LiDAR circuit models

I would have to agree with you to a certain point. When you say that FFB doesn't matter...well I suppose that's a matter of preference. However, just for the sake of debate, the FFB can be very very useful in certain places. Places like where you drive your car around the track, where the racing line is, and which curbs you take or don't take =) All of this can and usually is coded and can be felt by the FFB wheel. Is FFB the most important factor? No, it's not. I would say that all of these things combined allow for the overall experience. If you can't feel it, though... at least for me, the experience diminishes dramatically.

For example, if it's coded in the software that the racing line has more grip than off the racing line, you are going to feel that in the FFB. If you decide to take a curb that you know will pull your car onto it and unsettle the car, you are going to feel that in the FFB. All of this depends, of course, how well the FFB is coded and how well the wheel responds to the FFB itself. I do agree that all of the things you mentioned, matter. I just don't agree that it's not about the FFB.

I will say this, though. FFB isn't everything. There's also a great article on FFB written by Leo Bodnar. I highly suggest everyone read it. It's difficult to understand at times, but it's a great read nonetheless.

All good points. good discussion!

 

Do you or maniak respect me to not suggest what I have to do with my money (assuming I have it enough)? Maybe I have not enough resources to race IRL. Maybe I am handicaped or so. For me maniak's has no rights to do such statements. That's why I called it as I did

Can't agree. We are using such content because most of us are not able to do the job. But I would welcome better quality/accuracy if available.

 

I see. Thank you for the explanation =)

@MaXyM

I would not assume anything about anyone. I only comment on what I know. I happen to know that the comment you made was uncalled for. Do I assume that you have money or not? No, why would I? Do I assume that you are handicapped? No, why would I?
There is a difference between assumption, and commenting on something you know to be true.

My original post went into details about my "opinions" on the matter. They are not assumptions, nor are they accusations. I would suggest taking anything anyone says on any forum with a grain of salt. As always I respect your opinions just as I respect anyone else's opinions on any subject. Take LesiU for example. We agree to disagree and in some cases we misunderstand each other. But I don't think we resorted to name calling. As I said, your comment was uncalled for, but you have the right to disagree with that, just as I do.

Cheers

 

Well, I'm with you. The main track where this is apparent for me, and I won't name names but just keep in mind I'm in Australia , I find it quite annoying that the track layout is wrong - not just elevation, but where the corners are and their angle, and I was able to confirm this in 5 mins with a quick search for some aerial photography of the track in question. I find it frustrating that people spent hours upon hours making the track, and doing a great job on it (looks good, great detail, works perfectly), but they stuffed up what seems to be one of the more critical areas. It's a difficult thing to fix now that it's made of course (unlike track makers leaving out pit entry/exit lines, blend lines, or not properly stretching timing objects all the way across the track), but there was a reference used at the start and it was wrong.

Thing is, I've never seen anyone else make mention of it. So while I would appreciate more accurate tracks (and I think you can get a long way there with aerial shots and a few decent onboard cams from races, albeit with a fair bit of time spent on it) and obviously you would too, I suspect we're not in the majority

 

Slightly OT but putting things into perspective.

As those who have driven a racing car in real life would know, if you want a better racing experience on computer, it's not the detail of the track that will bring you closer to reality. If a track is modelled to within close-enough detail, say 95% for arguments sake, then that's good enough.

What makes a simulation feel 'real' is g-forces. If any of us could afford a fully functioning dynamic seat, trust me, you wouldn't care less if the track was laser scanned or not.

Having said that, obviously the main things in modelling a circuit should be correct - corner radii, track width, elevation, camber, kerbing, significant bumps and run-off. The rest are just bonus and nice-to-haves.

 

To my mind all the iRacing tracks have a unique character and set of challenges. Which is what makes things interesting. Often in other sims I couldn't tell you what track it was in parts without looking at the surroundings. It all seems flat and pretty similar. And when I've driven an iRacing track in another sim it's generally been unrecognisable without a conscious effort to think 'this bit is supposed to be that bit'. Even the VLM Mid-Ohio drives completely differently to the iRacing version. Elevations contours are all different. Whereas, plenty of people have said after driving a real track for the first time that it felt really familiar after iRacing. Sure, learning the Nordschleife even from a console has helped real racers there for the first time to know how it goes, but that's a special case. Generally, I can't see how most sim tracks would tell you anything very useful in regards to the RL version.

 

You can't always expect great tracks from a builder did it as a hobby. There are very few tracks out there built to a quality standard of say simbin. Remember the saying. "you get what you pay for".

 

But Cdn, that's the point. The quality is superb.

The work in creating the layout (top-view) is pretty much the same whether you base it on a 2-inch-square diagram in a newspaper or get some good quality aerial shots, and I don't understand why it looks more like the first was used.

I'm not belittling or undestimating the work that goes into creating a track; I just don't see why you wouldn't take an hour or two to find a good reference if you're going to make a real track.

 

I actually know a few drivers that have said that the iRacing version is off somewhat. Let's not forget, though, that you are comparing a real company that uses laser scanned data and a mod group that does this in their spare time as a hobby. In this case what CdnRacer says is true. You get what you pay for. I'm sure that VLM would be happy to correct the mistakes they made in their version, and as Tim Wheately has said "just write us a check" =) Even the laser scanning data has it's flaws. It helps the developer tremendously, but I would hardly compare a company that spent millions of dollars on their courses with a modding group that is mostly made up of hobbyists.

I'm sure you meant no offense to VLM, though. But I think it's just a bit unfair to compare the two. But perhaps you were trying to make some other point and I missed it.

 

I'm sure it is not about being fair or not. It is about making a point that even the best tracks done without LS support are not as accurate as they might be.
That's why some of us (including me) want to have LS tracks - because it gives best possible representation of real ones. Currently there is no other way to get such accuracy (and it's not about even bumps). Period.

 

+1 i totally agree with you.
Personally i think a laser scanned version of a track give you the maximum fidelity; i know a simulation will be not the same as real racing but..we're trying to simulate with maximum detail as possible so, laserscanned track are the best available and..i would pay a lot of money for a good laserscanned track. IMHO.

 

You know, reading this thread ..... there is a common ground here. We all have a preference for high fidelity representations of our favorite tracks.

Lazza's post #25 triggers for me the reaction .... OK, fair but I'm I've never driven that track so I'm not going to really know there's inaccuracy. Perhaps the kicker in his message is that it could be resolved using only the same set of tools that were used to create the original. Would LS be worth it.

So will LS provide the best solution to get the turn radi and general track characteristics accurate. Perhaps!

Is LS the only way to achieve this. Perhaps not!

From a practical point of view, is LS feasible .... Probably not at least yet!

Cost wise, it is affordable ... No, at least not today

Would I play a premium for it ... depending on price, yes I would consider it

Is there an alternative .... yes

Would the general market buy LS today... probably not enough to get my price point above

If the question really is about achieving a 'high fidelity' track, I think this thread has pretty much already covered what that definition needs to be over the various posts .... corner radi, slope, kerbs, length, width, etc etc etc.

What remains is an discussion that is largely academic about the simulation of reality.

And with more work, as per Lazza post #25, there is indeed a suitable solution that will meet the vast majorities needs and expectations.

FFB and the rest of the subjects are all valid points too but they serve to complicate the LS discussion hugely [and for good reason]

That common ground still is common ground and reaching it, I feel is a lot closer than we think and does not require LS!

Motivation from the track builders will deliver it in my opinion and with respect to motivation, that money backing my consideration to buy LS if it were available, is there for donations to good track builders also. Surely, I'm not alone on that!

LS will likely deliver it too but the cost:benefit ratio IS the reason it's not in regular use today. Simple economics I say!

 

PLease say something about. Of course alternative means something which gives almost the same result. Not "kinda-like".

Maybe you are not familiar with real costs of LS. I know it is possible to pay it from private, nowadays. Considering a time spent on blind modeling from blueprints, google, photos etc It is realy not a hi price even for a modder (or modding group). LS is expensive because of processing which must be done after LS. This work may be done by modder (today you may get even ready 3d scene as output instead of point cloud). So only cost of LS alone must be paid. In that case team will save about year and half of time and get possibility to produce most accurate result possible. Just count how many beers has been drunk by LM team while building LM track for more than 2 years


Of course accessing the device for low price may not be easy etc. But in short, it is not such expensive as you may think.

 

Just to put it in perspective, the maintenance plan alone for the laser scanner (not lidar, remember) they use is over $25,000. The costs are probably more than you imagine.

 

I look at it this way: if I can't tell that a track's wrong from looking at real life onboard and having a quick look at some aerial shots, it's probably close enough (I mean, you compare a track that close and a laser scanned track, and you actually don't know which one's more correct ).

The example I'm half giving, having done several thousand laps of it in rFactor I can watch an onboard lap and see that the lines through a number of the corners are subtly different (but noticeable), which led me to go and find some decent aerial shots... and that confirmed quite a few inaccuracies. So while I agree laser-scanning should give the most accurate results (unless you get lucky, and do a lot of research ) you don't absolutely need it. Of course, if you don't use it, then no matter how close to the real thing you get some people won't accept it, but by the same token some laser-scanned tracks might not be as perfect as people assume.

 

Lazza, I get your point. So, by analogy, can you confirm that you are ready to drop rF and race using nfs:shift just because you are not able to say for sure which version is closer to reality? I'm sure you are not. You are believing that Shift is arcade when rf is the sim.
It's all about believing in some reference. We believed SBT that their Spa is perfect (I remember 'gps scanned' slogan). Now I know it is several meters off. So we need some better reference to race off and be free to say: it is like IRL.

Finally I will accept that LS is to expensive. If it is indeed - there is no way.
But I cannot agree with statement that we don't need/want this, just because track mad manualy is "almost" the same.

 

To me there are two separate issues here. Fidelity and simply being interesting to drive on.

If you're not going to drive the real life circuit, nor want to, say, compare notes with a real racer on that track, then you might not care whether a sim version is accurate or not.

However, it's the bumps, contours and elevations on the laser track that make it so much more demanding to drive on the limit. Even if the lines, visually, might look the same to you between a mod and RL (or laser) track, the driver's going to be doing different things. You actually have to work hard with the steering (lots of corrections) and throttle control to drive on the limit on a laser circuit -even on a straight in some cars - because it's not flat like most sim tracks. I'm surprised people don't notice how little steering movement there is in most sims compared with RL cockpit footage.

But I don't see any reason that a non-lasered track couldn't be made as demanding/interesting - it just wouldn't be accurate...which might not matter to you.

At the risk of a Clarksonian moment: does laser scanning have to be so expensive? Is it just because of a niche market. Couldn't we just low fly a Lancaster bomber with a search light each end or something. Ok...that might need some work.

And BTW, I'm definitely not belittling the work of modders - it's their commitment and achievements which prove how short of the mark iRacing, with all its resources, falls in so many respects.

 

And if you didn't know it was several metres off (because it was only ~30cm, and you didn't have any reference material to indicate the error) you could say it's close enough, because you wouldn't know it's wrong. So then it becomes an issue of whether you need to know for sure that it's exactly right (if you can work out when you can trust such a claim) or you can do a few quick checks, decide it looks correct based on all the info you can find, and go racing.

It's fine to want it realistic, and I can understand you wanting something better if there's an obvious error.. but trusting claims of what version is most correct (bear in mind with the number of variables involved you could have two different groups laser scan a track and end up with discernible differences) is much the same as, for example, trusting iRacing that their physics are perfect (which makes you wonder why a patch came out...).

As has been pointed out, tracks change their behaviour day to day, week to week, let alone any resurfacing or repairs that are done... and what's the point of having a laser-scanned, supposedly 'perfect' version of a track that has since been altered? You might as well just have a good manually-done version and say it's an excellent representation. The difference between it, and a millimetre-perfect and prohibitively detailed track would probably be negligible by the time you throw in a physics engine that can't truly replicate real life, controls that aren't exactly like those in the car, sensations that you can't replicate in your home rig, and of course all the forces on your body that will probably never be emulated.