more of a technique at lower speeds.


I'd say, driver (or co-driver*) error by being too fast at that point.

*I don't know how detailed road-maps and notes are on the Dakar, so couldn't say if the co-driver can warn the driver about such an obstacle

 

Always amusing when a sim racer talks trash about a real life race car driver's sim racing abilities. I couldn't care less if he ever beats the fastest times on RSR, this dude does this stuff for real.

 

Suddenly setup matter, using RSR is dumb & or people tryharding restarting for hours. While I usually read the complete opposite in the past... It is fascinating for sure.

 

Oh, so suddenly it makes sense to compare in traffic, race lap times to no traffic hot lap times? BWAAAAAAAAAAAAAAAAHAHAHA hilarious!

 

Remember too that Nicki might not know how to accurately express everything he sees and feels in a sim into English too. Some of what he said I take with a grain of salt, both until he plays the sim more and comments more about it. Also, some of what he said I chalk up to how the trackside objects appear, not the accuracy of the scan. He pinned my comment concerning the track...I believe Kunos took the track from their earlier project, FVA...how that impacts laser data I do not know, but for certain the track has changed in appearance over the years, and resembles an iteration Nicki's probably not driven at in RL. He construes this as not as good as iRacing, perhaps.

 

Nurb GP is one of the oldest tracks in the AC library while it is one of the newest in iracing.
----------------
Teaser shot posted on FB from some mod group that is converting a dusty old rfactor mod to AC.

 

I really need Lancia in AC!

 

Obviously to get the most out of a car setup matters, however RSR or any such hotlap comparison leaderboard should be forced for all to use identical track conditions. It doesn't have to be optimal but the same for all, the people that purposefully look for advantages in weather setup be it wind and temps do so because they cant compete on a level playing field. This is one thing that likely brought down RSR and why its being abandoned after a good effort by all involved.

Still comparing hotlaps in static conditions mean very little beyond outright possible pace, when it comes to racing and even setting a proper qualy lap in traffic with varied conditions and limited time then it shows who the better driver is with all the same conditions. Then add to that being able to race close to error free and close to your limit while also being clean in traffic and very aware is another thing. Funnily enough many I see join servers and when they realise they are a bit off the pace, or if they make an error they simply leave. Look at regular pub server sprint races with 20+ in a basic combo less then half the field is still going 5 minutes in and less then 25% would finish a 10 minute race on the lead lap.

I would love to see @MrDeap in a server one day, after all he says how easy sims are. They take no skill and he gives the ultimate guide to help people with no skill be even less skilled, maybe its just because he used to always claim anyone faster then him is a cheat. This is what someone with lower skills that thinks they are at the limit usually says, they dont know how to see where others may be faster or how they can improve themselves. Its easier to simply say they are the best and everyone faster has no better skills but just cheats and uses cheat setups.

 

@pankykapus could you please separate textures for headlights and other glass panels in the LaFerrari? So we would be able to replicate this:
without having yellow windshield and backshield.
Thanks