The online racing simulator
If the clutch manual is an option the game because they do not use it? It's easy to say that with this clutch is going faster, because you play well, and instead of Tache cheats, get to practice and beat their WR.
Just had to test the hotlapping thing. Apparently all the very top times are done without H-shifter and manual clutch.. so i feel its more better to compare to people who use manual clutch and h-shifter. If i do it like that, im 3rd fastest in BL1 with XRG.. and i feel it would be close to impossible to be on page1 driving "correct". Im just thinking.. theres allready "aids" that help.. why the hell people need macros for? Its like doping that havent yet been banned. Its odd that you get satisfaction using some advantage most dont use. No matter though, its great that you see the input method in the HL chart so you can compare yourself rightfully with the like minded.
At the moment I am a Joypad user and use manual clutch and manual gears on a seperate buttons. Can someone correct me but during this thread I got the impression even using button clutch was seen as bad. I do realise the advantages of the button clutch though as in I would imagine I would be able to change gears ALOT faster than someone using a H shifter and pedals .
I actually talked about this subject of a macro to change gears and clutch before this thread but never thought someone would actually make one or even use it to their advantage .

To me using the macro shows laziness, and disrespect to the drivers who are doing it properly. I have been lapping on blackwood a few years but hit a brickwall at 1:33.45 at latest patch and see people doing much much lower times. Personally I put it down to my own driving but if people use this to get advantage over me then i would probably be a bit disapointed and upset? But am I doing the same thing by using manual clutch on a button to wheel users? if so I am sorry

Ps faster clutch DOES can increase stabilty and even maximum speed if you get the corners correct. I guess this macro helps with consistancey.
Quote from Glenn67 :And if your are very practised at starts and are tuned in (i.e. not in a daze ) you can actually be suprisingly consistent. My best runs of a serries of 5 will always be within 0.00 - 0.03 range of my possible best.

spr it or it didn't happen...

Quote from Glenn67 :
While it may only be 3 gear changes that is the amount on average between sectors on BL1 for example where I was under the impression myself that 0.1
per sector difference was possible because of a macro. But now I doubt it.

There is no comparison between the speed gain that benefits UF1 and eg FXO or RB4... also there is no comparison between the speeds that you have to change gear on a track and on drag strip.

The faster the car goes the more resistance is caused by aerodynamics and rolling resistance. Having a fast shift in high speed is way more beneficial than having a fast shift at 60km/h

Quote from Glenn67 :
And as for 0.1 - 0.2 difference in lap time, imo it is quite achievable at the top level. It happens all the time that is why records don't stay static for long periods of time on say BL1 which is hotlaped to death. Most people get stuck in a certain range because their brains are telling them there is no more time to be found on this combo. Now if you set a macro for gear changes and are of the belief that it is faster then I do definately believe that it is possible for that individual to go out and get a new pb or lap record. That is a phenomenon that is well documented in many rl highest level sports.

It is not achievable that easily by the same person after a really small amount of laps...
But anyway this depends on how close to the actual combo's lap time limit, is the current wr.
After the resent hotlap chart reset (patch Y) it's normal that the WR's are not so hard to beat.

Quote from Glenn67 :
So at this stage I remain unconviced it actually is any faster.

It is faster anyway... it is illogical to think that been able to reduce the time that the car stays with no power by shifting faster, has no speed gain.





Well even if I believe that a short drag strip is not the proper environment to measure accurately the speed gain cause of the use of uber fast clutch, I did some testing with the UFR using the default hard track set.
I have attached the replays for so you can judge them.

I did 5runs with auto clutch, then 5 runs with manual clutch (button rate 10) operated by my hand… because I am not at the place where I have my G25 to test with the profiler... so I am using the mouse.
Then again 5runs with auto clutch and another set of 5 runs with manual clutch.

Using my hand instead of an external script gives me much less consistency and of course still the script must be faster than my finger taping the clutch key a split of a tenth sec before changing gear.

Not counting the first 10 runs when I am still learning the rhythm of the start lights.

The average time at the 5 final runs with autoclutch is 12.76 and with button clutch is 12.65… more than one tenth difference at 400meters… (But I know that this is not so accurate)

The average time between the split and the finish is 4.342sec with autoclutch and 4.324sec with button clutch… at that part you shift only one gear… This is a more accurate measure because it depends only on the car's acceleration and it is not affected by my ability to start consistently when the lights go green.
Having a 0.2 tenth of a sec advantage at 200meters is more than enough. (For the UFR... other cars will show different speed gain)

The speed gain is more than obvious just by using the button clutch by hand… imagine having a script do it for you.
Attached files
UFR clutch drag test.rar - 24.5 KB - 228 views
Quote from kaynd :spr it or it didn't happen...

Just so you don't think I'm a total idiot

Autoclutch is sequetial style shifting
Manualclutch is clutch on axis (g25) and padel shifters
Macroclutch is c and s assigned to wheel buttons then creating a macro that is assigned to a paddle with buttonrate 10

Interesting thing here is manual clutch is fastest, followed by macro then closely by the autoclutch. Another interesting thing is I find the macro to be less reliable than the autoclutch, the macro seems more prone to miss-shifts for some reason.

And I'm not trying to be contentious by doing this, I'm just trying to approach this in an anylitical manner. The disscusion about button clutch has been around for a long time. And I'm pretty sure Scawen has taken steps to even it out. So the advantages guys are seeing from the button macro might be related to something else - perhaps stablitiy in braking or just merely consistency as has been mentioned.

The more investigation and discussion that is conducted on this though the more likely we might see something done about it.
Attached files
UFRTest.zip - 26.1 KB - 234 views
Quote from Glenn67 :
And I'm not trying to be contentious by doing this, I'm just trying to approach this in an anylitical manner. (...) So the advantages guys are seeing from the button macro might be related to something else

I am analyzing on the above post why the script is actualy faster even in straight line and you just pass that saying that the advantages must be related to something else...
This is not analytical manner, this is just your opinion because you are not able to separate the clear speed advantage from the driver stability.

[edit]

I did 20runs one after the other, including mistakes and took account only the last 10 where I was more consistent (5AC & 5 BR10) for the average times…
Also I measured the average time between the split and the finish which as I explained it is not affected by my inconsistent starts.
You can see when exactly each run happened by the “last modified” date of the file.

19/7/08
3:16 ->12.89 AC
3:17 ->12.99 AC
3:17 ->12.77 AC
3:17 ->12.79 AC
3:18 ->12.81 AC
3:19 ->12.85 BR10
3:20 ->12.67 BR10
3:20 ->12.85 BR10
3:21 ->12.74 BR10 (with miss shift)
3:21 ->12.64 BR10
3:23 ->12.81 AC
3:23 ->12.88 AC
3:24 ->12.81 AC
3:24 ->12.64 AC
3:25 ->12.69 AC
3:27 ->12.54 BR10
3:27 ->12.62 BR10 (with slight miss shift)
3:28 ->12.75 BR10
3:28 ->12.65 BR10
3:30 ->12.72 BR10


So in order to back up your opinion, you bring some your hand picked results which can be easily guessed from the time difference between the runs.

19/7/2008 11:45 -> autoclutch 12.67
___>>__ 11:56 -> autoclutch 12.66
20/7/2008 12:03 -> Manualclutch 12.61
___>>__ 12:06 -> Manualclutch 12.59
___>>__ 12:19 -> macroclutch 12.64
___>>__ 12:25 -> macroclutch 12.65

In that way I could just use my 12:54 done with manual button clutch and my 12:64 done with autoclutch… and not loose my time trying to explain what I explained at my previews post.
can someone explain what is this macro-script thing?.. I'm a bit confused in here
Quote from Tomba(FIN) :can someone explain what is this macro-script thing?.. I'm a bit confused in here

Short version or long version? Depending on the author any of them might help make you more than a bit confused.
Quote from Tomba(FIN) :can someone explain what is this macro-script thing?.. I'm a bit confused in here

It's just people using button clutch (enables faster shifting) but they don't have to press any button for the clutch because the "macro" means that when they press their paddle to shift up it also engages the clutch. So they do everything the same as us normal auto clutching people, but they get far quicker shifts.

Quote from kaynd :This is not analytical manner, this is just your opinion because you are not able to separate the clear speed advantage from the driver stability

Fair call

In my defence I was very tired and about to go to bed when I saw your post

Ok for some analytical stuff.

Firstly lets establish some terms we can all use so we understand each other.

In each action of the clutch we will break down the action into three components:
  1. Climb - the time it takes to fully engage the clutch.
  2. Sustain - the time the clutch remains fully engaged.
  3. Decay - the time it takes for the clutch to disengage and to engage the next gear.
Firstly I have only analysed the fastest runs one from each category for you and me. I intend later to analyse the slower runs to double check the results - but don't have time right at this moment.

Method of analysis:

view replay in 0.125 step slow motion and record on note ppaer action of clutch for each run for gear changes 2-3 and 3-4.

Analysis 1: your 12:54 run with macro clutch

2-3 gear change

Climb 12:80 - 12:85 = 0.05s
Sustain 12:85 - 12:90 = 0.05
Decay 12:90 - 13:00 = 0.1

Total time to change gear = 0.2s

3-4 gear change

Climb 16:63 - 16:68 = 0.05s
Sustain 16:68 - 16:72 = 0.04
Decay 16:72 - 16:82 = 0.1

Total time to change gear = 0.19s

Analysis 2: my 12:64 run with macro clutch

2-3 gear change

Climb 12:68 - 12:73 = 0.05s
Sustain 12:73 - 12:77 = 0.05
Decay 12:77 - 12:87 = 0.1

Total time to change gear = 0.2s

3-4 gear change

Climb 17:22 - 17:27 = 0.05s
Sustain 17:27 - 17:31 = 0.04
Decay 17:31 - 17:41 = 0.1

Total time to change gear = 0.19s

So both our macro runs match.

Analysis 3: your 12:66 run with auto clutch

2-3 gear change

Climb 12:90 - 12:90 = 0.00s
Sustain 12:90 - 12:99 = 0.09
Decay 12:99 - 13:10 = 0.11

Total time to change gear = 0.2s

3-4 gear change

Climb 17:55 - 17:55 = 0.00s
Sustain 17:55 - 17:65 = 0.1
Decay 17:65 - 17:75 = 0.1

Total time to change gear = 0.2s

Analysis 4: my 12:67 run with autoclutch

2-3 gear change

Climb 12:85 - 12:85 = 0.00s
Sustain 12:85 - 12:94 = 0.09
Decay 12:94 - 13:05 = 0.11

Total time to change gear = 0.2s

3-4 gear change

Climb 17:11 - 17:11 = 0.00s
Sustain 17:11 - 17:20 = 0.09
Decay 17:20 - 17:31 = 0.11

Total time to change gear = 0.2s

Again I rusults concure.

Finally a analysis of manual clutch (g25)

Analysis 5: my 12:59 run with manual clutch

2-3 gear change

Climb 12:71 - 12:76 = 0.05s
Sustain 12:76 - 12:81 = 0.05
Decay 12:81 - 12:87 = 0.06

Total time to change gear = 0.16s

3-4 gear change

Climb 17:02 - 17:07 = 0.05s
Sustain 17:07 - 17:13 = 0.06
Decay 17:13 - 17:18 = 0.05

Total time to change gear = 0.16s

I don't have time to really sit down and analyse these results just at the moment but my initial look at them and I felt quite suprised.

The auto clutch seems to be the one that is least realistic in its behaviour and that could be the reason it is at a disadvantage (auto clutch goes instant on!) this would explain why the car feels more balanced in transitions with a macro/button clutch as well.

So a quick summary:
  • Fully manual clutch with G25 is definately faster!
  • The time it takes for the clutch to act in both button clutch and auto clutch is pretty much the same.
  • The way the auto clutch behaves is not realistic in the climb phase and so therefore looses out to other methods as it stays fully engaged for a longer period.
i had the wr on fe green rb4 before fin eza beat it, well done
i dont use a script, just every time i want to change gears, i press the right hand button on my g25 which engages the clutch. i cant get the script to work.
a cheat becomes a cheat, only if someone knows you are using it
Quote from Fischfix :a cheat becomes a cheat, only if someone knows you are using it

Regardless of the fact that its not really detectable if one is or is not using a recorded macro to simulate the super fast clutch in/change gear/clutch out, the question becomes is it against the rules? only if it is against the rules of a league or server then someone will have to find a way of detecting it.

Of course I dont think anyone here would dissagree that it is against the spirit of good sportsmanship.

SD.
Glenn67 you must be doing something wrong, how did you script your macro? In my profiler I configured the macro to be as fast as 50ms, autoclutch must be around 100 ms, that difference is huge if you count all gear changes in a whole lap.
Quote from arrowkart4 :i had the wr on fe green rb4 before fin eza beat it, well done
i dont use a script, just every time i want to change gears, i press the right hand button on my g25 which engages the clutch. i cant get the script to work.

Well, you must have incredible fast reaction then, because from the attached graph you can see the difference in speed. Yours in red.
Attached images
sshot-4.png
Quote from Glenn67 :Fair call

In my defence I was very tired and about to go to bed when I saw your post

Ok for some analytical stuff.

Firstly lets establish some terms we can all use so we understand each other.

In each action of the clutch we will break down the action into three components:
  1. Climb - the time it takes to fully engage the clutch.
  2. Sustain - the time the clutch remains fully engaged.
  3. Decay - the time it takes for the clutch to disengage and to engage the next gear.
Firstly I have only analysed the fastest runs one from each category for you and me. I intend later to analyse the slower runs to double check the results - but don't have time right at this moment.

[...]

Yeah now you are saying something but I am not yet beaten, was I that wrong? Is really there no difference between autoclutch and button clutch?
Yes indeed the time the whole process takes, is 0.2sec in both autoclutch and buttonclutch (rate10)
But there is a little something that makes a difference… in autoclutch mode the clutch disengage happens instantly, leaving the driven wheels with no power at all for 0.1sec, till it begins to engage the engine with the transmition..
Button clutch takes 0.05 to fully disengage the clutch, leaving the driven wheels with no power at all for ~0.05sec.
In both cases clutch engage takes 0.1sec

And a short uber jerky slow motion vid showing the case http://www.box.net/shared/a4mhdcpc8s

(I use different words so a quick explain)
  1. Climb - disengage
  2. Sustain - no power on the driven wheels
  3. Decay - engage
Also because in button clutch mode the engine is less time disengaged from the drivetrain, it has less time to freely rev up (assuming you are flat shifting) before it gets engaged again… so you get less clutch slip and heat.
This is something really important on long races, which bawbag has already mentioned.
Further to the above analysis as promised:

I tested a slow run of each autoclutch and button clutch that kaynd posted to double check my results.

The slow autoclutch run - climb, sustain and decay pattern is identical in slow and fast passes on the drag strip. So can deduct that the time differences of each run are due to reaction times and/or spinning tyres.

The slow button clutch run - had consistent climb and decay sections but had longer sustain section. So when pressing the two buttons together there appears there can be some variation in how long it stays at the 100% engaged level. Which adds a new variable into the drag run times.

I know kaynd did this with buttons so I'll set my wheel up and do a few trial runs with a macro to see if the same holds true. I suspect it might as when I did the runs last night with macro it seems to "stick" sometimes.


Ok based on the data I posted above I've made some graphs to illustrate the difference between each type of shifting method so people can visualise it better.

The first graph ManualClutch.GIF is how the fully manual clutch works (i.e. autoclutch = no, clutch = axis, clutch pressed by my foot ) The three different phases are evident and on a quick gear change a shift of 0.15s overal is achievable. You can see that the climb, sustain and decay phases all take 0.05s respectively.

The second graph MacroClutch.GIF is how the clutch works using a macro to press s and c simultaniously. Again the three phases are clearly evident, but you can see in particular that the decay phase takes twice as long as what can be accomplished with a clutch pedal. The climb and sustain phases seem to be very consistent at 0.05s and 0.10s respectively, the sustain phase is the interesting one in that it seems it can vary somewhat - the quickest being 0.05s blowing out to 0.08s in some cases. (will check this further)

The third graph AutoClutch.GIF shows how the autoclutch feature behaves currently. The thing that is so startling about the autoclutch is that the first phase the climb phase is totally absent. It just goes from 0% clutch disengagement to 100% disengagement instantly! So this I believe is a bug. The sustain phase seems to be too long due to the absence of the climb phase. The decay phase is the same as for button clutch.

Some thoughts on the observations:

I think it is a good starting point to use the clutch assigened to an axis and autoclutch turned off as the reference for how the clutch should behave. ManualClutch.GIF

Having the clutch assigned to an axis and using fully manual clutch in LFS is definately the fastest method, which is extremely encouraging because it shows that there are rewards there for those that persue realism.

We can see that the way the button/macro clutch works is infact very similar to the way the clutch pedal works in a fully manual clutch setup. The one main difference you can see is that the decay phase is twice as long making the shifts on average 0.20s approximately 30% slower than a fast fully manual clutch shift.

The most suprising result from my analysis though is that of the autoclutch, which doesn't appear to opperate at all as expected. The climb phase is non-existent which does explain why people feel that the car is more stable under braking with macro as the wieght shift with the autoclutch will be sudden and jerky where with any other method of shifting the wieght shift will be more smooth. Also because the climb phase is missing the sustain phase (time the clutch is drpessed 100%) is much longer than expected thus giving a possible 0.02-0.03s time andvantage (taking into consideration partial engagement with other methods) for each shift when flatshifting to other methods of clutching.

At this point I am confident enough to say that I believe there is a bug in the autoclutch system. How ironic is this? Out of an argument that macros are cheats we may have dicovered the macro is not as bad as it seems and the autoclutch system that most people use has a fault Anyway time to raise a bug report and see where it goes me thinks.
Attached images
ManualClutch.GIF
MacroClutch.GIF
AutoClutch.GIF
Quote from RiGun :Glenn67 you must be doing something wrong, how did you script your macro? In my profiler I configured the macro to be as fast as 50ms, autoclutch must be around 100 ms, that difference is huge if you count all gear changes in a whole lap.

Post an spr so I can compare it to kaynd's and mine please

I've just looked at slow motion replay of Worms AS3 RB4 macro example replay and the results are in line with what we are seeing in the other tests as to the pattern of expected behaviour.

Two Acceleration changes itemised:

1st acceleration shift example
1:13;32 - 1:13;37 climb phase 0.05s
1:13;37 - 1:13;40 sustain phase 0.03s
1:13;40 - 1:13;51 decay phase 0.11s
Total shift time 0.19s

2nd acceleration shift example
1:57:09 - 1:57:14 climb phase 0.05s
1:57:14 - 1:57:16 sustain phase 0.02s
1:57:16 - 1:57:26 decay phase 0.10s
Total shift time 0.17s

1st shift example under braking
1:41:09 - 1:41:14 climb phase 0.05s
1:41:14 - 1:41:22 sustain phase 0.08s
1:41:22 - 1:41:32 decay phase 0.10s
Total shift time 0.23s

2st shift example under braking
1:41:99 - 1:42:04 climb phase 0.05s
1:42:04 - 1:42:10 sustain phase 0.06s
1:42:10 - 1:42:20 decay phase 0.10s
Total shift time 0.21s

Here's the results of analysis of worms AS2 RB4 wr using AC.

Example of shifting under acceleration.
0:47:05 - 0:47:05 climb phase is instant on 0.00s
0:47:05 - 0:47:15 sustain phase 0.10s (longer than you would expect almost like it's compensating for the lack of climb phase)
0:47:15 - 0:47:25 decay phase 0.10s
Total shift time 0.20s

Example of shifting under braking.
1:08:31 - 1:08:31 climb phase 0.00s
1:08:31 - 1:08:40 sustain phase 0.09s
1:08:40 - 1:08:51 decay phase 0.11s
Total shift time 0.20s

From this we can clearly see that the total shift time is pretty close in most cases although it does appear you can get faster shifts with the macro on some occassions approaching similar performance to an actual clutch pedal. We can also see that the autoclutch does not behave as one would expect in that it engages 100% instantly and seems to compensate by staying engaged longer before releasing. This would explain the speed advantages that people are seeing along with how the car feels in transients. I actually think the macro clutch is simulating how the autoclutch should be behaving, will be interested in Scawens views when he sees this.
And to add to that Glenn, the manual/macro/button clutch don't have to be 100% engaged to acquire a shift. For example, while auto clutch is at 100%, the macro may only be at 60% but the gear has still engaged.
Blah Blah Blah
My vote is for the exclusion of the clutch button in the next patch!
Quote from [DUcK] :And to add to that Glenn, the manual/macro/button clutch don't have to be 100% engaged to acquire a shift. For example, while auto clutch is at 100%, the macro may only be at 60% but the gear has still engaged.
Blah Blah Blah

You obviously didn't read my posts, but I wouldn't expect a Duck to anyway! So that's ok

You are right about the engagement as well, but something is abit odd there the pic acro posted of a "macro" user with analysis clearly shows only 60% engagement but the gear changes I looked at in Worms replay had 100% engagement. Not sure what going on there, I have some ideas though but would be near on impossible to confirm without alot of work
Yesterday I did my own test too, I was in CTRA when Need For Weed first humored me as his No1 enemy and then asked me for my setup for the XFG@SO2R. I already knew that his wr with XFG@SO2 was most probably done with the macro, since neither AC nor CL flag is on in the charts. So I gave him the setup (which btw was not made by me, its partyboy's) and when he told me he intends to brake my (not so good) wr in that combo and disconnected I left him a message in LFSW asking him not to use the macro, in the name of the fairplay.

What's left for this test to get completed is to see what will Need For Weed do (no answer in my msg so far)

PS. Btw, with this setup the current XFG@SO2R wr has at least 1 more tenth (according to my best sectors combined from different laps: 45.460+18.790).
Gearshift not activate if level manual clutch (axis or button) is less 90% (89,88 ...). Instant Hotlap Analyser on site lfsworld.net don't accurate instrument and cann't be argument.
Quote from arco :Well, you must have incredible fast reaction then, because from the attached graph you can see the difference in speed. Yours in red.

mate, im telling you, ask duck if u dont beleive me, every time i have to change gears, i press the right hand button on my g25
Quote from arrowkart4 :mate, im telling you, ask duck if u dont beleive me, every time i have to change gears, i press the right hand button on my g25

One side-effect of all this fuss is that now all hotlaps NOT showing the AC and the CL flag on the charts, become pretty much suspicious of using the macro.

The only safe way to have all people be 100% convinced that they are ethically valid laps is to be re-uploaded with the AC flag on

Is this a cheat?
(625 posts, started )
FGED GREDG RDFGDR GSFDG