Yeah, But Does Your Tune Work ONLINE? (Possible Future Hassle for Tuning Community?)

[budious]

I am now going turn around and say it is all down to lag. More seriously, I do not know.

My observation after racing online over the last few days is that online physics seem to change like the weather here in the UK. A good example is my Elise 111R RM using madmike's tune. Offline it is fantastic, online it was a pig, another day online it is about 90% right. Similar results with other cars that I have tuned.

Something is not right somewhere, but I am just going work on the basis that every online day will have different physics.

edit: ok i see the RM is stuck with semi-race now and can't be down spec'd which sucks cause it has the same problem the 111R did with semi-race flywheel, workaround is to set your LSD accel up to about 50, should straighten you out.

I'm not familiar with madmike's exact tune, but the conclusion of my testing is now updated with Sports flywheel testing on my article.

As tested for me on my Elise 111R;

"The ultimated drivetrain for the Lotus Elise 111R is the Single Plate Clutch, with a Sports Flywheel, and Customizable Limited Slip Differential with an acceleration value of 50."

 

[StevenDunn99]

workaround is to set your LSD accel up to about 50, should straighten you out.

Thanks I will try that. 👍

 

[xlx_Bobby_xlx]

I am now going turn around and say it is all down to lag. More seriously, I do not know.

My observation after racing online over the last few days is that online physics seem to change like the weather here in the UK. A good example is my Elise 111R RM using madmike's tune. Offline it is fantastic, online it was a pig, another day online it is about 90% right. Similar results with other cars that I have tuned.

Something is not right somewhere, but I am just going work on the basis that every online day will have different physics.

what?! i find his 111r tune perfectly fine online, ive ran 6:40s round the ring online destroying 500hp+ cars

 

[MadMike'86]

what?! i find his 111r tune perfectly fine online, ive ran 6:40s round the ring online destroying 500hp+ cars

👍 She's a beast lol

 

[StevenDunn99]

what?! i find his 111r tune perfectly fine online, ive ran 6:40s round the ring online destroying 500hp+ cars

I can only report on my experiences. I am a fan of Madmike's tunes but I find online 'physics' inconsistent.

 

[SS_1182]

Okay, friends. Here's a little test for you then. Imprezas are possibly my favorite cars on the game. One of my obvious favorites for street car racing is the WRS STi Spec C Type RA '05 Impreza (NOT RACE MODIFIED VERSION) which I think I have set up really well for ONLINE Nordschleife.

So anyway, just for kicks I went to the practice area (OFFLINE), to see what times I could put up on my personal leaderboard. Immediately I had trouble steering this car due to offline UNDERSTEER. Not only that, the anti-roll bars were far less effective offline, really it was quite striking. I couldn't even safely run two tires over any rumble strips.



My theory is true that there is more oversteer online, test this tune for yourself. Drive it FIRST in an ONLINE lobby, and I don't mean the lounge. Create a Nordschleife track with very high race quality and run at least a lap with this tune. Do this FIRST, and see how it feels. My setup may not suit you perfectly, but what we're looking for here is the presence of oversteer/understeer. Just see how the car handles after running a lap or two, and gauge those characteristics in cornering.

When you're done, go straight to the practice area OFFLINE track and run two laps on Nordschleife again and try to tell me that the car doesn't understeer when you get there.


Again, this tune is for the street version of the Spec C Type RA '05 Impreza WRX, NOT the race modified version.


And you're welcome for the tune. (assuming you like it)


**Tune this car to 500hp if you have different engine upgrades than I do for this test.

Aero: 0/20
Engine: Stage 1 only, ECU Tuning
Trans: 193mph
LSD: 10/40/30 front 10/55/35 rear
Torque Split: 35/65
Height: -15/-7
Springs: 10.6/9.7
Ext: 5/5
Com: 5/5
Roll: 2/3
Camber: -2.0/-0.5
Toe: 0.00/0.00
Brakes: 6/3
Tires: Racing Soft




Curious to hear the results.

 

[budious]

Did not try your test but you could be experiencing the same differences I record in offline practice when there are variations in the rate of tire temperature, if the front tires warm quickly but the rear ones remain shades of blue longer, then you experience understeer during cornering until all tires reach equilibrium. Since you run a 35/65 split but favor the front tires with a 6/3 brake balance then the rear tires may be wearing faster negating the understeer effect of the tires not having reached temperature equilibrium.

Understeer and oversteer are influenced by all the following factors:
Tire temperature
Tire wear and circumference difference
Camber behaves differently due to tire temperature variations

 

[Corthalis]

Online difference has little to do with tire temperatures and wear. It's just different physics and if you've played online enough then you'd notice. Too much theorycraft on this thread on what's different when it's so easy to just jump in, test the damn tunes, and notice the difference straight away. You can do a few laps to heat up your tires if you want, without wearing them too much, and on your fastest hotlap the cars will still feel a lot different from offline. You can even use arcade garage cars to eliminate the tuning variable. Guaranteed you'll be lapping at least 1 second quicker around most tracks in arcade mode or practice mode in GT life compared to opening your own lobby and doing the same laps in the same arcade car.

 

[budious]

Well if you mind testing a control condition along with your problem condition to compare the results; then retry the same tune but with 50/50 torque split and brake balance 5/5 and compare how both handle on and offline.

 

[kilttraeger]

Online difference has little to do with tire temperatures and wear. It's just different physics and if you've played online enough then you'd notice. Too much theorycraft on this thread on what's different when it's so easy to just jump in, test the damn tunes, and notice the difference straight away. You can do a few laps to heat up your tires if you want, without wearing them too much, and on your fastest hotlap the cars will still feel a lot different from offline. You can even use arcade garage cars to eliminate the tuning variable. Guaranteed you'll be lapping at least 1 second quicker around most tracks in arcade mode or practice mode in GT life compared to opening your own lobby and doing the same laps in the same arcade car.

Fully acknowledged.

If you don't want to take SS_1182' tune and give it a try, take a real tricky car (Ruf BTR, Yellowbird) that you come to love and manage (or even master) offline and take it online.
No matter how much time you give it to heat up your tires, or how much fuel is in, I bet you will spot the difference immediately.

Taking a rather civil and easily drivable car (M3's for instance) will make it less obvious to spot.

 

[xSNAKEx]

I've heard this a lot from people, but ive never noticed anything mysel.f Im fine with the online physics.

I would think any difference felt is cold tyres and then tyre wear but many of you seem to be saying thats not it either.\

Though, why would you even bother tuning for offline? It makes no sense to me at all. YOuou can beat the AI on any event easily. The real competiton is online so you shoulnt even consider tuning for offline purposes.

Anyway, some said standard model cars have better aerodynamics in this thread.

Is this true? for example instead of using the premium r34 youre better off using a standatd r34?

 

[budious]

.

Anyway, some said standard model cars have better aerodynamics in this thread.

I know it's confusing because I throw out theories seriously sometimes, then make seemingly serious sounding theories as a sarcastic pun directed upon myself... that didn't answer the question did it?

 

[HKS racer]

I absolutely quote the OP. This problem is true.

My team is running a championship with GT300 cars, and in the first races we were tuning our cars offline.. then after we realize the difference, we all tune our online cars in an online lobby, possibly the one we are going to race. It's a big shame that some serious setup work in here at gtp won't be very useful online, I remeber the days of GT5p, everything was easier.. and that love/hate PP system.. I think GT5 really need an improved PP system from the prologue. We need something for balance the performance of the cars, seriously Mr. Yamauchi, we need that. HP/wieght is not enough

 

[budious]

I'm still of the opinion if PD simply patched the game to allow the race cars to be equipped with lower (not really lower but condition specific use cases apply) drivetrain components this would fix most of the cars without major changes in the setup for online play.

Standard Clutch - slow shift, lots of inertia; typically least efficient

Single Plate Clutch - faster shift, less inertia but enough to remain balanced in deceleration and acceleration, best option for rolling starts and multiple lap time trials

Twin Plate Clutch - fastest shift, least inertia, produces a great first lap time from a stand still grid start but slower average lap times than single plate clutch; under similar conditions as a grid start, if you spin out or come to complete stop, it is quicker to get back to pace.

Stock Flywheel - has the most inertia; and slowest transition between low to high to low engine rpm changes; can be beneficial on tracks where frequent tight corners require the driver to be on/off the throttle

Sport Flywheel - has less inertia than stock; but a slightly faster transition between low to high to low engine rpm changes; can be beneficial on tracks where tight corners and hairpins are part of the layout accompanied by long straights

Semi-Race Flywheel - has least inertia; but the fastest transition between low to high to low engine rpm changes; can be beneficial on tracks where speed is maintained through wide radius corners or those with banks and momentum changes are less significant.

Apply parts where most beneficial for track layout on standard cars and you will notice improvements; unfortunately you don't get this level of drivetrain flexibility with RM, LM, and race cars.

 

[Ghost Rydor]

Tune your car in your personal online lounge, not time trial. Generally tuning done online will be backward compatible with GT life but if you tune in GT life it often is too loose for online.

 

[budious]

As for the qualities of GT5 physics in general, you only need to look insofar as the video posted in this thread: https://www.gtplanet.net/forum/showthread.php?t=178977

 

[HKS racer]

Would be interesting see how the new PP system will work in GT5. Let's see if it will advantage particular type of cars. I remember the prologue PP was a bit weird, Renault Clio faster than Corvettes in some tracks..
I hope this new system help to balance the cars better and help us eventually reduce the hp value if you cannot enter a particular time trial for just 3 hp

 

[BlueShift]

Thoughts?

You're rigth but it's only because of tire wear.

Wheel diameter change in offline, not in online, as err... somebody said in another thread.

For DS3 users, it's a great pain "where the sun never shines", because all accel LSD must be set -4 if the race goes more than 2 laps. For wheel users, they just don't care. 5° less while steering and that's it.

I hear PD : "You should all buy driving wheel, buy ours, it's the best !"

 

[budious]

I remember now where I saw the option to use tire wear enabled in offline practice, you only get the option if you select One Make Race on the offline practice mode screen. You'd think it be available for Free Run and Drift Mode also...

Also, if you make custom courses, and open the My Courses icon in GT Life, click one of them, there are three green icons at bottom, the first is Single Race. This also has the option for tire wear and fuel consumption enabled, it's slightly different as the computer will pick similarly powered cars to yours like the Arcade mode does rather than only one make races.

Either of these practice race modes work like free run, just exit anytime you want, make changes, start racing again except you have to put up with traffic on the course. For the argument of eliminating other factors such as network connection speed, this should put you into a purified offline endurance tire wear and fuel consumption mode for practice tuning.

 

[nomis3613]

I have just produced online versions for 5 cars, each at a different track, here are my observations:

DB7 on Sports Soft at Trial Mountain
After bending over backwards to reduce the understeer offline, taking it online suddenly gave me an oversteering demon! To fix the balance, I removed the ballast and rake. Once this was fixed, I noticed a strange phenomenon: the handling was smooth, but it would slowly start to slide at the rear of it's own accord. Stiffer anti-roll was needed to fix this.

Ferrari F40 on Sports Soft at Tokyo R246
This had the least difference between online and offline of the cars I tested. Online, it felt like the suspension was softer and there was less tyre grip but the balance/transients/etc were very similar.

Golf I on Sports Soft at Autumn Ring Mini
A bit more understeer online. Needed more -ve rear toe to fix it.

Integra DC5 RM on Racing Soft at Tsukuba
My offline tune has very crashy suspension, taking it online felt like a smooth magic carpet ride! Also a bit more understeer online.

Audi RS6 Avant on Racing Medium at Laguna Seca
My tune which was nicely balanced offline (IMHO...) became very oversteery and unpredictable online. Less rear TCCD bias was needed and less rake. This improved things, but I can't get it as nice as offline.

I have seen posts suggesting the difference is due to lag. I disagree for 2 reasons:
1) it seems GT5 only uses PSN to get info about other cars location, not the behavior of your own car (a common programming technique for network games). This is why you see opponents cars jumping around if you have a slow connection. So all of the physics calcs are done on your own PS3 without any waiting for the network.
2) it doesn't feel like lag. If you have an LCD, turn on 100Hz mode to experience what driving with lag feels like. Online doesn't feel like this.

My belief is that it's due to fuel weight differences (yes even though the fuel guage reads full for both online and offline). Removing the rear ballast on the DB7 got rid of most of the handling differences; the DB7 is otherwise very nose-heavy so maybe the fuel is acting as rear ballast. That the F40 has 50/50 balance and didn't suffer any balance changes also supports this, I reckon. As for the Integra feeling "soft" online, 120L of fuel makes a big difference to a 950kg car, and heavier cars need stiffer springs to control the weight. Anyway, that's my conspiracy theory for online vs offline handling.


Budious, if we could take a brief tangent about your drivetrain theories:

Standard Clutch - slow shift, lots of inertia; typically least efficient

Twin Plate Clutch - fastest shift, least inertia, produces a great first lap time from a stand still grid start but slower average lap times than single plate clutch; under similar conditions as a grid start, if you spin out or come to complete stop, it is quicker to get back to pace.

What do you mean by "efficiency"? (I've never noticed any straight-line speed reduction (from a rolling start) with any of the "inferiour" drivetrain parts)

So a twin plate clutch only has a benefit for launching from a standing start?

Stock Flywheel - has the most inertia; and slowest transition between low to high to low engine rpm changes; can be beneficial on tracks where frequent tight corners require the driver to be on/off the throttle

Totally agree. Thanks to you I don't just slap on the "best" drivetrain parts any more 👍

 

[V4.SKUNK]

I think tyre wear on, grip reduction @ realistic and no skid force recovery is the reason we got noobs here complaining about tunes not working on-line.

 

[CSLACR]

I think tyre wear on, grip reduction @ realistic and no skid force recovery is the reason we got noobs here complaining about tunes not working on-line.

There are major differences online to offline, if your offline built tune works online, it means your offline tune is under steering fairly heavily. In every car I've driven online, and every track I've driven online so far, over-steer *should* be in pretty much every turn with a proper offline tune.