GT5's Game-Breaking Online Flaw (OP Updated: 11 Feb)

[CSLACR]

I really don't know how you can be so sure of that. You do know people use entire clusters of ps3s to build super computers right? You can bet your arse they're running for more than 8 hours straight!

If your cpu needed a break after running your fluid simulation for more than 3 hours then you did not have adequate cooling in the first place. Fact.

And btw. In no way does every game use 100% all the time. I own Pac Man on ps3 for example. If there was some kind of task manager for ps3 you might see it showing 100% cpu usage or something. It may well consume all available cpu cycles but I bet it doesn't even touch the Cell chip!

I'm not sure what any of that has to do with anything, but it seems to contradict itself.
I do know we have this awesome little orange light that tells us how active the HDD is, and if that's not solid orange, it's not going 100%.
Installation of GT5 data is a good time to see it at solid orange. Fan runs more, system gets hotter.
I sure as hell don't want to try keeping that light lit like that for hours on end, take that for what you want.

 

[rsergio007]

Yes.
Click the name, then click "view profile", at the top of their profile page, click "user lists" and the bottom option will be "add to ignore list".

Thank you Sir!

 

[BWX]

Saw people talking about missing sounds.. That's been a problem since 1.06.
(Just for the record)

Feb 24 2011, 11:00 PM :
https://www.gtplanet.net/forum/showthread.php?p=4919947#post4919947

A couple days after the update I noticed that in a race around Nuburg that my 2010 (sedan) Impreza's tire sound were gone completely. I restarted the race a few times, same problem. I had to end race, go to main menu, then start race again for tire sounds to work again.
Yesterday I was in a b-spec race and one of the ai cars - it was an RX-8 prototype LM Car with 787B paint job had NO exhaust sound at all. I could hear the engine and gears whine clearly, but from any view there was no muffler/exhaust sound at all. This bug where a specific sound will be gone has only ever happened in GT5 since 1.06.

May 18 2011, 5:33 PM
https://www.gtplanet.net/forum/showthread.php?p=5331039#post5331039

Cainius:
anyone notice a lack of tiresqueal at toscanna?

That's that audio bug from a few updates ago. Basically, with certain cars on certain tracks, they sometimes have a soundmissing. It could be engine sounds, transmission sounds, but mostly all tiresounds will be gone from your car. The only way to get them back is to jump out of event completely- not just restart the race, and sounds will come back, but only for one race. If you re-run the race the same sounds will be gone again. I've tested this thoroughly.. it's annoying as hell and PD are not fixing it. It's been there for 3 or 4 updates.. they probably don't even know it exists.

 

[xSNAKEx]

As I see it, there are thing it cant be and things it can be. 'The PS3 can't cope' angle is IMO a dead end. EVERY SINGLE GAME USES 100% OF THE PS3'S ABILITY. It is not like any developer is not going to use all the power available. Some games look better than others because more time has been spent finding ever more clever and time consuming ways to do more in the same amount of time. And as has already been said, offline should push the game engine quite a lot harder.

I also don't believe that a persons internet connection is going to change code on a PS3 as has been suggested. Time needs to be spent thinking of what is possible. For example:

What if the room settings are incorrectly sync'd? I have personally once entered a race with a disallowed car. I can not remember exactly how I did it but I remember it was because there was a lot of lag between pressing the enter track button and my car appearing on the track. I think I was on track and I knew the host was about to change the settings. I hit the back button to exit, then went to hit it again because my car would not leave (lag), but I hit free run button. The game process my first request pulling me back to the lobby, the host changed the settings, then the game processed my second request putting me back on the track. I've never been able to repeat it but I know it is possible to have an illegal car on track (I was 50PP over).

What if something like this happens: Room is set to state 1. Room settings change to state 2 and for some reason a key variable does not get transmitted to some players. Some players start with state 1 and some with state 2. Seems reasonable yet I cant think of a room setting that could account for the loss of grip... Grip reduction does not affect lap times right?

SRF, grip reduction, boost all affect lap times I would set all these to off they are just clutter

 

[CSLACR]

Saw people talking about missing sounds.. That's been a problem since 1.06.
(Just for the record)

Feb 24 2011, 11:00 PM :
https://www.gtplanet.net/forum/showthread.php?p=4919947#post4919947


May 18 2011, 5:33 PM
https://www.gtplanet.net/forum/showthread.php?p=5331039#post5331039

I know I heard of it before, but I never encountered it, nor did anyone in any room I was in until 2.02, for whatever that might mean.
It was present again tonight, but only 2 people in the room had it.

SRF, grip reduction, boost all affect lap times I would set all these to off they are just clutter

Grip reduction doesn't affect my lap times.
I've heard a few theories on what all grip real/low affects, but for me, it only takes effect when wheels leave paved surface.

 

[xSNAKEx]

I know I heard of it before, but I never encountered it, nor did anyone in any room I was in until 2.02, for whatever that might mean.
It was present again tonight, but only 2 people in the room had it.

Grip reduction doesn't affect my lap times.
I've heard a few theories on what all grip real/low affects, but for me, it only takes effect when wheels leave paved surface.

Exactly so when you leave the paved surface it does affect you thus affects lap times.

On some tracks all my fastest laps involve leaving the paved surface

 

[jon-]

Exactly so when you leave the paved surface it does affect you thus affects lap times.

On some tracks all my fastest laps involve leaving the paved surface

We first noticed this on SSR5 where real / low grip settings have no effect on your laptime.

 

[Dave_Panic]

I'm not sure what any of that has to do with anything, but it seems to contradict itself.
I do know we have this awesome little orange light that tells us how active the HDD is, and if that's not solid orange, it's not going 100%.
Installation of GT5 data is a good time to see it at solid orange. Fan runs more, system gets hotter.
I sure as hell don't want to try keeping that light lit like that for hours on end, take that for what you want.

Just correcting some technical misinformation. There's nothing in my post that contradicts itself; HDD access is not equal to CPU, GPU or Memory usage, as I'm sure you're aware.

Of course it gets hotter from constant use but only up to a point, provided the cooling system is man enough for the job, which you would hope it would be or Sony will be in trouble pretty soon.

 

[kilttraeger]

Grip reduction doesn't affect my lap times.
I've heard a few theories on what all grip real/low affects, but for me, it only takes effect when wheels leave paved surface.

It's not that important regarding the current (and very useful) discussion, but grip reduction has an influence on handling and maybe on lap times. I mostly race in private lobbies; if one of the hosts set grip reduction to low, I can sense it through a change in behaviour of the car. Handling becomes more tight, showing less oversteer / more understeer (on a dry track).

 

[HydroFIN]

We had a race yesterday. I get 6-7 extra seconds between the qualification and the race. It was a real WTF -moment. Three guys left the room, and my lap times got better. Immediately I had five seconds better lap times!! 💡
I think that this may be the problem. There was 15 racers when the race started.

Sorry for my bad english.

 

[chuyler1]

Had major issues last night reported by everyone with a mic. As soon as the car count went over 12, people had to quit before the remaining drivers could race properly. The issue was worse if drivers were near you, but some saw choppiness and lack of grip/control while driving solo.

My lobbies will be restricted to 12 from now on...and I'll see if it happens again.

 

[carracerptp]

EDIT: something when wrong.

I'm thinking a number lower than 12 in the room is save for lag free rooms if everyone has a good connection.
Lower than 10 if you have connection worries.

I'm going to try and find out how boost works. What does it take to "activate it" on your car, in a race with it on.

 

[dudeondacouch]

Grip reduction doesn't affect my lap times.
I've heard a few theories on what all grip real/low affects, but for me, it only takes effect when wheels leave paved surface.

It also (greatly) affects grip when the track is wet.

With it set to "low" you can drive on a wet course with racing slicks, but with it set to "real" this becomes impossible.

As far as the regular dry track surface goes, no, it should not affect laptimes.

 

[jon-]

As we keep finding out, "should not" and "does not" are 2 VERY different things in the world of Gran Turismo

Anyone for Forza?

 

[dudeondacouch]

Well, by "should not" I mean "does not unless you either race differently or go off the track like a no-driving hack."

 

[windmill]

I think we are on the wrong way .

With some dutch guys we each called sony nl helpdesk and urged them to make separated case nr's and send them to the UK .

Sony NL says the problems are new and they never heard of it .

I advice everyone make a summary of problems u got and call your local Sony Helpdesk that way they get lots of complaints .

We have got an official case nr and they have to reply them if not we call Sony again in a few days and ask them how things are

 

[NCRthree]

Had major issues last night reported by everyone with a mic. As soon as the car count went over 12, people had to quit before the remaining drivers could race properly. The issue was worse if drivers were near you, but some saw choppiness and lack of grip/control while driving solo.

My lobbies will be restricted to 12 from now on...and I'll see if it happens again.

We've been having all this same crap for the last several months in the AIWJT mixed class series and Super Tuner Chmpionship Series. Everyone in the group has a Ping of B and Very High connection status. We've had to drop from 16-12 racers. I've personally experienced the lap time/ grip weirdness.

Last night we ran 28 laps @Spa in an Open Lobby-Fixed Host no mics. 12 racers. Grip real, no boost, low draft as usual. This race was run w/o any Visual Damage or Damage however. NO issues that I know of. Just makes sense to lessen the amount of data being transferred. I don't know if the lack of damage or fixed host helped or it was just a lucky night.

FWIW, even before the physics update I couldn't have a consistent race in a pub lobby if the count was over 12. That's the main reason I've pretty much stopped racing in pubs.

 

[chuyler1]

Any thoughts about changing the lobby bandwidth settings? I usually run it at the recommended "high" setting for game and "low" setting for audio. I wonder if running it on "low" for game would improve the playability with 16 drivers.

 

[NCRthree]

I don't host our races. I'm thinking race quality is not on high but I'll try to confirm.

 

[windmill]

Any thoughts about changing the lobby bandwidth settings? I usually run it at the recommended "high" setting for game and "low" setting for audio. I wonder if running it on "low" for game would improve the playability with 16 drivers.

I did try all different combinations of those settings nothing makes a difference for us

Also set my router to ip number , dmz zone no results

 

[chuyler1]

Ok, good....just making sure someone tried that.

Also, if this were a PS3 performance / graphics related bug, and not a network bug, wouldn't the issues worsen with weather enabled? We all have trouble driving in the rain, but I wonder if 16 cars with no rain is equivalent to 12 cars with rain...when it comes to cpu cycles and rendering graphics.

 

[windmill]

Ok, good....just making sure someone tried that.

Also, if this were a PS3 performance / graphics related bug, and not a network bug, wouldn't the issues worsen with weather enabled? We all have trouble driving in the rain, but I wonder if 16 cars with no rain is equivalent to 12 cars with rain...when it comes to cpu cycles and rendering graphics.

It can be thousand different reasons we can all keep guesing them for years to come .

My opinion lets do something now and stop talking https://www.gtplanet.net/forum/showthread.php?t=238610

 

[nasanu]

Hint: The system idle process is not part of of any game

I have made a couple games myself (PC, Mac ported, C++, some ASM).
Not an expert on the PS3, but if every game ran 100%, it would die in about a day (maybe a week, not sure how good the cooling is).

With every one of the many times I have maxed out a CPU for more than 3+ hours (fluid, light simulations), it needed a cooling system or a break.

With the PS3 not having a liquid cooling system, I can safely say that if it ran 100% for more than 8+ hours or more, it would die.

Considering some people run it for 24 hours in offline A-Spec, it would be safe to say that any game sitting at 100% usage would kill it.

I notice you say resources. I'm assuming that includes RAM and CPU.

There is no need for a game to use every byte of memory. It completely depends on the game. Some games use more, some games use less - that's the way it works.

I can't think of any title (PC) that treats memory like PHP does, and takes it all for itself. If you could name one, i would be interested in looking into it.

Whatever OS that the PS3 uses for the XMB and all couldn't let the game use all the resources. Could you imagine how troublesome that would be, listening to the PS3 thrash as it tries to load the Quit Game menu?

I always wanted to prove a point for no good reason

Well you failed.

http://www.google.com/url?sa=t&rct=...B3xVOFYlOQBLEjbvA&sig2=TAvvo7S4mxVHi8Az15v4pw

Damn google stuffing up links, anyway that should still work. Its an SCE power point on using their performance analyser. The PA is a visual tool developed by Sony to allow developers to see visually the load on each bus and chip. The basic way to use it is to see where the lines are not reaching near the top and move code around to max everything out. Take a look at the 'high performance' slide and note that the CPU chart is almost all green. Obviously they have tried to max the CPU 100% of the time. Also take a look at the section on VU usage tips. Note the line 'should run almost 100% of the time'. The whole thing is dedicated to pushing everything harder, nowhere does it even hint at not using everything to maximum. Oh and it also says the the CPU is the most common bottleneck in most games. How could it possibly be a bottleneck if it is not getting maxed?

A console is not your PC, it is designed from the very start to run maxed. If it had headroom then why do we get slowdown?

Games programers do indeed try to max everything out. This is VERY hard to achieve though, that is what some games look like GT5 and some like Ferrari Challenge. Show me something official telling developers not to use the power of a console and perhaps I (and the industry) will change my mind. All I can find is platform holders tying to help developers exploit more and more of the hardware.

 

[BWX]

As I see it, there are thing it cant be and things it can be. 'The PS3 can't cope' angle is IMO a dead end. EVERY SINGLE GAME USES 100% OF THE PS3'S ABILITY.

So if someone designed a game like ''Pong''.. with black and white 2D GFX, one dot for the ball, and two other lines as the paddles, for the PS3, this game would tax the PS3 as much as GT5 running SPA in the rain online with 16 players all running premium cars? Are you serious or just serious about trolling for arguments?

Well you failed.

No you failed. The theoretical example I just gave disproves what you are claiming. You get into trouble when you make huge claims you cannot possibly prove and state them as fact with terms like ''always''. It's complete BS. Just stop man.

EDIT- Using ''ignore'' feature now so I won't be responding.

 

[brettcee]

EDIT- Using ''ignore'' feature now so I won't be responding.

I've got him "ignored" , now if everybody would just stop quoting him, I wont have to read his posts at all

 

[cicua]

Any thoughts about changing the lobby bandwidth settings? I usually run it at the recommended "high" setting for game and "low" setting for audio. I wonder if running it on "low" for game would improve the playability with 16 drivers.

I did try all different combinations of those settings nothing makes a difference for us

+1 for real info. I have been wondering about this too.

Also, if this were a PS3 performance / graphics related bug, and not a network bug, wouldn't the issues worsen with weather enabled? We all have trouble driving in the rain, but I wonder if 16 cars with no rain is equivalent to 12 cars with rain...when it comes to cpu cycles and rendering graphics.

Good point, weather should be rather taxing just like more cars. It would be interesting to test, it could prove or disprove if this is CPU related.

It could be tested without rain and on dry track so lap times can be consistent and predictable. The dynamic lighting is still processing heavy.

 

[nasanu]

So if someone designed a game like ''Pong''.. with black and white 2D GFX, one dot for the ball, and two other lines as the paddles, for the PS3, this game would tax the PS3 as much as GT5 running SPA in the rain online with 16 players all running premium cars? Are you serious or just serious about trolling for arguments?



No you failed. The theoretical example I just gave disproves what you are claiming. You get into trouble when you make huge claims you cannot possibly prove and state them as fact with terms like ''always''. It's complete BS. Just stop man.

EDIT- Using ''ignore'' feature now so I won't be responding.

Don't be so childish. If not from me then take the word of Sony Computer Entertainment who made that slide I linked too. But anyway keep claiming that games do not use all available resources while also speculating that GT5s issues are because the resources are being taxed too much.

 

[esh]

Totally speculating but I don't think framerate dips in this case are hardware related. More to do with people in room not providing the correct/appropriate amount of information or there is an averaging of all connections to induce this lag on purpose as to simulate an equal playing field. That would be a netcode issue.

Not sure if many people are familiar with Quakelive (or Quake III) but Quakelive uses an "un-lagged" technology in it's netcode as to average latency between clients within a certain range as to not give an advantage to low ping players versus high ping players in general but more specifically when using hitscan weapons:

Hitscan weapon is often used colloquially to refer to a projectile weapon which uses unmodified hitscan information to dictate whether or not it has hit its target — deploying the weapon calls the hitscan function, and if an object is detected in the path of the projectile, a hit is registered

 

[CSLACR]

Ok, good....just making sure someone tried that.

Also, if this were a PS3 performance / graphics related bug, and not a network bug, wouldn't the issues worsen with weather enabled? We all have trouble driving in the rain, but I wonder if 16 cars with no rain is equivalent to 12 cars with rain...when it comes to cpu cycles and rendering graphics.

Yes, it happened to me all race long a few weeks back at Le Sarthe, weather and time change. Room started at 12 or 14, I can't remember, but even all the way down to 10 it was there.
This was before I knew exactly what it was, and just believed it was solely poor frame rate.

It didn't rain, but it definitely seems just having the day/night tracks up makes a difference. there are the ones with the most detail, so it makes sense I guess.

 

[nasanu]

I think this effected me today, or something similar. In my case from the start the frame rate was perhaps about 20fps, very noticeably slow. Was not just me either, everyone in the room started complaining all at once.

It is very strange. You don't lose grip, the physics don't change but on occasion inputs seem delayed. It does cause you to run wide as if you have less grip but you can adjust to it, steer and brake a little sooner etc.

The race was at the ring 24hr, we had just finished race there and started a new one. However in between races the host messed quite a bit with the room settings resetting the track about 5 times and also two British players joined in (room: about 8 usa, 1 aus, 2 british). I highly suspect something to do with the British players over room settings though, the way the frame rate seemed to lag then jump ahead, very much like network activity.

Maybe this should be split into two separate problems because what I experienced (and others in this thread) and what speedy described are different. Speedy said it was just physics that changed and everything else was normal but the bug that I experienced was obvious. Though some people can seem to tell when frame rates drop and the missing frame data would account for a faster replay... They might be the same thing.