Cars Caterham 7 Super Sprint | 1995 1.35

[JackCY]

This is a great mod. That said, I noticed a few issues with the latest release. Didn't notice if anyone covered these:

1. The left rear "spinning" wheel center on the LHD model does not move with the suspension, it remains fixed relative to the body.
2. Something seems to have happened with the damage/collision box from previous versions. I buried the entire nose cone of my LHD car into the rear of a car in front last night without any actual contact occurring, no ffb bump, no car movement or visual damage.

I will check the wheels. Don't know if Ben will be on today but wanted to do some patching.

As has already been mentioned, I too have encountered some problems with the LHD versions collision box.
On Lake Louise West Forward, it basically does its best impression of a bucking bronco, flipping around, and pointing the front wheels to the sky...
The other versions are fine, though!

Yeah the LHD has a new collision box, even I didn't have the latest model files before release, tested what I could.
Just copy the RHD collider.kn5 for now, that one works for sure.

This happens at Okayama by @Mitja Bonca

Only this car and as far as tested only at this track.

Mentioned before, copy the RHD collider.kn5 for now. I even had Mitja driving prerelease version on his track with the LHD, but you know the new collider was released later so we didn't catch it.

The tires sounds when blocking is a bit low, this one has custom sounds, right?
I find it very understeery but maybe it's like that it'll, tried on okayama and the you need lift the gas for the second to last corner and on the last corner as well, but I do not know how the car behaves irl

I can hear sounds fine.
Yes all the Cats "struggle" with understeer in their stock road as sold setup.
Also check the aero app, you will see why it understeers at speed.
There is a balance between slow speed and high speed understeer in dampers, nothing more can be done without making it a race car. The springs = wheel rate (in AC, there are no springs and spring rate in AC, everything is at wheels) are as close as can be including the progressive part that's done via suspension travel as motion ratio changes, same for bumpstops, 0-100mm bump movement by 10mm and approximated using the only option AC has the progressive rates. It helped with front dive/lift a bit.

Otherwise you just need to learn to drive this type of car, there are not many like it, LFS had similar car and it would struggle with front grip too, you can watch some Caterham driving and drifting instruction videos they note how to move the weight to front as well. The car 51/49% empty, around 48/52 with 75kg driver and around 45-46/55-54 with full fuel tank. Open differential, terrible aero lift, makes it lose front grip a lot. It's not such a big problem with the more powerful versions that have much higher rear slip. But with the Academy, this one and other lower powered Cats it will understeer compared to what you are used to, entering corners too fast, not enough weight transfer to front. You have to look at weight on front and rear axle as you drive, you will see why you're understeering. The springs and dampers are aggressive enough to get as much weight to front as possible (they sell it like this and for track use people change it). And the usual setup and spring changes work, I've left an option in setup menu to stiffen which ever spring you want but not at extreme levels. Sort of middle of the road to track oriented car. This is from real values people use again. As such you can change the handling quite a bit in setup menu.

I can make some rounds around Okayama. In CM my best shows as 1:54.282 on the older version.

Everything default, no setup changes, 100% grip, 12:30 time mid clear what should be RSR Pro settings.

And I don't know the track, it's all about getting heat into tyres and getting used to the softer front, it has plenty steer in if you know how, I'm even turning in a little early, not used to it. Could probably get to 52s, with setup changes probably 51s. WR... I don't know guessed 50s or less. Hard to tell, I've only driven the track so few times and the car mostly on roads not race tracks.

It takes YT forever to spit out the 1440p60 higher quality versions, then even the 1080p60 might be better, it's as if YT runs 2 encoders: old crappy for 1080p60 and lower source and this old first and then new afterward for 1440p60+ sources, once the new is processed it replaces the old, or they keep the old and only add 1440p60 with the new. Nah I think they are all new, it's starting to show up one by one 720p60 first then the rest, even 360p is gone. They literally waste resources and compress uploads twice, one faster older and then slower newer method. And both usually look almost equally ugly for FPS view uploads.

I've tried to match the camera to what they use but there is no fisheye lens in AC as an option so it's impossible and at high FOV it looks bad because no fisheye. Might move it a hair foward, I thought the helmet mount is out of view, it should be.

 

[Leonardo Ratafiá]

thanks for the detailed explanation man!

 

[JackCY]

We can't find the LHD left nor right rear wheel center not spinning emblem. Can you make a video or screenshot if it persists? I can't see it in live gameplay nor replay.

 

[JackCY]

Last sector seems to fluctuate for me.

 

[Ben O'Bro]

Ben O'Bro updated Caterham 7 Super Sprint | 1995 with a new update entry:

Hotfix LHD + NEW FOLDER NAME

• Please delete the old folders! (we updated the folder names to shorter ones to avoid an error)
• LHD flames fixed
• dash cam + f6 camera updated
• LHD shifting updated
• LHD collider fixed (pivot orientation was wrong)
• LHD chrome skin updated (removed the lines effect)

View attachment 214634
​

Voila, hoping we didn't mess up something else, as well as didn't forget anything

Read the rest of this update entry...

 

[JackCY]

 

[Joseph Maxx]

Thanks for the update guys. I came across a little issue with the left rear rim on the LHD version, haven't tested the others yet. Found it while watching replays, strange thing is if I'm watching in slow motion this doesn't happen.

 

[JackCY]

That's odd, I drove around with camera on that wheel and nothing, replay nothing either. Is it only with the blurred wheel? I know Ben mentioned some oddity for blurred wheels overall but it was more texture/quality related or something and for all wheels possibly on any car. Ok I could replicate it now on suspension test track when the wheel moves and jumps a lot at high speed, it indeed shows only in 1x playback and not in slow motion, that's one odd AC behavior. One has to go very fast so the wheel switches to the blurred version. There are kind of 2 rims, normal and blurred, switching automatically. Right side is fine, probably some attachment point for left wheel is missing, wrong named or decoupled.

 

[mclarenf1papa]

[...] that's one odd AC behavior. One has to go very fast so the wheel switches to the blurred version. There are kind of 2 rims, normal and blurred, switching automatically. Right side is fine, probably some attachment point for left wheel is missing, wrong named or decoupled.

It's a configurable switch though. They could swap at 1 kph if you wanted them to (or have as many switches as desired IIRC).

 

[Ben O'Bro]

Thanks for the update guys. I came across a little issue with the left rear rim on the LHD version, haven't tested the others yet. Found it while watching replays, strange thing is if I'm watching in slow motion this doesn't happen.
View attachment 214677

thanks for that, i have no idea what's happening there!
Something with wrong hierarchy maybe, i'll check tonight

 

[garyjpaterson]

Difficult to tell with a screenshot but it could be that its not centred in the dummy or the dummy pivot it off - or it could be parented to the wrong dummy (RR instead of LR for example), done that before.

 

[Ben O'Bro]

Difficult to tell with a screenshot but it could be that its not centred in the dummy or the dummy pivot it off - or it could be parented to the wrong dummy (RR instead of LR for example), done that before.

 

[chili pepper]

Thanks for the update guys. I came across a little issue with the left rear rim on the LHD version, haven't tested the others yet. Found it while watching replays, strange thing is if I'm watching in slow motion this doesn't happen.
View attachment 214677

That's the issue I was trying to describe. It almost seems like the spinning wheel center got "unlinked" from the rest of the left rear suspension and just stays put relative to the body. In a right turn, the wheel center punches through the lower portion of the wheel rim, left turn and it punches through the upper, following the body roll.

Still a great mod and probably one of my favorite vehicles of all time IRL. Aside from some time spent in a new virtual Ferrari or two, this is literally the only thing I've been driving lately and is what weened me off my Lola T70 addiction for a bit. Kudos to Bo'B et al!

 

[Ben O'Bro]

for now please find the kn5 to replace for the LHD :
http://benobro.org/sub/bo_caterham_s7_1700_ss_lhd.kn5.7z - 13mo
i wanna wait few more days before patching it again... just in case we spot another mistake! I'll also take that opportunity to improve few other details

and a new skin

 

[Joseph Maxx]

for now please find the kn5 to replace for the LHD :
http://benobro.org/sub/bo_caterham_s7_1700_ss_lhd.kn5.7z - 13mo
i wanna wait few more days before patching it again... just in case we spot another mistake! I'll also take that opportunity to improve few other details
View attachment 214824 and a new skin

Thanks Ben that was really quick.

Just one more thing I've noticed but not sure if it's something on my end. The dash seems to blow out when lit by direct sun, probably to do with your specular and diffuse settings but it almost goes white when the sun is behind the driver. Not sure if it would be a quick fix but maybe you could have a look at it. BTW I'm using the default PP.

Thanks again.

 

[JackCY]

I didn't notice dash blow out it gets bright but that's not surprising it's flat and when sun is behind it's kind of lit to it's max, it's like staring into mirror when sun is behind you. There is this surface reflection for the details that blinks on C1 night but that's quite common for all cars to do because of the moded weather and all.

The default PP, is quite contrasty and will blow out a lot, personally I can't use it, made my own presets long ago and still I gotta switch between target brightness depending on weather, yeah weather can really mess up the brightness especially now when they removed the profile area to use and it's forced to fullscreen all the time, can't be tuned to react only to what you see and not the dash etc.

Default PP is set to -1 for tonemapping which depends on what Kunos inializes the Yebis with, so far it has been 2 (PFXTM_SENSITOMETRIC), personally I can recommend using 5 (PFXTM_LOG) and then play with gamma, mapping facor = how much compression do you want and the automatic exposure. There is a whole help file for the Yebis in .chm but you have to download it from Yebis website.

Patch wise we will wait if there is something else.

 

[Dave K]

Thank you so much for correcting the long name issue .
It drives great can see myself driving this for hours!

 

[JackCY]

Thank you so much for correcting the long name issue .
It drives great can see myself driving this for hours!

Longest should be the Clamshell with 32. All I can say is that Ben discovered capital letters don't work somewhere, I don't know where, ask Ben. And the name has to be 33 characters max? I've read some of the app APIs but don't remember seeing name lengths, thanks for giving us the heads up. 33 char max and small letters only. Are you sure it's 33 max usable or 32+0? As in C style uses \0 at the end to end the string. But if it's just a structure with predefined 33 length, one reads it and is expected to \0 terminate it when using C or does it come already terminated? Or it has it's own name length number in the structure that can be up to 33? If it's pure C API then it's probably 32 usable + 0 termination = 33. The 32 clamshell we have should be fine.
Would bet it's 32 usable chars max, small letters and alphanumeric + underscore + dash only. Anything else = possible issues.

What app are you using? It was failing to detect the car or it was displaying empty name before? If done right then it should cut the name if it's too long to say 32 chars by itself, both when writing it and reading it. Passing data via shared memory can be a real pain (and almost impossible to debug at times) especially between programs done in different programming languages.

 

[Dave K]

I am using the shared memory static carModel[33] Name of the player’s car
http://www.assettocorsa.net/forum/index.php?threads/shared-memory-reference-25-05-2017.3352/
Converted to c#
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 33)]
public string CarModel;

wchar_t carModel[33] Name of the player’s car
wchar_t track[33] Name of the track
wchar_t playerName[33] Name of the player
wchar_t playerSurname[33] Surname of the player
wchar_t playerNick[33] Nickname of the player

You are correct it is the C API

 

[JackCY]

And when it's larger than 33 AC writes nothing to the name? It should, by common logic write only 32 wchars and then a null terminator. The can of worms here hiding is that wchars are variable size that depends on compiler and overall handling strings that aren't just ASCII is not trivial especially in a platform independent way. Plus what locale and encoding etc. are they. What if the name is in Chinese or Russian and what kind of Unicode.

The C# reading C written data can be almost undebuggable and that's with access to code of both. There can be some odd size differences between types or something and causing random issues, random = the worst you can get and near impossible to debug.

The wchar_t can be 8, 16, 32bit probably. I'm guessing here, Kunos is using C++ and on Windows wchar_t in MSVCC would be 16bit = 2 byte.

https://msdn.microsoft.com/en-us/library/s3f49ktz.aspx

Might be better if char16_t, char32_t were used, these are newer since C++11. wchar_t never worked the way it was supposed to.
I'm guessing again, they take C++ datatypes and convert them to C style. 32 wchars per name is a poor choice, way too short.

"1995 Caterham Super Seven 1700 Super Sprint (model for Europe)" is the full name used in catalogs. Just 62 chars. And I've probably met people with full names longer than 32 chars too. Such memory saving for 1 instance makes no sense, it would make sense to optimize this in a different way (variable length) if one had millions or billions of instances, then sure the memory savings start to matter.