Assetto Corsa PC Mods General DiscussionPC 

  • Thread starter Paiky
  • 124,705 comments
  • 30,892,132 views
what does the ai.ini look like? Is it all garbled and weird? If yes, then the ACD is also encrypted

Nope
upload_2020-12-27_19-22-1.png
 
Hello! Can someone help me with the templates for this car? whatever I do, all the cars turn white. am i wrong converting .psd to .dds? Thanks a lot! has small adjustments in the handling and I changed the sound!
https://mega.nz/file/VhAXAAZL#jSWh2l49FclOvmyaOSzu-9EOeq2HIN4oXaNRxm30200

I used the templates from @DaBaeda (thanks for doing that) but created a PSD file with additional layers for you. Paint whatever you want on the "Paint Layer" and then save it as a DDS when you are finished.
If you open the car in the CM showroom and save it as a PSD each time you do something it will update in real time to reflect your changes.

Shout me in PM if you want any more help.

Remove the .txt from the attached file

(oops sorry tagged the wrong person)
 

Attachments

  • UNO_BODY.psd.txt
    5.9 MB · Views: 19
Last edited:
Then there’s no reason for the mod to not work with unpacked data... try removing the ACD?

I'm really unsure what's going off I tried the "Use unpacked data if it exists" to see if that does anything and comes up with the fail to decrpt (this is without removing the data.acd) - again to clear out stuff I am trash at figuring stuff out :lol:
 
VR - NeckFX or GEFORCE setting in CM only - I am NOT interested in using RHM (real head movement add-on)

I am trying to configure a setting where I can get body roll and g-force movement in the vehicles in VR. I have seen several videos from members here who race VR, post videos on youtube with body roll in the vehicle. I have asked but they arent sharing their settings. Its not the holy grail but would be nice to get correct like i get in rfactor 2.

If anyone would be so kind as to post their VR settings for NeckFX or View/UI setting G-force i would appreciate it, or any other tips to do so.

Thank you.
 
Last edited:

I'm not sure how I missed @norms optimization, but wow, it's sooo much better. In looking "underneath the hood" of the ai, I noticed the fast lane looked like it was done by coasting around the track. It actually works pretty well as is, but from what I understand it could throw off fuel calculations for the ai (not mention there was no ideal_line.ai file in the data folder). This should improve the speed and consistency of the ai. Tested at 100% strength and aggression with Kunos GT3, and several of the RSS open wheel mods.

AI for Baku 2019:
edit: now version 2 with rain fx added
https://www.racedepartment.com/downloads/ai-for-baku-2019.38220/updates
https://mega.nz/file/b0JHVaya#8hZZvuFttjsEMnMUyMqAria1LLJ2d8E5oj2DYlOCz1o

Baku 2019 v1.1 track link:
https://mega.nz/file/6xhiAAwT#tLPKYU3qRT_YgdZsQFmD6mihFE3hXWCc8__NF2djYSc


AI hint and pit lane for Kyalami 1967:
https://mega.nz/file/Sl41Ca7L#jcqBF3Yl47iuCsrlJElvjZiME4PiT0PssVLzFo_4teg

This is a small tweak- an ai hint for last corner, and smoother entry and exit for pit lane. This was done with the GPL 1967 mod in mind. Keep aggression level at 80% or below for this car/track combo.



AI for Spa 1966:

https://mega.nz/file/C54TXS5a#hN9Qdw_Q8TghgKiDDc2YQ8dABYhEEaDQnnVHSP_o2z0

Speaking of GPL 1967's, a member over at RD asked me to look at Spa 1966 for this car. It seems like a natural combination, but I'm having hard time taming this car on Spa 1966. The look ahead numbers in the car's ai.ini is quite aggressive, and at high speed, they tend to fly off the track. I literally almost had to write an ai hint for every corner. Reducing aggression did not help as much as I would have hoped, surprisingly, but at 80% I mostly was able to get them to stay on track. The only problem is the member thinks they are too slow now, although in my testing the the lap times were equal, if not a hair better with my reworked ai. So if anyone is a fan of this mod car, please test this out with this track, and let me know what you think. Detailed info would help, like problems with specific corners, etc. I've been pulling my hair out on this one... :banghead:
 
Last edited:
is it normal that exporting the FBX from blender is twice as large as the version from RTB while not adding anything but only modifying some mesh?

Depends on what "modifying some mesh" means, a single subdivide will double mesh size instantly.
Suspect the RTB exported file is a customised export that does not include anything absolutely neccesary for kseditor (example no backface as kseditor cannot use them etc) , because the fbx it exports does not load properly for me in Rhino/Maya or 3dsimed, in the case of rhino it actually tells me it is "fixing" it, maya/rhino both export a file roughly double the size as well. if you open with hex editor yuo will also see the textures etc directy references INSIDE the larger size fbx, and the ini will reference a LOT more objects, suspect instancing is not being respected, however all a mute point, no matter what size fbx you have the outcome KN5 is alwasy the same size from any source

below is 3dsimed trying to open the pure rtb export which includes an AC ini file.
upload_2020-12-28_7-3-50.png


and here is after a simple import/export through blender.
upload_2020-12-28_7-4-16.png




the problem: importing the modified FBX with blender on KSeditor gives me the result in the picture:

if you "explode" geometry to move its points you can induce flipped normals, always double check before you export that you fixed normals or use the "make doublesided" function in 3dsimed.
 
Last edited:
Anyone have a RSS Formula 90 - Setup file please? I am getting my ass kicked at Suzuka. I am currently using the default and its got wicket understeer and I am not good at setups.
 
Depends on what "modifying some mesh" means, a single subdivide will double mesh size instantly.
Suspect the RTB exported file is a customised export that does not include anything absolutely neccesary for kseditor (example no backface as kseditor cannot use them etc) , because the fbx it exports does not load properly for me in Rhino/Maya or 3dsimed, in the case of rhino it actually tells me it is "fixing" it, maya/rhino both export a file roughly double the size as well. if you open with hex editor yuo will also see the textures etc directy references INSIDE the larger size fbx, and the ini will reference a LOT more objects, suspect instancing is not being respected, however all a mute point, no matter what size fbx you have the outcome KN5 is alwasy the same size from any source

below is 3dsimed trying to open the pure rtb export which includes an AC ini file.
View attachment 981082

and here is after a simple import/export through blender.
View attachment 981083





if you "explode" geometry to move its points you can induce flipped normals, always double check before you export that you fixed normals or use the "make doublesided" function in 3dsimed.

thank you so much for the reply, have see the face of the object on blender, this is the result:
Screenshot_10.jpg

pratically all the trees have 1 face...

what is the best way to modify a track just born, RTB to 3Dsimed to blender to 3dsimed and later kseditor?
 
thank you so much for the reply, have see the face of the object on blender, this is the result: View attachment 981098
pratically all the trees have 1 face...

what is the best way to modify a track just born, RTB to 3Dsimed to blender to 3dsimed and later kseditor?
3dsimed is the very last tool I use and only if I really have to, it can induce a lot of issues with it import/export and handling of textures, (I actually only use it now for porting tracks from other games), generally I only use to use RTB for sattelite imagery and height data until I found out I can grab that in a better fashion with BlenderGIS anyhow as a bit of poking with the scripts allows you to grab gistorical sattelite imagery from google earth as well, and you dont need to jump through all the hoops of signing up for googleapi etc. (and it supports real ACS data as well, the google height data is very incorrect and compared to DEM files I have found it to be out by as much as 20m in some cases.

When I was using RTB workflow was RTB>>FBX>>ACEDITOR>>KN5>>3dsimed>>FBX>>Blender, because then the required fbx ini files are created properly and you can return to a normal blender>>fbx>>aceditor>>kn5 cycle.
 
Last edited:

That's looking really good. Normally I don't like when tracks have been merged together, but I think this combo will be great.

I may have missed it, but are there going to be additional layouts, and will there be versions where you just stay on one track? I think I'd enjoy the quality of your tracks if I'm on one or the other, and the combined version as well.
 
I never knew Mid Field was always connected to Apricot Hill

It wasn't, I just decided to make both tracks at the same time, then add some connecting roads.

That's looking really good. Normally I don't like when tracks have been merged together, but I think this combo will be great.

I may have missed it, but are there going to be additional layouts, and will there be versions where you just stay on one track? I think I'd enjoy the quality of your tracks if I'm on one or the other, and the combined version as well.

Yes both tracks are able to be driven separate of the other. But there is also the new combined version.
 
I'm also trying to decrypt the LMP1 ACFSK to modify their performance but it's complicated... Maybe @Caddy_DPi could help us, it seems to me that he has already succeeded in decrypting the DPI of Rollovers
 
@TheRealCeeBee has a blender script to turn trees to y, x or any number of faces, search his posts



are you talking about the rtb data or the blenderGIS here?
RTB and if you use the GIS basic google data then yes, especially over tile seams in the data, google height data is only meant to be an indication and is between 30-100m sample points often, and I think trees also screw it up, but BlenderGIS allows you to use real DEM data instead, If you watch the Z numbers in here and reference the overlay of 2m dem versus RTB, about 10m in some place, but it goes wrong up and down. RTB is "ok" for make believe tracks, but if you trying to grab real nuance, then you need to go at DEM "bare earth" 2m or better 0.5m which is available for a few locations in Australia as well. Still not true laser scan but 1m error is heaps better than a 10-30m error.

We all waiting for @LesNeilson to make his DEM plugin for RTB available ;)

 
Last edited:
Due to the seasons adjustments function, the hill in the background turns white, but the trees and grass stays green. Looks very odd (not a big deal to me, just want to mention it). Furthermore, why didn´t you include some realistic billboards instead of the Kunos ones? At least I recommend to use these ones from our fellow PeperoniAC:
https://www.racedepartment.com/downloads/misano-skin-blancpain-gt.31035/

Some people, also on RD, ask me about skins, billboards etc
The track is open for skinning and you can put some of the banners of pepperoni skins in the skin/default folder, it works.
I didn't care too much because in this track there are really few billboards.....

About seasons adjustments, i keep always deactivate, so i honestly did not think about it before your post, maybe i'll look for a solution, thank you. ;)
 
Sharing a mod? Host it on GTPlanet Downloads. Free, public hosting for files up to 10GB in size.
Back