If anybody is borred on christmas 25 evening we will host some pachanga races at www.cesav.es 22h15 CET Come and join the fun!

Last edited by loopingz, Wed 20 Dec 2023, 11:27.

Thanks for the good work! A lot of new and polished feature that's cool! And even though I have been using test patches a lot, I am glad to have now everyone the same official version!
I did a couple of races yesterday, no issue as expected.
I wish you the best on the following development and hope it will be the end for you to maintain two branches!
Merry christmas!

Looks promising!

Ok my bad, I did more testing and I was wrong, getting super low on the force feedback in LFS can kill the clipping. So yeah, nothing else needed!

I am not sure you want to go in this rabbit hole:

But great work as always!

Quote from rane_nbg :

You're the one who fails to understand how ffb in lfs works. Lfs can't send certain torque in units of Nm to a wheel, because it has no idea what a wheel can reproduce. It only sends a value between -10000 to +10000 and that's it. What wheel does with that is up to each wheel and its firmware.

In this situation you mentioned, you just have to lower lfs ffb below 10% to a desired value where there is little to no clipping.

What you probably are asking is for Scawen to re-scale the ffb strenght such that optimal value for most cars is about 100% and not 10%. For me personaly I don't care at this point, it's just a relative number.

Well explain me where I fail then as I do understand that LFS and force feedback work with a limited amount of possible values. So ffb is integer between -10000 and 10000 yes cool.
On one side I do like to feel some differences between a truck, a high downforce powerful car or an old school 700kg low power stuff, small wheels.
Let's say the laterwill range around -2000 to 2000 no problem.
The truck -8000 to 8000 no problem.
The formula -25000 to 25000 will be clipped to -10000 to 10000 whatever I multiply that gain to my wheel.

When the engine sound clipped too much a quick correction was made.

Imagine you like music like acoustic stuff to make the baby sleep. OK no problem. That sound will be recorded and normalised. It doesn't matter if the original recording was made at 70db or 90db, you will be able to play it on your sound system adjusting the output to your liking.
Now you want to record a live Slayer show, well you will choose different settings because it will blast around 105db at the mixing table on a day they refrain from turning up too much. But if you have the same settings you will have a huge clipping mess. The idea is to be able to put some volume slider before the sound is clipped. That the basic of numerical signal treatment. That the same for force feedback.

Currently we have
A- Physic calculation
B- FFB -10000 to 10000
C- FFB settings in game 0 to 100%
D -FFB settings drivers side

If B is clipped there will be no miracle in C and D even if C is car specific it is too late to do something. Pros: perfect physical difference of non clipping cars independant of C and D.

What we could have without changing the physic, just the ffb rendering:
Currently we have
A- Physic calculation
A'- Normalization factor for high FFB cars
B- FFB -10000 to 10000
C- FFB settings in game 0 to 100%
D -FFB settings drivers side

A' could be either set in stone by developer or modder manually or through random test calculation (I mean putting AI on some tracks and measure force output for a variety of cases, crashes values being not relevant or interesting to me)
It could also be a user slider. Like FFB Norm slider it could even replace C
Anyway if you don't care having some multipliers to 1 would not bother you.

Please test AS5R with FORMULA RC6 at various force settings and tell me how much clipping do you have. I think you did not test.

This awesome car will be featured on Monday 11th in Cesav race at 22.15 cet.

Can be moved to improvement suggestion...

Rane, Johney, if you don't understand what clipping is don't put me as crazy.
Let's say high down force can put 30Nm on the ffb. Then If LFS clips that at 10Nm. And I put 10% force in LFS I will have have the following:
Real force 1 2 3 4 5 10 20 30
Wanted force .1 .2 .3 .4 .5 1 2 3
LFS calculated 1 2 3 4 5 10 10c 10c
Force on my wheel .1 .2 .3 .4 .5 1 1 1
Force I would love to have .1 .2 .3 .4 .5 1 2 3

About the force feedback.
I did some laps at AS5R with FORMULA RC6 https://www.lfs.net/files/vehmods/850339
The force feedback is saturated most of the time. My driver (for the wheel) is at like 20% force my lfs force value is at around 15%. Max force would be dangerous.
What I mean is that I would rather not clip LFS input ever except for crashes or glitches. We would need either some extra factor or percentage for the max FFB to a car. Either set by the moder or manually by the player. I would really prefer to use higher force settings on my side and have the game not clipping all the time.
All the cars with big tires high aero and horsepower are concerned.
Clipped input just give a direction input, no more info.

Quote from ElPvtoV2 :

Hi, i bought a new pc and i had installed this patch, but now my lfs dont work, i cant see the events in the main menu, i cant see the multiplayer servers and when i close lfs cause it didnt work, i have a permanent black screen and i have to restart the pc. (It happends when i install patchs, how can i fix it?)

Try first to install and run 0.7D.
It should work.
Then patch it.
It should work too.

Report what happens.

Thanks for the good work.

I don't want to be pulling a request but on the incompatible branch it would be great to have some basic handling of the electric. Like if you are pulling 200W with all the lights on then the alternator should charge them back from the alternator let's say 400W with 0.5 efficiency.
A
In this case it would be 0.5hp wasted which will not be very noticeable but this way nobody would run lights on for no reason adding to realism.

Other than no bugs noted!

Well most bike are generating lift (like most cars) and it extends to racing motorbikes.
This paper give some real values for racing motorbikes. The little wings on new motoGP are reducing a bit the lift and cleaning the flow for drag removal... 32 pages but pretty interesting!
https://upcommons.upc.edu/bitstream/2117/393241/1/energies-16-04793.pdf
It is still a pretty complex matter with ground effect appearing on leaning bikes and Cl and Cd varying depending on angle of lean and wheel speed...

Quote from Scawen :

Thanks, although that didn't affect the AI, it's an interesting issue. I've fixed that today by using the drive speed (as used on Speedometer) instead of the true speed. It's more realistic and solves the problem.

Was it AI or speed limiter issue? Because it happens to me on powerful stuff but I guess that would be for incompatible update.

At Cesav.es we did an event with bike some month ago (I would say 12-2022). So it was old physic, it was not perfect but we all had great fun. Out of comfort zone learning new stuff having some laugh... Can't wait to seriously try these improvements!

Great job!

Excellent !

Yes but that is not my point or request. I can go to LFS world and see all of my pbs...
But no I want my same button /w pb to adapt to the situation. I am spectating there is only one car type I want my pb with this car. Or if there is various cars I want my pb with the car I follow.
I usually play in VR and if it is not on a button it doesn't exist to me...

Also mod car name are funny to remember. Peace of cake for rain man
I would say!

If I am spectating I cannot know my pb by /w pb... It could autofill that for most cases...

I would like to be able to do /pb when spectating and having as a result my pb of current combo or my pb corresponding to the vehicle I am spectating.

I did a try with a full AI random grid of 32 at ky2r...
Bike are interesting but some cars just like to kill them.
One formula got stuck in the sand did not reset. Bursted tires and ended fuel pretty stupidly. One other car got stuck did not reset did not ended fuel.
I have to admit I have less fun watching them but there was some huge improvement done!

Quote from matze54564 :

I play Single Player with AIs. I drive the BF1 with 4 AIs. The AIs sometimes crashed before the update and they also crash after the update. I have smooth graphics in single player with a FreeSync Monitor. This doesn't work in Multiplayer. The players in multiplayer have very high skills. I guess to get these skills, you need training in the single player mode. I need training in the single player mode.

You have every right to play single player!
You could find some multiplayer that goes fine with not very fast player and make good improvements but you don't have to and that's cool.

Quote from Scawen :

Thanks for the video.

I think they initially run wide as a result of pulling out to pass another bike, which causes problems if they end up on the grass. I'd be interested to know if any AI take a corner too wide when they are not behind another vehicle.

The inability to reset is because LFS thinks they might spawn in the way of another vehicle, but it looks as if that code needs to be adjusted. For example it's not really an issue if the other vehicle isn't moving.

The AI reset is not working very well to say the least. I admit from time to time I set up a offline race full of AI and I just watch for the mess. It relaxes me, I know it is strange. Do we need better or not? Not sure but they are a lot of situation not handled properly, like reset wife or declare race ended after a while. No fuel auto spectate after some time should be auto DNF. Stuck for more than a minute auto reset or DNF...

I understand the regard d'anchois idea and the obvious it looks like an anchois! But you have to be french speaker to come with that! It made my day anyway!

Detect & Monitor from @kingofice does also have this feature. But it also shows how Lazy was great at that. Lazy was able to detect layout or even multiple layouts.
I do understand that it would time consuming to natively add.
Currently the issue is some people me included (depending on the race) might stick to an older version to have lazy or detect working.And that's bad for test patch testing and feedback.