Physics changing between empty and full online rooms

  • Thread starter jon-
  • 62 comments
  • 6,351 views
J

jon-

75
United Kingdom
London
TyreReviews
Right, I'll try and keep this short. I'm not sure if it's been covered before, I couldn't find it using a search.

I race in an online league which has full grids (16 people) most weeks. We mostly run standard cars with no modifications on sports soft tyres. Damage full, grip reduction real, tyre/fuel wear on.

During the week we all practice in small groups and as we're all pretty consistent, we find our groove and can mostly run within a few tenths of that.

A couple of times I've noticed on certain courses, courses I'd call "high draw" courses (lots for the playstation to process) the physics of the cars change when the room fills up and (mostly the slim users) get a little frame rate drop.

Case in point: Last night we raced on Special stage route 5 reverse with standard Lambo Merc LP640 SVs.

During mid week practice a number of us were running low 1.25s and high 1.24s in practice races without problem. Come race day, those of us who were running the 1.25's qualified 1st, 2nd and 3rd (all of us on slim PS3s). When the race started and the sheer number of people gave us all a little FPS drop the cars felt totally different, with the best times from all of us 1.29s.

Other people, who didn't qualify as strongly and are traditionally slower drivers were able to run their normal pace. Fighting them on track with identical cars and identical tyre wear it was like they were on different tyres! They had so much more grip to hit the apexes and power out of corners.

We even changed hosts for the second race and I checked the replay and we were all on the same tyres so we can rule out the following:

- cars
- setup
- tyres
- tyrewear issues
- driving style
- room
- internet connections

The only variance I can see trending is it seem to hurt those of us with a 'slim' PS3 more.

Any ideas? Known bug?
 
Last edited:
I do a lot of shuffle races, and I've noticed when it's a full room of 16, braking distances are longer. My usual braking points leave me overshooting the turn.
I have a slim, but I can't say that's the cause.
 
dzidza
I have a phat model, and when we have full lobby, lap times are drastically reduced.
Click to expand...

So it's not just the slims, interesting.

Have you ever noticed it affecting some people more than others? Because I race with the same 15 guys every week, the results are fairly similar. Last night, all those who should have normally finished at the front struggled, and those who normally don't quite have the pace didn't! :nervous:
 
This use to happen on prolouge too, I would race Suzuka when the room had 10 -12 players it would put 3-5 seconds on my lap time if there was only 2 or 3 players it would run faster and would be closer to offline times.
In both cases there was no lagging or car juddering.
 
Last edited:
Alfabob46
This use to happen on prolouge too I would race Suzuka when the room had 10 -12 players it would put 3-5 seconds on my lap time if there was only 2 or 3 players it would run faster.
In both cases there was no lagging.
Click to expand...

I wouldn't call the issue we had lagging, eg the cars weren't jumping around, it was more a slight framerate drop where the ps3 couldn't render everything going on.

PS, I live in Hitchin, small world.
 
dzidza
I have a phat model, and when we have full lobby, lap times are drastically reduced.
Click to expand...

Correction: lap times are significantly different through the grid with the same car, non-dependable of any known factor.

While some can run "normal" times - same as in Free Race (qualification) - some players are getting handicapped with no reason for up to 15 seconds per lap.

It happens in rooms with 12+ players.

My personal opinion is that is bandwith/ping issue where both host/client consoles have problems with crunching so much data because of large player filed (12+ players) and P2P connection resulting with "slowdown" for certain cars because their bandwidth somehow "shrinks".
 
amar212
My personal opinion is that is bandwith/ping issue where both host/client consoles have problems with crunching so much data because of large player filed (12+ players) and P2P connection resulting with "slowdown" for certain cars because their bandwidth somehow "shrinks".
Click to expand...

I think you're on the right track, but I'm not sure it's bandwidth related, more processor related.

In our example race last night, for the first race I was one of the last to join the room, and the second race I was host. It seemed to be the same mix of people going slow for no reason and others able to run their normal pace.

Being host made no difference to me, and I was still 5 seconds a lap slower than my raw pace.
 
Double-T
I do a lot of shuffle races, and I've noticed when it's a full room of 16, braking distances are longer. My usual braking points leave me overshooting the turn.
I have a slim, but I can't say that's the cause.
Click to expand...

That might be because your using CS tires.
 
Wow, are we talking the possibility of 3 physics engines? Offline, online full, online, empty?

What's going on there?
 
Spagetti69
Wow, are we talking the possibility of 3 physics engines? Offline, online full, online, empty?

What's going on there?
Click to expand...

I don't think it's another physics engine, that would be extreme, more that the online engine can't keep up with everything else going on and the car handling suffers.

Whatever it is it's making Forza look inviting :(
 
jon-
I think you're on the right track, but I'm not sure it's bandwidth related, more processor related.

In our example race last night, for the first race I was one of the last to join the room, and the second race I was host. It seemed to be the same mix of people going slow for no reason and others able to run their normal pace.

Being host made no difference to me, and I was still 5 seconds a lap slower than my raw pace.
Click to expand...

No, he's right. The thing is, if there's any lag the game's timer will keep running even though you weren't moving during the lag period. This can happen multiple times per lap equaling slower lap times. I noticed it while hot lapping, someone would jump on the track or off the track, the game would stop for a second, and I'd be down roughly a second or two. It could be a combo of internet and processor, but regardless the timer will still run even if the game freezes for a split second.
 
Large rooms are awful with my slim I usually back out when there is more than 8 people because it ruins my races, it even affects my drag racing.
 
Are you sure it's not the catch-up mode? Like GT5 Prologue, if you were at the back it gave you more grip. At the front you'd has less grip.... I remember people talking about this in GT5p.
 
Kazanski
Are you sure it's not the catch-up mode? Like GT5 Prologue, if you were at the back it gave you more grip. At the front you'd has less grip.... I remember people talking about this in GT5p.
Click to expand...

As much as I would LOVE it to be this, we race every week and there's 16 of us who are very anal about the settings so we're 100% sure it's not 'boost'.

So far the leading theory is

- Frame rate drops
- Your PS3 time slows down
- Host server time doesn't
- Your laptimes get longer
 
Kazanski
Are you sure it's not the catch-up mode? Like GT5 Prologue, if you were at the back it gave you more grip. At the front you'd has less grip.... I remember people talking about this in GT5p.
Click to expand...

That's the infamous 'rubberband'-effect, now called boost which you can set at high, low or not at all (which I prefer) so I hardly think that's the cause here.
I've experienced wildly different car behaviour too online, and in hindsight it may be due to a large number of players (roughly more than 8) as I hardly experience any major differences when in a room with a small number of players.
I have simply no idea what causes it but it sure is annoying.
 
analog
I have simply no idea what causes it but it sure is annoying.
Click to expand...

It wouldn't be /that/ annoying if the effect was uniform across the grid, but it becomes REALLY annoying when it only affects half the field so you've some people drastically handicapped and others able to run at their normal pace!
 
I have done a fair bit of online racing myself and I have not noticed my times being slower due to those reasons and I also have a slim.
Maybe the rooms I race in are not large enough (usually 6 - 10).

Jon, have you checked out this thread?

Would be interesting to hear you and your buddy's comments on those too, since you also have some experience racing online.
 
Small_Fryz
I have done a fair bit of online racing myself and I have not noticed my times being slower due to those reasons and I also have a slim.
Maybe the rooms I race in are not large enough (usually 6 - 10).

Jon, have you checked out this thread?

Would be interesting to hear you and your buddy's comments on those too, since you also have some experience racing online.
Click to expand...

I've never noticed it below 10, I think I've only noticed it above 12 people and only on 'high draw' tracks (SS R5, Daytona Road course etc).

I did see your thread but sadly it's not something I've ever noticed. I'll keep a look out though as sometimes my tyres do seem to wear surprisingly fast.
 
jon-
It wouldn't be /that/ annoying if the effect was uniform across the grid, but it becomes REALLY annoying when it only affects half the field so you've some people drastically handicapped and others able to run at their normal pace!
Click to expand...

Agree, to me that suggests it might indeed have something to do with individual bandwidth or connectivity aspects since if it the physics were toned down due to the limitations of the game/hardware the effects would probably be the same for everyone.
Might the quality (indicated by those stars) of the room have an effect too or is it simply potentially happening in any room populated by a large number?
Although even when it's due to bandwidth/connectivity/whatever outside reason, that perhaps should've been a consideration to take onboard when implementing the online aspects of GT5, if it can be dealt with that is.
But it's also a perfect excuse to use when you can't keep up I guess. ;)
 
analog
Agree, to me that suggests it might indeed have something to do with individual bandwidth or connectivity aspects since if it the physics were toned down due to the limitations of the game/hardware the effects would probably be the same for everyone.
Might the quality (indicated by those stars) of the room have an effect too or is it simply potentially happening in any room populated by a large number?
Although even when it's due to bandwidth/connectivity/whatever outside reason, that perhaps should've been a consideration to take onboard when implementing the online aspects of GT5, if it can be dealt with that is.
But it's also a perfect excuse to use when you can't keep up I guess. ;)
Click to expand...

I'm still of the opinion it has little to do with bandwidth. We do 2 races per weekend back to back, the first room was so bad for me I got everyone to move to a room hosted by me and still had the issue.

There are lots of variables on the playstation itself that could cause different CPU loadings - Full HF or 720, progressive smoothing, having bluetooth on for pad / headset, having to process data from a wheel etc.

I'm going to research this some more with my 15 helpers :)
 
Bit more of an update. I've canvassed all the drivers for the following

Resolution
Control method
Mic status
View
Version of PS3
How much they think they were affected.

Interestingly we almost all ran 1080i/p, there was a mix of slims and fats affected, it didn't seem to matter if the mic was on or not, however there was one clear difference.

Of 12 drivers who replied 7 were on a wheel and 5 on a pad.

Of the 7 drivers on the wheel 6 drivers reported being slowed by around 5/seconds a lap.
Of the 5 drivers on the pad only 1 reported being slowed, and only very slightly (compared to the wheels huge loss)

If there any official way of submitting a bug report to PD? I don't think we're going to get any further here.
 
Oh, the other interesting thing is we timed some of the replay laps against a stopwatch.

1.28.9 in game, 1.27.1 on stop watch
1.30.6 in game, 1.28.3 on stop watch
1.32.6 in game, 1.30.6 on stop watch

Second race 1st lap 1.49.2 - stop watch 1.45.0.

I timed someone who "wasn't affected", and while there was still a variance between the in game time and the stopwatch time, it was less...
 
Hi Jon,
Thanks heaps for sharing all your testing, it would be great to get to the bottom of this.
jon-
Bit more of an update. I've canvassed all the drivers for the following

Resolution
Control method
Mic status
View
Version of PS3
How much they think they were affected.

Interestingly we almost all ran 1080i/p, there was a mix of slims and fats affected, it didn't seem to matter if the mic was on or not, however there was one clear difference.

Of 12 drivers who replied 7 were on a wheel and 5 on a pad.

Of the 7 drivers on the wheel 6 drivers reported being slowed by around 5/seconds a lap.
Of the 5 drivers on the pad only 1 reported being slowed, and only very slightly (compared to the wheels huge loss)

If there any official way of submitting a bug report to PD? I don't think we're going to get any further here.
Click to expand...
Wow, that wheel thing stands out heaps. You know that GT5 has much less tearing etc when you come down to 720p? Maybe you could switch to a low res and see if you're still slowing down. Also, turning off any "distractions" for the PS3 (mic / bluetooth / chat sessions) might give some insight.

I've noticed that when you pause a race (offline), it always causes a slower lap. For example, straight line acceleration suffers even if you are holding full throttle all the time. Perhaps this is related.

jon-
Oh, the other interesting thing is we timed some of the replay laps against a stopwatch.

1.28.9 in game, 1.27.1 on stop watch
1.30.6 in game, 1.28.3 on stop watch
1.32.6 in game, 1.30.6 on stop watch

Second race 1st lap 1.49.2 - stop watch 1.45.0.

I timed someone who "wasn't affected", and while there was still a variance between the in game time and the stopwatch time, it was less...
Click to expand...
Yikes!

I can accept it is necessary to "fudge" time to account for lag or whatever, but if this affects drivers differently, then this totally stinks.
 

Similar threads

hawkeyez
Performance Point Differences
Replies
16
Views
1K
Monkdapunk
Monkdapunk
benjoi84
Allow full screen "drone footage" demo during lobby waiting screens.
Replies
0
Views
541
benjoi84
benjoi84
M
PD Firmly stuck in Hard Tyre Syndrome again!
2
Replies
32
Views
3K
Mudda Chod
M
TheWeapon874
Tuning help request: M3 GT '11 at Monza
Replies
2
Views
1K
TheWeapon874
TheWeapon874
F
Compulsary Tyre in online lobby
Replies
2
Views
1K
Fidjeb
F

Latest Posts

Back