As a bit of a spin off from this thread discussing allowing a server side option to force drivers to use manual clutch rather than auto clutch I think it would be interesting to get some fast, consistent and experienced drivers to do some runs with auto clutch + sequential then manual clutch (axis) + h shifter and compare the laptimes against each other.

An idea for a test plan that I have to try to get some good, meaningful data is the following:


The same setup must obviously be used for each shift mode in the same test. You must either flat shift in both circumstances or lift for shifting in both circumstances (clutch heat will probably be the limiting factor in this).

If anyone is willing to participate in this test I would really appreciate it. Please reply attaching SPRs (taken from Single Player practice or race mode with just yourself on the track) and give us some details like controller type, whether you lifted on upshift and whether or not you throttle blipped on downshifts.

Thanks

I'll Do some laps.. I've been running 900* with manual clutch lately and it's an absolute blast compared to autoclutching everything.

I was always faster with manual clutch and h-pattern than with auto and h-pattern or auto/paddles

With auto clutch on, the shift time is whatever time the game uses for the auto-clutch duration. With auto-clutch off, the shift time is the time it takes to press and release the clutch button if you exploit the manual shift sequence. I'm not sure which is faster. This is an example of the auto-clutch off exploit shift sequence:

For a sequential shifter the sequence for each shift is:

press and hold shift button before you want to shift.
tap clutch button to cause shift to actually occur
release shift button

Quote from JeffR :

Due to a quirk in LFS, auto clutch off will provide the fastest shifts. This is because you can select the next gear and hold it (if sequential shifter) and nothing happens until you press the clutch button, in which case the shift occurs near instantly.

For a sequential shifter the sequence for each shift is:

press and hold shift button before you want to shift.
press clutch button to cause shift to actually occur
release shift button

Oh, you can't really do this in the test ! It's not a cheat test !　

The tranmission system in LFS needs much improvement.

Quote from JeffR :

For a sequential shifter the sequence for each shift is:

press and hold shift button before you want to shift.
tap clutch button to cause shift to actually occur
release shift button

Quote from Keling :

Oh, you can't really do this in the test ! It's not a cheat test

Although my example is a bit extreme, a player can just press both shift and clutch at about the same time, and it doesn't matter which button is pressed first.

Quote from JeffR :

Due to a quirk in LFS, you can shift before using the clutch. With auto clutch on, the shift time is whatever time the game uses for the auto-clutch duration. With auto-clutch off, the shift time is the time it takes to press and release the clutch button. I'm not sure which is faster. This is an example sequence:

For a sequential shifter the sequence for each shift is:

press and hold shift button before you want to shift.
tap clutch button to cause shift to actually occur
release shift button

Axis clutch, not button clutch.

Quote from JeffR :

Although my example is a bit extreme, a player can just press both shift and clutch at about the same time, and it doesn't matter which button is pressed first. In case you missed the correction to my previous post, I'm not sure if this is related to auto-clutch or not.

Yeah. Almost forgot. We may be doing it without noticing.
LFS is not strict enough with shifting input.

Damn if i would have G25 i would use h shifter and manual clutch.

I've done some of the tests for you.

First of all, I used G25 wheel, 900 degrees rotation, when using manual clutch I heel'n'toed, with auto-clutch I braked with the left foot and blipped with the right one.

I've done two of the tests so far. I'm using H-gate shifter for more than a year now and got quite used to it, still mis-shifts happend from time to time.

In my tests I've had several attempts, the best ones are uploaded. With XRT I used ABS, to minimize the spreading of lap times.

ALL TEST DONE WITH Z15 Patch

Well, shouldn't really be harder to beat the clutch kick that auto clutch produces, it's very slow imo.

It's not hard to beat with manual Clutch.

Quote from Dennis93 :

Well, shouldn't really be harder to beat the clutch kick that auto clutch produces, it's very slow imo.

It's not hard to beat with manual Clutch.

Indeed, the shift time with manual clutch is faster, however what really brings the diffirence is the consistency of the driver, because you have to do several actions simultaniously which takes quite some time to learn doing right.

Quote from [RF]-art555 :

Indeed, the shift time with manual clutch is faster, however what really brings the difference is the consistency of the driver, because you have to do several actions simultaneously which takes quite some time to learn doing right.

exactly, doing a fast time with for example with the FZR is very hard to do with the H-Shifter in my opinion. I would never try to do anything with H-shifter in the GTR, but in the road cars, maybe

AFAIK CR|troy is very fast with h-shift, ask him to do this stuff.

Quote from NightShift :

I'm really curious to see what will come out of this, despite some people endlessly insisting that manual is faster than autoclutch,

Manual is faster ovar auto-clutch when you compare only one shift, fact. But when you add to the equation things like need for blipping the throttle, contol of braking, steering, moving the gear lever and pressing clutch pedal at exact times it really changes the picture. thats why AC rules in WR charts.

You need to be a robot to be faster than auto-clutch in all conditions...

@Dennis: FZR is indeed hard with H-shifter, but not impossible. You just need to practice a bit more and you can always skip gears

I'm not sure if you can get to a proper conclusion if you test it like that, there are to many variables playing a part if you just compare laptimes: tyres, driver mistakes, flatshifting, even on a short combo like AS2 you still got plenty of room for different lines/brake points and if you are 0.1 faster in the manual clutched run does not conclude that you lost those 0.1 in the other run because of the auto clutch.

edit: Maybe you should just go for one corner, with some practice it should be possible to be in the same time range every run, add a start/finish point and you've got a short test track. Then just do 10 runs on each clutch, take the average and you're a little closer to something meaningful (at least I guess so maybe not ^^).

Quote from [RF]-art555 :

I've done some of the tests for you.

First of all, I used G25 wheel, 900 degrees rotation, when using manual clutch I heel'n'toed, with auto-clutch I braked with the left foot and blipped with the right one.

I've done two of the tests so far. I'm using H-gate shifter for more than a year now and got quite used to it, still mis-shifts happend from time to time.

In my tests I've had several attempts, the best ones are uploaded. With XRT I used ABS, to minimize the spreading of lap times.

ALL TEST DONE WITH Z15 Patch

Here's the hard data:


So far the manual clutch has been slightly faster both times. In the XRT testing though you weren't all that consistent with the manual clutch. Was this the clutches fault or were you getting used to the track? Do you feel like the slight difference in best lap was due to your skill or the different types of clutches?

I will note that for the XFG test ART got a better start with the auto clutch... by about .10-.15 of a second. Also he flat shifted, but that just shows that he was able to get the clutch in just as quickly as the auto-clutch.

Quote from [RF]-art555 :

I've done some of the tests for you.

Interesting data, thanks

Quote from pik_d :

Here's the hard data:

Thanks for the analysis.

Quote from troy :

I'm not sure if you can get to a proper conclusion if you test it like that, there are to many variables playing a part if you just compare laptimes:

Agreed, the only meaningful way to test is to extract rafs for every lap, measure shifting time and then average out.

Quote from amp88 :

Thanks for the analysis.

No problem. I'm not promising I'll do that for everyone, but I have just as much interest in this as you.

How are you organizing these results (and whatever others come in)?

Quote from pik_d :

No problem. I'm not promising I'll do that for everyone, but I have just as much interest in this as you.

How are you organizing these results (and whatever others come in)?

I'm planning on writing a little InSim application that will provide a bit of in-depth analysis of the replays, but it depends on how many people actually take part.

Quote from pik_d :

Here's the hard data:
...................

Thanks, didn't had time to do it back then in the morning

Quote from pik_d :

So far the manual clutch has been slightly faster both times. In the XRT testing though you weren't all that consistent with the manual clutch. Was this the clutches fault or were you getting used to the track? Do you feel like the slight difference in best lap was due to your skill or the different types of clutches?

Partly both. I knew this track quite well, but I didn't had much time so I had less practice with the manual clutch on this track, I'm not the best driver in the world also

Quote from pik_d :

I will note that for the XFG test ART got a better start with the auto clutch... by about .10-.15 of a second. Also he flat shifted, but that just shows that he was able to get the clutch in just as quickly as the auto-clutch.

Yeah, I flat-shifted with both clutches in this test