FM Vs GT - Discussion Thread (read the first post before you post)

  • Thread starter Scaff
  • 8,743 comments
  • 531,689 views
The ARB settings worked a little too good for my liking. No matter the chassis, a setting of 1/40 would almost always produce the best lap times. I could count the exceptions from this rule on maybe 2 fingers. And that's a bit too good to be universally true without any drawbacks. Also the gearing should raise an eyebrow. Myself I'm quite lazy shifting gears, even more so with my G25 and clutch, but a single gear on Mugello? Thats a bit optimistic even for my liking. Just watched Gan-San doing the Nirdschleife in a Scooby in an old Best Motoring episode and he was changing gears so often I wondered if he simply did this to make a statement.

Very soft springs and especially very very soft compression (bump) settings along with extremely hard extension (rebound) settings and a low ride height should result in a car that's riding on the shock stops a lot. Or is constantly bottoming out. Please do correct me if I'm wrong. One downside was also poor braking performance, but this was compensated by the ABS.

Forza 3 dealt with this, as far as I could tell, by adding an impact penalty on the car underside. It did encourage folks to mind their settings as to avoid invalidating lap times, but was there a cure for the problem or only the symptoms?

Very good garages I remember from the top of my head were VVV (very fast and aggressive), WWR and SloPoke. Don't know if anybody is in touch with them, or if they are even still around.

Now don't get me wrong, I was fascinated by the effects of minor changes to the suspension. They reflected rather well on lap times. Instantly in fact. Quite enjoyable. It's just the realism (used in a casual sense, it's still a console game after all) I felt was hurt.

Take your time though, and I'm really open minded to the results.

OK - I've now spent some time with the above and broken it down into three areas:

  1. Anti Roll Bar settings
  2. Soft front springs combined with min bound and max rebound
  3. Single gear around a track

The car used was a Integra Type R (DC5) pretty much maxed out at 650ish PI (S class) and at B class, I will post up details of the exact upgrades and settings when I get the chance (posting this at lunch from work).

Anti Roll Bars
I'm looking at this one with an eye both to the ITR and other tunes I have put together and a setting of 1/40 certainly doesn't act as a golden bullet, certainly once you get out of FWD tunes. Even looking at the two different ITR set-ups the S class one I found better with a higher ARB setting that the B-class tune and neither warranted going as low as 1/40.


Soft front springs combined with min bound and max rebound
I set the front springs as soft as the would go along with the front bound and then maxed out the rebound and the only way to describe the results is just plain nasty.

Straight line throttle application - It wasn't too bad at B-class, however the S-class tune had problems maintaining contact with the road.

Braking - Nasty on both set-ups, with the front suspension hitting the bump stops almost immediately which resulted in a massive loss of traction occurring, telemetry shows this very clearly. The firm rebound also showed in a massive miss-match as to what the suspension wanted to do, resulting in longer braking distances (even with ABS) and a very unsettled car that was not easy to control. Shows internally as the front end bobbing up and down under braking.

Rumble strips - Unsettled the car quite significantly particularly if combined with braking

Cornering - Understeer city. The car just slams to the bump stops and as the spring rates effectively go through the roof it just looses all front traction, meaning you have to wind back on the steering and throttle to get grip back (which happens suddenly as the firm rebound delays it occurring) and then as soon as you steer or apply throttle again your back to square one.


Single gear around a track
Simply doesn't work effectively at all, you step so far outside the power and torque curves that it doesn't work as a method of lapping quickly.


Results
None of these areas acted as a glitch on the car in question at all, in either of the tunes or on any of the tracks I used (Mugello and Tskuba). I compared lap times from the these set-ups to a quick three lap burst from my own set-ups and in the B-class tune it was at least 4-5 seconds a lap slower despite putting in significantly more laps in the test set-up. The S class set-up showed an even bigger difference at around 10 - 11 seconds per lap and it was even more difficult to get consistent laps out of the car.

End result is that you do not want to set your cars up like this at all, ever, the springs didn't control the front end of the car at all, the soft bound setting allows the suspension to hit the bump stops when almost any load transfers forward (which means you go from the softest possible spring rates to a stupidly high value as the bump stops are hit) and the firm re-bound then unloads it very abruptly. Not a nice combination of factors at all.

What I also hope to get the time to do is expand this to GT5 and first see how the values we have to play with compare for suspension settings and then see how GT5 reacts under similar set-ups.



Scaff
 
Last edited:
So the glitch seems to be not present.

Try locking the Accel dif on FWD cars for better traction when coming out of corners.
A friend of mine on Xbox live who ran in a weekly series as a team mate of mine gave that tip.
He used to do a lot of grass racing in RWD cars. Everyone had there diffs locked for better grip.
Not sure why that worked myself. But it worked in real life.

The FWD locked diff Seems only marginally better to me but it helped the front end stay in the direction you wanted to go coming out of a corner. Certainly not on rails glitchy though.
 
OK - I've now spent some time with the above and broken it down into three areas:

  1. Anti Roll Bar settings
  2. Soft front springs combined with min bound and max rebound
  3. Single gear around a track

The car used was a Integra Type R (DC5) pretty much maxed out at 650ish PI (S class) and at B class, I will post up details of the exact upgrades and settings when I get the chance (posting this at lunch from work).

End result is that you do not want to set your cars up like this at all, ever, the springs didn't control the front end of the car at all, the soft bound setting allows the suspension to hit the bump stops when almost any load transfers forward (which means you go from the softest possible spring rates to a stupidly high value as the bump stops are hit) and the firm re-bound then unloads it very abruptly. Not a nice combination of factors at all.

What I also hope to get the time to do is expand this to GT5 and first see how the values we have to play with compare for suspension settings and then see how GT5 reacts under similar set-ups.



Scaff

Oh bugger. Well Scaff you certainly went through a lot of trouble there. Thanks for that.

Good choice of car btw. Now I should try it myself shouldn't I.

I'm very surprised to say the least. Got to put the kettle on and take in all the facts. Hm.
 
OK so I've now recreated the 'glitch' tune from FM2, which is of course no longer a glitch in FM4, in GT5 to see what the results would be like.

I used the FM4 B class tune as it was not possible to match all the upgrades in GT5 from the S class tune, I also found that running Racing tyres in GT5 pretty much removes a lot of the detail from a tune.


Soft front springs combined with min bound and max rebound
I use the values from FM4 to recreate this in GT5 - details of both below.

Straight line throttle application - Minor issues with laying down power but nothing major. However the front of the car could be seen to bob indicating the dampers were not able to fully control the spring oscillation.

Braking - No issues at all, car remained remarkable stable under braking and I didn't notice any issues with the car hitting the bump stops.

Rumble strips - Noticeable oscillation of the front end that continued for a short while after the rumble strip was finished.

Cornering - The car did understeer quite heavily, however the suspension did not appear to hit the bump stops and as such it didn't have such a marked effect as the FM4 set-up.


Results
The end result is that again this is not a good set-up at all, however it does lack that definitive hitting the bump-stops and the issues associated with doing so. Its does still suffer from a soft front end not being corrcetly controlled and not being 'tied down' to the road as it should be.

I do however think its important to take a look at the settings used in both:

BHP - FM4 290 : GT5 289
Torque - FM4 203 ft/lbs : GT5 190 ft/lbs
Weight - FM4 2,692 lbs : GT5 1,221 kgs
Lat G 0.95 - FM4 Street tyres : GT5 CS tyres
Ballast adjusted in GT5 to ensure 61% front weight distribution

FM4

ARB - Front 1/40 : Rear 24/40
Spring Rate - Front 118.9 lb/in : Rear 535.9 lb/in
Ride Height - Front 5.7" : Rear 6.0"
Damper Rebound - Front 12/12 : Rear 7.8/12
Damper Bound - Front 1/12 : Rear 6/12


GT5

ARB - Front 1/7 : Rear 4/7
Spring Rate - Front 2.5 kgf/mm : Rear 9.6 kgf/mm*
Ride Height - Front -3 mm : Rear +6 mm
Damper Rebound - Front 10/10 : Rear 7/10
Damper Bound - Front 1/10 : Rear 5/10

* The rate used at he front should have been 2.12 kgf/mm, however it is not possible to set this on GT5


As you can see its not actually possible to set the spring rates as soft in GT5 as it is in FM4, I also have some concerns that the more limited range of settings values in GT5 may also make direct comparisons difficult. While it is possible that these factors resulted in the car not hitting the bump stops under braking and cornering it does also remain possible that GT5 simply doesn't allow it or doesn't model it.

Unfortunately it does mean that at present I can't narrow it down any more than that, I can however clearly say that this kind of tune doesn't act as a tuning glitch in either GT5 or FM4. What I intend to try next is two more steps:

The first being to match the GT5 spring rate in FM4 to see if that still hits the bump-stops, which will at least make that area a little clearer. The second is to then take a tune that I know works well in FM4 and see how it does when I recreate it within GT5, more than anything just to see how to two sets of tuning values compare and if the wider range in FM4 does actually help or just confuse.


Regards

Scaff
 
OK so I've now recreated the 'glitch' tune from FM2, which is of course no longer a glitch in FM4, in GT5 to see what the results would be like.

I used the FM4 B class tune as it was not possible to match all the upgrades in GT5 from the S class tune, I also found that running Racing tyres in GT5 pretty much removes a lot of the detail from a tune.


Soft front springs combined with min bound and max rebound
I use the values from FM4 to recreate this in GT5 - details of both below.


I do however think its important to take a look at the settings used in both:

BHP - FM4 290 : GT5 289
Torque - FM4 203 ft/lbs : GT5 190 ft/lbs
Weight - FM4 2,692 lbs : GT5 1,221 kgs
Lat G 0.95 - FM4 Street tyres : GT5 CS tyres
Ballast adjusted in GT5 to ensure 61% front weight distribution

FM4

ARB - Front 1/40 : Rear 24/40
Spring Rate - Front 118.9 lb/in : Rear 535.9 lb/in
Ride Height - Front 5.7" : Rear 6.0"
Damper Rebound - Front 12/12 : Rear 7.8/12
Damper Bound - Front 1/12 : Rear 6/12

Regards

Scaff

Ah, ok, I see what you tried there.

The tyre compound should have been the stickiest around, in FM2 it was the Avon Racing tyres.

I guess the class to go for should be A600 now on Mugello.

So for a FWD car, first set the ARB front to 1 and back to 40, and tune the rest around these settings. There were a few minor changes, but even a setting of 3 front and 37 back were not considered competitive back then. So I think that's were tune differs quite a lot. The challenge from there would to design a car that's actually driveable, wants not to eat his own face and doesn't lift off the curbs by just looking at them.

Next step would be a very hard setting, like you did, for the rebound of the shocks. Then use very soft compression. Start at the extremes, but don't limit yourself to them. Last the springs should be used to give the tune some sort of balance. What can't be achieved through this, the caster setting come into play. Mostly there were aero tunes, but that's what I set last. Camber looked pretty normal, so was ride height. LSDs were of minor concern sometimes, though very low settings were used like in GT5.

Now it would be interesting to see what times you get if you once set a time with a tune you'll think of as "proper", and one you started building around the ARB and rebound/compression settings I described earlier on. Don't try to get any ideas from the telemetry though, simply by looking at corner entry/exit speeds and maybe straight line speed.

If the gearing still applies this could be found out also by looking at LB replays, also I think telemetry should be still available, so that would be a good thing to go by whether the LB tunes produce good telemetry data (heating of tyres, camber in crucial corners).

Now, I don't run competitive tunes in GT5, because (a) I don't do online anymore and (b) I'm fed up with all the "creative use of what's possible in the game." Power limiter is effectively a glitch, as it doesn't affect torque. Or at least that's what I make of it. Same goes for ballast. Now I don't mind optimizing a car's natural balanced, but once you go there it's only a little step to bringing a car closer to a theoretical perfect ride. Which is very hard IRL because the chassis as some limitations one can't overcome. Well, that's to say if one doesn't actually build a full blown race special which reassembles the appearance of the road car only.

I'm really annoyed I can't do this myself because of RL issues, but come August I'll have the chance to take a look into things. Even if it would be out of sheer curiosity. Anyway, I expected to get good results by bogus setups much easier, so that's good new for me so far.

Won't mind if you can't be bothered Scaff, you've invested enough of your time already.
 
Ah, ok, I see what you tried there.

The tyre compound should have been the stickiest around, in FM2 it was the Avon Racing tyres.

I guess the class to go for should be A600 now on Mugello.

So for a FWD car, first set the ARB front to 1 and back to 40, and tune the rest around these settings. There were a few minor changes, but even a setting of 3 front and 37 back were not considered competitive back then. So I think that's were tune differs quite a lot. The challenge from there would to design a car that's actually driveable, wants not to eat his own face and doesn't lift off the curbs by just looking at them.

Next step would be a very hard setting, like you did, for the rebound of the shocks. Then use very soft compression. Start at the extremes, but don't limit yourself to them. Last the springs should be used to give the tune some sort of balance. What can't be achieved through this, the caster setting come into play. Mostly there were aero tunes, but that's what I set last. Camber looked pretty normal, so was ride height. LSDs were of minor concern sometimes, though very low settings were used like in GT5.

Now it would be interesting to see what times you get if you once set a time with a tune you'll think of as "proper", and one you started building around the ARB and rebound/compression settings I described earlier on. Don't try to get any ideas from the telemetry though, simply by looking at corner entry/exit speeds and maybe straight line speed.

If the gearing still applies this could be found out also by looking at LB replays, also I think telemetry should be still available, so that would be a good thing to go by whether the LB tunes produce good telemetry data (heating of tyres, camber in crucial corners).

Now, I don't run competitive tunes in GT5, because (a) I don't do online anymore and (b) I'm fed up with all the "creative use of what's possible in the game." Power limiter is effectively a glitch, as it doesn't affect torque. Or at least that's what I make of it. Same goes for ballast. Now I don't mind optimizing a car's natural balanced, but once you go there it's only a little step to bringing a car closer to a theoretical perfect ride. Which is very hard IRL because the chassis as some limitations one can't overcome. Well, that's to say if one doesn't actually build a full blown race special which reassembles the appearance of the road car only.

I'm really annoyed I can't do this myself because of RL issues, but come August I'll have the chance to take a look into things. Even if it would be out of sheer curiosity. Anyway, I expected to get good results by bogus setups much easier, so that's good new for me so far.

Won't mind if you can't be bothered Scaff, you've invested enough of your time already.

I've run it now in a wide range of configurations, to be honest the use of stickier tyres does nothing but make the issue of uncontrolled spring oscillation worse. The additional grip available just means the transfer of load forward is happening slightly quicker and the bump stops get hit a bit harder and faster.

Setting the rear ARB to 40 does slightly reduce the more severe end of the understeer, but only once you are into the corner (not a massive surprise that), as such it tends to go understeer on entry moving to massive understeer, slight reduction in understeer and then back to massive understeer as you try and apply power out of the corner.

The gearing is a dead end now, you simply can't make any effective laps using a single gear, so that I have thrown in the bin as it was a good 4-5 seconds a lap slower over a sensible gearing set-up.

In particular with the S class configuration this kind of set-up is just not effective at all, now at this kind of level a FWD car is never going to be easy to get the power down with, but in comparison to my own tune I was looking at a difference of around 10 seconds a lap with the single gear set-up and around 5 seconds per lap without it.

To be honest no matter which way I try this its not a glitch of any kind in FM4, nor is it even close to an effective way of setting up a FWD car, regardless of the power or tyres fitted. Which I must say is to a degree as things should be for a tune of this level of extremes.

The only part of it that does have some real world bearing in the very low front ARB, however that is normally in shorter wheelbase cars such as the Clio Cup car, in those I have even heard of the front ARB being removed. However that would still be done with front spring rates and dampers set to correctly balance and support the car.


Scaff
 
Scaff
I've run it now in a wide range of configurations

(...)

To be honest no matter which way I try this its not a glitch of any kind in FM4, nor is it even close to an effective way of setting up a FWD car, regardless of the power or tyres fitted. Which I must say is to a degree as things should be for a tune of this level of extremes.


Scaff

Thats cool then I suppose and a vast step in a good direction. Thanks for making this clear and looking into the matter.
 
Erm... So, when there's no torgue steer, you're willing to assume that that's down to the track? To all tracks, rather than the physics engine? :odd:
 
Sorry for replying so late. I've been blitzed with a large load of work, so this led to me doing a mini test on torque steer.

I started with the Silverado SS, but then found out that it uses a trans-axel:tdown:. I then shifted to the Chevelle SS. First test was to see if there's any torque steer influence with the tires under different loads. Results: Torque steer in the direction of the tire with the lesser load. Second test: see if there's any torque steer influence at launch with tires of different ware amounts. Results: No torque steer(torque steer with tires at different loads), but this test needs to be executed again in the future for me to validate it. End... Yes.. you know the obvious (no torque steer), but there needs to be a explanation to whether it is a physics problem, or a problem due to bad development of the tracks. Will do a full in-dept test in two weeks.

One thing I love about GT5... It uses a real time simulation of mesh deformation, while Forza sticks with the same old damage models (pre made models for damage). With more development PD can make its damage system be WAY more flexible(more realistic damage awesomeness).

The only track in the entire world that can influence torque steer (to my knowledge) is Brands Hatch with its cambered start line. Its entirely down to the physics model.

Real time simulation of mesh deformation? It looks crap, FM4s may be pre made but it looks hell of a lot better than GT5s poor excuse for damage.
 
MildAshers
Real time simulation of mesh deformation? It looks crap, FM4s may be pre made but it looks hell of a lot better than GT5s poor excuse for damage.

Personally I couldn't care less for visual damage, but as far as console games go, Grid is yet to be matched. So none of the two really cut it visually. The few models with full damage modelling look good in GT though. Even in the rain or at night :P
 
Agreed on Forza's damage looking a lot better than what is seen in GT5. I don't think PD concentrated enough on textural damage (i.e. scratches and paint loss) when compared to Forza. Another thing about the deformation in GT5 is it applies to the whole car rather than only things that would deform, for example plastic bumper covers and glass will usually crack or shatter long instead of bending to the weird shapes as seen in GT5. This isn't just with GT5, but I've noticed games that don't use pre-rendered damage and have a 'dynamic' form of deformation can end up with some pretty unrealistic damage. If you know what you're doing, you can make a 'banana car' in Grand Theft Auto IV, and Viper Racing from the late '90s goes as far as to allowing you to compress your entire car into the size of a soccer ball if you try hard enough.

That being said, I think few video games get damage completely right. The first is few video games actually allow you to "crash out" a car these days, in that you wreck a car to the point it just does not run any more. Forza comes close when the damage is set to 'simulation', but even then the car will still slowly sputter along. The second is the accurate modeling of the actual structure of the car crumpling in collisions, ironically one particular arcade franchise Burnout actually gets this right at times. However, for the most part I've noticed with a lot of games the car will never deform beyond its basic original proportions, which is true with GT5.
 
Personally I couldn't care less for visual damage, but as far as console games go, Grid is yet to be matched. So none of the two really cut it visually. The few models with full damage modelling look good in GT though. Even in the rain or at night :P

Any Race Driver game is yet to be touched! They are each amazing in their damage rendering.
 
The only track in the entire world that can influence torque steer (to my knowledge) is Brands Hatch with its cambered start line. Its entirely down to the physics model.

Real time simulation of mesh deformation? It looks crap, FM4s may be pre made but it looks hell of a lot better than GT5s poor excuse for damage.

Erm... So, when there's no torgue steer, you're willing to assume that that's down to the track? To all tracks, rather than the physics engine? :odd:

Sorry. I had the wrong view on torque steer, forget I ever even bought it up. I wasn't critiquing on which title has a better damage system. Also it is pretty evident that GT5's damage looks like crap, hence why I said they need more time to develop on it.
 
In theory such a system of procedurally generated damage could produce better results than pre-rendered but I doubt it's possible during this generation of consoles.
 
Real time damage is certainly possible, it is however a logistical nightmare for a game of this size. The programming would have to take into account the material, the material properties, the underlying body structure, and how each panel would affect the next. This is why games such as GTA use the more basic dynamic system that allows a person to almost crumple the car into a cube. Realistic dynamic systems are used by the car manufacturers, however since they are working on one or two cars at a time per team, they can actually produce good results unlike PD trying to tackle this for every different car in the game, especially for older models.
 
FyreandIce
<a href="http://www.youtube.com/watch?v=nDKDWae-7UU">YouTube Link</a>

Nice little video just showing how crazy advanced Forzas tyre model is.

No offence but T10 has been boasting how good their tyre model was supposed to be since FM2. I have no doubt they pushed things a good deal since FM 2/3, in fact Scaff has shown they actually improved on a lot of things.

But I won't buy their marketing talk just like that.

Tim Schrick did a RL vs. Forza comparison. He was pretty pleased with the game besides the M5 having too little grip on corner exit. And he was 3 seconds down a lap on Nürburgring GP compared to RL. Quite a nice video if you happen to understand German.

Edit forgot the link

http://www.youtube.com/watch?v=wcwstdwjTE8&feature=youtube_gdata_player
 
HKS Racer posted an interview with Shrick in November.


However I thought this was Forza vs GT, not Forza versus one driver's impression of real life.

Is anyone buying their marketing talk just like that? I thought what Scaff did was to post demonstrable physical improvements with the tyre response that couldn't be reproduced in GT5.
 
I played both games and think that both games have pro's and con's...!

But for me there is be one aspect about FM4 that GT5 just can't deliver.

578249_407193342636912_185465731476342_1342204_1372969628_n.jpg


There is no other Console-Racing-Sim out there that coul'd get the right feeling for a Porsche. I think only Forza and GT are capable of doing so.

So I'm very exited about the upcoming Porsche DLC.

For me personaly Forza has got the better carselection compared to GT5. But now Forza is just heaven for every Porschelover.

And with the great sound of Forza driving those Beasts will be a pure orgy.
 
I've only played Forza, what, 3 or 4 times, but every time I've played it I've found the physics unsatisfying. The times I've played GT5 I've found the physics very satisfying and realistic. I've barely played Forza or GT5 so I can't talk. I'd say Forza and GT4 are pretty much tied from what I've seen of Forza though.
 
Not sure if this is mentioned but I had a look at a replay on eiger with my mach 1 mustang and when it landed after that jump on the first straight it didnt show the correct bounce that would normally follow a jump like that.

Im going to try it with a jump in forza 4 see what happens.
 
What's your basis of whether the physics are satisfying or not? Just curious.

I define satisfying physics as physics that give you a very direct, instinctive feedback. Most racing games don't have that, but GT always has. You can just feel what the car is doing, even through a controller. Race 07 is the other good example I can think of of a game that has physics like that.
 
I define satisfying physics as physics that give you a very direct, instinctive feedback. Most racing games don't have that, but GT always has. You can just feel what the car is doing, even through a controller. Race 07 is the other good example I can think of of a game that has physics like that.

If you were speaking of using a wheel on both games, I wouldn't really contest that, but I find when stuck with a pad, FM4 definitely excels. I actually feel when my cars are overstepping the boundaries of their grip, on either end, whereas GT5 gives very little feedback. I want to say it wasn't always this mute though; I think one of the more recent updates did it :(
 
SlipZtrEm
If you were speaking of using a wheel on both games, I wouldn't really contest that, but I find when stuck with a pad, FM4 definitely excels. I actually feel when my cars are overstepping the boundaries of their grip, on either end, whereas GT5 gives very little feedback. I want to say it wasn't always this mute though; I think one of the more recent updates did it :(

Put force feedback to 10 and the wheel on simulation...
 
I try to play FM4 but the wheel feedback is so damn weird. I just dont get how the wheel was made for FM but it feels better with GT5. The only thing I like FM4 does is simulate the tire grip sensation it is very good. Other than that GT5 feels better and natural to me. Also whats up with the engine rumble in FM4? GT5 has got great engine rumble after you turn off the ABS on the CSR. The wheel feels so dead while i play FM4 now I just don't get it. I love the customization but the driving is not doing it for me.
 
Back