There's a good recently released laser scanned conversion/rip of Laguna Seca on internationalsimracingI have to say first, that I am super happy, that i got laser scanned Nordschleife. I never had thought that a small team like S397 will do that. Cause of that I am very happy and surprised.
My votes for future modern laser scanned tracks goes too:
* 1. Laguna Seca
* 2. Spa-Francorchamps
* 3. Brands Hatch
Though not listed above in the top three tracks, but for what my heart really beats is to get Targa Florio for rF2. Probably not laser scanned, but maybe still based on a scanning technique that is very accurate.
Thanks for the info! Problem is illegal conversions/rips can only be used in private. If you want to race it online, no chance. Same for AC Nordschleife rip in the past. Good we have now a very high quality Nordschleife for rF2. Cause when new Online System comes, we can race it. Not possible for Laguna Seca rip.There's a good recently released laser scanned conversion/rip of Laguna Seca on internationalsimracing
https://forum.studio-397.com/index.php?threads/circuit-d´azur-ai-crash.65067/Quality tracks with no ai bugs
I would say a correct version of Sebring. The current laser scanned version has a surface nothing like the real track.
Sure the bumps are mostly in the right place, but the depths are way off, probably from laser shadowing.
What's your reference for stating this?
So your feel is that the real one is bumpier?Ergo, I've driven Sebring a number of times between 1982-2001. I've rode bikes there a few times, last time in 2017. Moreover, I support two IMSA teams with full access to their telemetry files (both MoTec users). One of the IMSA drivers uses rFactor2 as training tool, who also agrees.
As for the laser scanning procedure there are several ways to introduce errors.
Shadowing is somewhat common among concrete block pavement where pavement blocks are uneven with large chips broken on the trailing linear edge of the block when scanning at resolutions of less than 2-3". Sebring has hundreds of those potential error points.
Not all laser scanning systems have the same features such as 9 axis accelerometer internal adjustment feedback loops. As the scanner mounted vehicle travels the path it not only moves in the direction of the path it also moves vertically over the bumps. Not much of a concern for lower end scanner units (lacking 9 axis feedback loops) as the vertical changes are detected in multiple directions milliseconds apart however they suffer from sine errors which typically tend to increase the actual vertical rise. Moreover they can't compensate for vehicle suspension travel very well.
Another error point comes from the differences between the methodology used to convert scan point sample averaging to surface model features and waypoints. Generally an ad-hock black magic art.
Its not too much of a concern overall to me. I'm glad S397 put an effort to creating Sebring, I just hope they can learn that not all laser scanning systems/contractors will produce accurate results at the resolutions required. Laser scanning is definitely the best way to accurately model a track when its done right, but just sending a purchase order to a an offshore vendor to produce data and deriving a method to convert that data without actual comparison to actual track data is far from any guarantee of pavement accuracy.
A perception of mine is that S397 doesn't seem to follow typical SW Dev practice by first defining and documenting features and change requirements before a new a release of code and then testing those changes to the requirements it would be consistent that when contracting to a vendor for laser scanning data they may be a bit weak on requirements. That's only my perception, based on the fact that they rarely publish any sort of details in advance of changes, and some changes maybe just bug fixes are not mentioned. And that's every bit heir right to practice in that manner, so its not something I'm faulting them for just pointing it out.
I'll keep buying content to support S397 because at the end of the day its the only true race simulator thats affordable to the general public.
Thanks for the info. Could you elaborate on the issues? Like where you think the elevation is off up/down? Just out of curiosity. I love the track myself and i have no issues with if its a bit off here and there. Its a really well made track and super fun to drive.Ergo, I've driven Sebring a number of times between 1982-2001. I've rode bikes there a few times, last time in 2017. Moreover, I support two IMSA teams with full access to their telemetry files (both MoTec users). One of the IMSA drivers uses rFactor2 as training tool, who also agrees.
As for the laser scanning procedure there are several ways to introduce errors.
Shadowing is somewhat common among concrete block pavement where pavement blocks are uneven with large chips broken on the trailing linear edge of the block when scanning at resolutions of less than 2-3". Sebring has hundreds of those potential error points.
Not all laser scanning systems have the same features such as 9 axis accelerometer internal adjustment feedback loops. As the scanner mounted vehicle travels the path it not only moves in the direction of the path it also moves vertically over the bumps. Not much of a concern for lower end scanner units (lacking 9 axis feedback loops) as the vertical changes are detected in multiple directions milliseconds apart however they suffer from sine errors which typically tend to increase the actual vertical rise. Moreover they can't compensate for vehicle suspension travel very well.
Another error point comes from the differences between the methodology used to convert scan point sample averaging to surface model features and waypoints. Generally an ad-hock black magic art.
Its not too much of a concern overall to me. I'm glad S397 put an effort to creating Sebring, I just hope they can learn that not all laser scanning systems/contractors will produce accurate results at the resolutions required. Laser scanning is definitely the best way to accurately model a track when its done right, but just sending a purchase order to a an offshore vendor to produce data and deriving a method to convert that data without actual comparison to actual track data is far from any guarantee of pavement accuracy.
A perception of mine is that S397 doesn't seem to follow typical SW Dev practice by first defining and documenting features and change requirements before a new a release of code and then testing those changes to the requirements it would be consistent that when contracting to a vendor for laser scanning data they may be a bit weak on requirements. That's only my perception, based on the fact that they rarely publish any sort of details in advance of changes, and some changes maybe just bug fixes are not mentioned. And that's every bit heir right to practice in that manner, so its not something I'm faulting them for just pointing it out.
I'll keep buying content to support S397 because at the end of the day its the only true race simulator thats affordable to the general public.