QUESTION How long does it take for a basic model?

[andyl]

I'm competing in the Vincenzo & Sons Bristol MX5 Challenge in 2018. It's an 11 round championship run alongside the Bristol Clio Cup. I'd like to get some practice in on AC, but whilst there's an MX5 to drive, none of the tracks are modelled. One benefit of being in the UK is LiDAR data of all but one of the tracks is available in 0.5m resolution. I've read lots of guides, but I'm still not sure how much effort it'll take to make a track that's representative in terms of track layout. I'm not worried about how pretty it is, just as long as the corners are accurate enough that I'm learning the track.

Alternatively, anyone fancy taking on building one of the tracks?

 

[mantasisg]

I don't know those tracks, but with a Lidar data file in hands and with a know-how then it would be quick. But there are many tracks...

Projecting a layout mesh over the LiDAR mesh (using actual Lidar Mesh and intensity data image only as a guide for layout, elevations, cambers and more significant undulations). Using a spline and as simple as possible elements, which would mean that track would have almost no appearance except road and very limited offtrack surface space and super basic approximate kerbs, maybe simple box objects to suggest key objects if some are used as brake refferences or something.... Preparing to be usable in AC...

Then IDK,.... 1-4 days on a track, depending of care level, size of the track, hours spent per day (lets say 4-5). Also if everything would go according the plan.

I have used actual LiDAR mesh for Goodwood. Received some fire for it, and it was quite time consuming. Projecting mesh would be much quicker. But you wouldn't get the very fine details. However, you'd also avoid artifacts from lidar mesh if there are some.

I don't really have time to help directly, sorry for that.

BTW, I was really surprised that Goodwood LiDAR data was available publicly at least for a couple of years. The data itself is from 2011.

 

[andyl]

Thanks. So 20 hours might get me something vaguely drive-able? I've probably spent 3 just reading guides/forums and the same again watching YouTube videos! Probably opening a can of worms here, but BTB/RTB as the best route forwards?

I was driving the LiDAR Goodwood track last night. It's a wonderful track but I've not actually driven it outside AC, so I can't say how faithful it is to the original. Being drawn from LiDAR data, I'd hope it's rather good. I have a friend who instructs at Goodwood - I wonder if I could convince him to stop driving real cars for a moment to test out the virtual track...

 

[mantasisg]

Well getting to know the things takes lots of time, Good beginning also is important to the end result, so if I would have a slight bit of hope that the projects would evolve anywhere further in future. Then I would try to make better start. It is very time demanding even when you know things.

I have zero experience with RTB or BTB, but I don't think that they are good for something what has to be precise.

My Goodwood should be quite close to real thing. I have slight doubt about few bumps in the straight between Madgwick and Fordwater, and then few little bumps after going up from Fordwater before braking into noname, but they don't cause me discomfort, just perhaps they shouldn't be there. Also a bump right before turning in to St.Mary's curve - that one is there for sure, but IDK if not too strong in my model. Would be great if your friend would ever consider trying it. That would be a real deal ! They could use a simulator to build some courage for fresh people, or to take it away

Maybe I could look at Castle Combe if so, but I don't promise.

 

[andyl]

I can certainly provide information from Castle Combe as to accuracy.

I'll have a chat with him about trying out the simulator. If that fails, I have another friend who has his own simulator and is vastly experienced in the real world. I'll find out how much experience he has of Goodwood - I know he's competed there a few times. I can point him towards your track.

 

[mantasisg]

I would be very happy, I'll keep my expectations low though. I have already had one similar situation, and I have had to admit that it was not going to happen.

We could talk about Castle Combe later, I could also try to do some kind of step by step instruction what I would be doing in order to get it going in AC as a basic version. I can't imagine how long the instruction would turn out to be, and how practical it would be - just a wild idea now haha

Will be looking forward

 

[andyl]

I've fired him an email to ask him if he's running AC (he's been focussed on rFactor previously) and if he can give it a test. He's only 15 minutes from me, so I could invite him down to try my setup if he's keen. He races a fair amount at Castle Combe too.

 

[andyl]

Any instructions gratefully received, BTW...

 

[mantasisg]

I forgot to mention that I am working with Blender, it is free to use. But it is very very functional. Also IMO more comfortable to use than 3Ds max. So you should learn using Blender if you'd like to get better use from my help. It is not difficult to use once you get the basics.

You could send me Castle Combe LiDAR data file if you have it, I could do few glances at it when I'll have more free time.

 

[andyl]

Thanks to the Environment Agency, LiDAR is available here. I've not actually checked what data is in there yet, but that tile has DTM/DSM (.asc) and Point Cloud (.laz) data from multiple years down to 50cm tile resolution on the DTM/DSM.

Update: Now I've downloaded and opened the LAZ file, it looks like there's very little there and nothing of the track itself. I guess this means Google Maps terrain data is as good as it gets unless I can find a reasonably cheap LiDAR device! Disappointing...

 

[mantasisg]

Yes. It is getting close but not quite... Track layout looks very similar to goodwood. If nothing then google earth.... And if lucky - a map with isohypses. Elevations seems all quite gradual and subtle. I noticed that there are tons of quite rough bumps

Could be fun track.

 

[andyl]

It sounds like someone else is close to completion on a Castle Combe implementation. I've posted elevation data from GPS here.