How different are actually offline/online physics?

  • Thread starter Thread starter Pseudopod
  • 31 comments
  • 2,426 views
Messages
1,079
Norway
Norway
I see people mention this again and again. I don't doubt the validity of these claims, but I also don't know much about these differences, as I don't play online very often (almost never, actually). Even if I did, I'm not really a very good GT player, so I'm not sure how much I would notice anyway.

Could anyone give a good explanation of which things are different, and possibly why they are different?
 
It's hard to pinpoint exactly what is different, but from my experience, offline has more grip. So i think it's more related to the track and not the physics.

When online, if you set the grip to real and goes off track, you will have to wait at least one corner to get the normal grip back. Offline, you almost instantly get the normal grip when you come back on track.

So, if you want to use a car online, tune it online because the sweet spot will be different. I personally never use arcade mode, if i want to have a drive alone i make a room, it's much easier to change the car and tracks.
 
So, if you want to use a car online, tune it online because the sweet spot will be different. I personally never use arcade mode, if i want to have a drive alone i make a room, it's much easier to change the car and tracks.

I never thought about that! You're right. I tend to take a car onto a track and run 4-7 laps with it, then want to switch to a different car and this can go on for a couple of hours. Having to back out all the way to the main menu and then change cars and re-select the track is a pain in the buttocks. :mad:

Like the OP, I too have heard that "offline times are faster than online times". It must have something to do with grip, right? I'd really like to learn why this is true. I can accept that it is, no problem, but I need to know why...
 
Like the OP, I too have heard that "offline times are faster than online times". It must have something to do with grip, right? I'd really like to learn why this is true. I can accept that it is, no problem, but I need to know why...

This is exactly my finding, especially after trying the LMP1 Nissan Nismo in the final round of GT Academy at Le Sarthe.
I only managed to get bronze in the competition no matter how hard I tried with the car fighting for grip when cornering and bouncing / grounding out on the straight.
Now when I tried the same car, same default settings (suspension, tyres, gearing etc) offline in time trail, I managed a lap that was almost 4 seconds faster and I found it was cornering grip that was the reason. Bouncing and grounding was about the same on the straights so that didn't help with my lap time, only the cornering did so I'd like to find out the reason why ?
 
I never thought about that! You're right. I tend to take a car onto a track and run 4-7 laps with it, then want to switch to a different car and this can go on for a couple of hours. Having to back out all the way to the main menu and then change cars and re-select the track is a pain in the buttocks. :mad:

Like the OP, I too have heard that "offline times are faster than online times". It must have something to do with grip, right? I'd really like to learn why this is true. I can accept that it is, no problem, but I need to know why...

Online is so much better for that, changing tracks just takes a couple of seconds, you have all of your garage available, and when you change cars the track doesn't load again, so you have the exact same conditions.

Don't forget to set the room privacy to "friends only" and limit the players to 2, this way no one will bother you. And you can save your room settings, so it's pretty easy and much faster to just turn on the game and play.

PS: You can be disconnected from your own room. :lol:
 
I use offline because it's faster to change car or its settings. I don't change tracks much.
 
The physics MAY be more simplified for more efficient online usage, but my suspicion is that online differs more because the server is correcting where you are at on the track.

Offline, the box computing your car's movement is right in front of you. Online, everything that happens locally is sent to a sever for verification. You might find that with slower connections, you have more understeer. I always notice that I am a FAR better driver when I am the host. :)
 
The lap times are offline definitely faster then online. I use to drive constant laps on Nordschleife with my tunes and my cars so I notice every grip change. Offline I'm minimal 5 sec faster.
 
The physics MAY be more simplified for more efficient online usage, but my suspicion is that online differs more because the server is correcting where you are at on the track.

Offline, the box computing your car's movement is right in front of you. Online, everything that happens locally is sent to a sever for verification. You might find that with slower connections, you have more understeer. I always notice that I am a FAR better driver when I am the host. :)
In that case, making an online room to practice alone in wouldn't really help you get better, as you would always be the host. You wouldn't get the lag and corrections that you would get if someone else was the host.
 
The physics MAY be more simplified for more efficient online usage, but my suspicion is that online differs more because the server is correcting where you are at on the track.

Offline, the box computing your car's movement is right in front of you. Online, everything that happens locally is sent to a sever for verification. You might find that with slower connections, you have more understeer. I always notice that I am a FAR better driver when I am the host. :)

In that case, making an online room to practice alone in wouldn't really help you get better, as you would always be the host. You wouldn't get the lag and corrections that you would get if someone else was the host.

Just my 2c, i never experienced that and i did a lot of league racing and single races online here in GTP, that means that everytime before the race i take a few days to just test the car and simulate the event that i will participate. And i always end up making the same times and having the same level of grip,understeer,oversteer when i am the host or not (i am never the host on the official race day).

I'm not saying it can't happen, is just something that is not very common for me. 👍
 
In that case, making an online room to practice alone in wouldn't really help you get better, as you would always be the host. You wouldn't get the lag and corrections that you would get if someone else was the host.

I'm not saying it happens for everyone, but I definitely notice it. In races that have tire wear and real grip, I notice a BIG difference between pre-race practice and the race itself.

I know for a fact that optimized physics for online play is a common practice, but I can't say for certain that it exists in GT6.
 
I've also found this.

To be specific, when trialing a car to use in Quick-match events.
Turns out I simply didn't have the correct settings for my offline tests.

To my experience, once offline tracks are set appropriately to the race online settings, handling/lap times are similar.

Personally, after being 'tricked' into running faster lap times offline as opposed to online, and then finding the reason why, I'm not convinced of this 'apparent' difference.
 
I never thought about that! You're right. I tend to take a car onto a track and run 4-7 laps with it, then want to switch to a different car and this can go on for a couple of hours. Having to back out all the way to the main menu and then change cars and re-select the track is a pain in the buttocks. :mad:

Like the OP, I too have heard that "offline times are faster than online times". It must have something to do with grip, right? I'd really like to learn why this is true. I can accept that it is, no problem, but I need to know why...

Yea man I really hope they fix that in the future. You shouldn't have to be online to switch tracks and cars like that. If you can do it online that quickly, you should be able to do it offline even faster I'd assume.

Hopefully PD realizes it.

You we should definitely add that as a feature request if it's not already on there.
 
I'll play the devils advocate. :lol:

Personally I've found very little difference between online and offline regarding changing tracks.
And from memory (a little rusty granted) cars either.

Am I missing something significant here?
 
Online always, you can change car & track without having to back out of the session which is a lot faster, Also I run tyre wear on real for a more realistic level of traction within my setups.
Personal opinion of course.
 
I use offline because it's faster to change car or its settings. I don't change tracks much.
No it isnt. In offline, you have to exit the track everytime if you want to change car. In online free run, you can change your car right there anytime.
 
It's much quicker for me to test as many cars as I do, offline rather than online. I can save my best lap replay, too.

I'll have to host a room and see if I can meet or beat any of my Mid-Field lap times to test this theory...
 
Online physics take the fuel weight into account... Offline the tank will show full, but it's not. The difference in feel online is that the car is more planted but slower due to the weight. In my biased opinion, of course. ;)

Can be verified by running a car down to 10L of fuel and pitting for tires only. Then you can match offline/Arcade times.
 
That would be an interesting explanation. Maybe someone should do some testing for that. I can't, because I'm not really consistent enough with my lap times to make any meaningful statistics :p
 
Online physics take the fuel weight into account... Offline the tank will show full, but it's not. The difference in feel online is that the car is more planted but slower due to the weight. In my biased opinion, of course. ;)

Can be verified by running a car down to 10L of fuel and pitting for tires only. Then you can match offline/Arcade times.

Interesting theory. Got to test that :)
 
Online physics take the fuel weight into account... Offline the tank will show full, but it's not. The difference in feel online is that the car is more planted but slower due to the weight. In my biased opinion, of course. ;)

Can be verified by running a car down to 10L of fuel and pitting for tires only. Then you can match offline/Arcade times.

^^exactly that^^
 
How do you know online takes fuel weight into account?

But putting 2+2 together, as gravitron says if you're online and run your fuel down and put on fresh rubber it is possible to match the quicker offline times.
 
But putting 2+2 together, as gravitron says if you're online and run your fuel down and put on fresh rubber it is possible to match the quicker offline times.

But if you run at full fuel and no tire wear the entire time, then your lap times online and offline should be the same.
 
Yeah online with tyre/fuel off, same story. Over a 2min lap online you'll get about a second difference between tyre/fuel on/off ... until you run your fuel down to almost empty, fresh tyres and the times match again.
 
My (almost completely unfounded) theory is this: The physics simulation calculates a certain number of times per second how the car behaves with the inputs it gets. Offline, this can be done much quicker than online (it doesn't need to send that in the same frequency to the server), which is why they may use a slightly lower frequency in these calculations for running in an online lobby to avoid lag. Now, if the frequency of these calculations has an effect on grip, then you would naturally get slightly more grip offline (the same "amount" of grip for each occurrence, but offline would have just more of these "events").

This might sound far-fetched, but have you ever wondered why other people's cars in an online lobby sometimes produce pulsating tyre squeal? This could be an effect of that. Or it's just the lobby quality setting.

Or the whole thing could be something completely different. :lol:

ps: I'm not convinced by the "offline is the same with low fuel" theory. It cannot be disproven, but just because a car can approximately match the offline-times with low fuel online doesn't prove that's the reason why the speed difference is there in the first place. It could still be something completely unrelated to fuel that has about the same effect in terms of lap time performance.
 
@snowgt Isn't offline updating 1000 times per second? Even a fifth of that would be simply impossible to arrange with the expected comms distances.
 
Back