Talk:Comparison of racing simulators

From Wikipedia, the free encyclopedia

Famicom style controller This article is within the scope of WikiProject Video games. For more information, visit the project page, where you can join the project and/or contribute to the discussion.
Start This article has been rated as Start-Class on the assessment scale.
Low This article is on a subject of Low priority within gaming for inclusion in Wikipedia 1.0.
VG To-do:

Contents

[edit] Evolution of the ISI game engine

In the main page, it was said that GTR, GTL, GTR2, and rFactor represent the evolution of the ISI game engine. But why did F1 Challenge 99-02 is not included? It's still being played and if http://forum.rscnet.org is right, it's still popular. and all three SimBin games are based on it (GTR, GTL, GTR2) compared to rFactor where ISI made a different game engine from scratch.

[edit] GTR Graphics: Specular Mapping

I highly doubt that GTR uses Specular Maps. As far as I know, you can only use specular maps when rendering using shaders, the fixed pipeline does not support specular maps. And since GTR doesn't use shaders, specular mapping doesn't seem to be possible...


[edit] GFX -> Shadows

I think it's a good idea to erase the current column "soft shadows" and replace it with "shadow mapping" and "light mapping", which will contain the certain shadow technique that is used. This makes sense because none of the games in the list supports soft shadows, sepcially not virtual grand prix 2 :)

As an example, LFS uses smoothed shadow maps, projected from a lower-poly object, and gouraud shading as lightmaps (not sure about this though though). LFS viewer uses stencil shadows (point light, "Depth-pass stencil shadow volume rendering"), without reducing the polygon count for the source mesh, and with merging the shadow volumes of different objects, resulting in one plain shadow rather than overlaying shadows with darker intersections.

Soft shadows info: There are various different approaches regarding the implementation of soft shadows, which will never reach the quality of real global illumination / raytracing based shadows which are caused by area lights, though. The term "soft shadows" is meant to be used only for shadows which have a danymic amount of smoothing, depending on the distance to the object, which causes the shadow. Shadow maps like in LFS are no "soft shadows", even though they look soft. Actually it's just basic shadow mapping, which is not as exact as for example the point light based shadows in Doom3. Some links: http://www.gamedev.net/reference/articles/article2193.asp http://dee.cz/fcss/

[edit] No Original Research!

I wish to remind everybody that it is official wikipedia policy to require that all facts mentioned be verifiable and NOT the result of original research by editors. Please read the policy page. If you do not agree with this policy, please, write about your findings or original analyses elsewhere. 68.165.188.60 01:33, 7 May 2006 (UTC)

Would a screenshot be classified as OR? IMO strict adherence to WP:OR hurts some articles, like this, because the sources ie the games are rarely objectively studied: The number and quality of available material is often dependent on marketing budget. Besides, anyone with an access to the game (or even the demo) can verify the content of this article - One could argue that in this case it's not original research but open research. - G3, 09:50, 27 August 2006 (UTC)
This should be a special case. I don't see how these can be verified any other way than just loading it up and trying it. --Darth Borehd 11:14, 8 November 2007 (UTC)

[edit] Live for speed Suspensions

Can samobody confirm that LFS models wishbone and MacPherson strut suspensions?

Maybe this one helps somewhat? Scawen talking about the double wishbone here: http://www.lfsforum.net/showthread.php?p=32566#post32566

Yes it does. Since it has double wishbone, then the wishbone component must be supported.--BMF81 10:40, 23 March 2006 (UTC)
Press shift+L in-game and you will see how suspension links work in realtime. Nitemare 22:25, 18 March 2007 (UTC)

[edit] GTR and GTL real tyres data

Do really GTR and GTL use data from real tyre manufacturers. Which ones?--BMF81 01:14, 23 March 2006 (UTC)


[edit] deformable tyres affectiong grip

What is meant with "deformable tyres affectiong grip"? How is it supposed to not affect grip? LFS tyres deformation certainly has an effect in driving (therefore in grip, i would guess), so i wonder why is it marked as "no"?

Also, RBR deformation most probably has an effect on grip, judging by some of its configuration files. However, it's also well known that RBR has a lot of unfinished features, therefore this wiki article could be right when stating "no". But can anyone actually confirm that? Unless it can be confirmed, i think the best would be to leave it as "unknown". --STenyaK 10:31, 23 March 2006 (UTC)

Isn't the LFS tyre deformation just in grahics and not in physics??--BMF81 10:36, 23 March 2006 (UTC)
judging from personal experience, i would definitely say yes. Ways to check it:
  • Stand still in a car wide tyres and low pressure. Move for your force feedback steering wheel. You'll feel the wheel "lagging", as if the tire was bending.
  • Or more easily: go to garage, choose lowest pressure possible, hit the "jump" or "drop" button (can't recall the name). See how the car rims effectively reach a lower point (closer to ground) than when sitting still. That's because of tire deformation. And if the whole car position is affected by that (rims obviously connected to car body), therefore car handling is affected too.--STenyaK 10:50, 23 March 2006 (UTC)

Isn't that more a tire pressure issue than dynamically deformable tire issue?

  • By "deformable" I understand a change in tire shape. By "dynamic" i understand that it changes during the simulation. In lfs, tyre pressure does change tire shape (tyre radius is changed, as desmonstrated above), and it does change during the simulation (you can check it yourself in the garage, which runs physics engine too, or in a regular race). So it's a "tire pressure issue" in part (collision forces and tire wear also count), but also a "dynamically deformable tire issue". STenyaK 19:21, 6 January 2007 (UTC)

[edit] active differentials

"A computer uses inputs from multiple sensors, including yaw rate, steering angle, and lateral acceleration and adjusts the distribution of torque to compensate for undesirable handling behaviors like understeer. Active differentials are common in the World Rally Championship, though they may be eliminated in coming years."

RBR certainly simulates this. Torque is electronically controlled while you drive depending on the situation. Knowing the heaps of posts at BHMS RBR forums regarding setups for active differentials, authored by both real rally drivers, gearheads and racesimmers (oh, i forgot: and the RBR physics coder himself too!), it's impossible that active differentials are not implemented in RBR. Therefore, unless proven otherwise (which would need some serious scrutining of reverse engineered code), i'll leave it as yes.

Supposing torsen diffs are not a subgroup of active diffs, or viceversa, torsen diffs are therefore not implemented in RBR, therefore i'll set it to no (RBR only models one type of diffs afaik). Feel free to correct me though (please discuss in this page).--STenyaK 11:19, 23 March 2006 (UTC)

Great work on the article Stenyak ;) --BMF81 00:11, 24 March 2006 (UTC)

I'm pretty sure, that Live for Speed doesn't support active differentials. Can someone check this and correct the article?--Florian Jesse 12:41, 31 March 2007 (UTC)

[edit] More rigorous def needed

The column "grip calculation" distinguish between "interpolated lookup tables" and "realtime equation calculation": this is not very rigorous, and better classification should be found. I found the current info in a forum discussion with the nkPro developer: [1].
ISI physics engines have an approach similar to pacejka, but instead of re-compute all equations in realtime they interpolate lookup tables. Insted in other sims (nKPro and others) the whole formula is computed at every iteraction.

An example that give simple explanation of the ISI engine behaviour is this: there is a table with the grip values for a slip angle x1, a load y1 and a camber z1; there is another table with the grip values for x2, y2, z2; at every interaction of the physics engine, we have the values at xN, yN, zN, and we interpolate the grip values in previous tables to obtain the current grip value. This is of corse a less accurate approach than computing the whole formulas at evety iteraction (unless we have a huge number of tables).

Anybody know how this could be expressed in a more formal manner?--BMF81 16:09, 24 March 2006 (UTC)

Since when did we have conclusive proof that the model rFactor uses is less accurate? Sure, it uses a lookup but it's untold if that harms the handling much at all, let alone the more important one, if the formulae used by the other sims are remotely accurate.

[edit] Lack of bugs == Feature?

I wonder why something like "Significant join lag is NOT present" is present in the tables? All sims have bugs everywhere, but i don't think it's relevant to cite them all in the tables, it doesn't make much sense (IMO at least). --STenyaK 13:41, 26 May 2006 (UTC)

I'm not sure that join lag has to be considered a bug. I looked at join lag absence as an optimization, therefore as a feature.--BMF81 14:25, 26 May 2006 (UTC)

[edit] Sound system

It is currently mentioned that netKar Namie (v0.9.9) has "Dynamic" sound, and not "Sampled". This is wrong. Just browse your harddisk, go to "netKar099namie/cars/modena_street/sfx", and you'll find an "engine.wav" file which lasts for 4 seconds. So I'm correcthing the table ¬_¬ --STenyaK 13:51, 26 May 2006 (UTC)

It could be run through dynamic effects?
Depends on what's considered "dynamic effects" and "sampled". If a change in pitch (or in volume, or doppler effects, or reverberations...) are considered "dynamic", then all games have dynamic sounds, whether based on samples or not.
Also, think about:
* A .wav long sample created by putting a mic on the real car
* A .wav long sample generated thanks to a highly accurate simulation of engine+exhausts :-)
* A .wav short cylinder explosion sample, used by the .exe for creation of final engine sound.
* The .wav sample, hardcoded in the .exe
* The sample, generated by the .exe thanks to some parameters lying in a .ini file.
* The sample, generated by the .exe thanks to a harcoded array of values.
* The sample, generated by the .exe thanks to a couple of hardcoded variables in some function.
* The sample, generated by the .exe thanks to a highly accurate simulation of engine+exhausts ;-)
STenyaK 21:40, 6 January 2007 (UTC)

[edit] RBR sound system

In the article there is that RBR uses two different (sampled and dynamic) sound engines, anyone got a source for this information? I have always thought it uses sampling for engines. The engine sound is generated from a single sound file (per car) that is 2-3 secs long and named e.g. "subaru.eng". For me it does sound like a sampled system, at least not as dynamic as in netKar Pro or LFS.

Please, read my 2nd reply to Sound system STenyaK 21:42, 6 January 2007 (UTC)

[edit] Add TORCS?

Wouldn't it make sense to add the characteristics of TORCS to the article? Unfortunately I can't do that myself, I stumbled upon TORCS in the German Wikipedia, where it is proposed for deletion :-( and have no firsthand knowledge of TORCS or racing sims in general. However, I'd like to see the Open-Source software compared to the other racing sims. Cheers --Make:en 21:55, 28 March 2007 (UTC)

I agree. I play TORCS and it is realistic enough to be included in this list. --24.7.150.166 21:39, 7 November 2007 (UTC)
I added snub entries for TORCS. I will try to fill them as my time permits, but please go ahead and do it if you have the information. --Darth Borehd 11:09, 8 November 2007 (UTC)

[edit] Missing from comparisons

It looks like two of the most popular simracing games ever are missing from this list: Crammond's Grand Prix series and the EA F1 series. --Mal 03:05, 5 April 2007 (UTC)

Please, feel free to add data about them, since they have traditionally been considered as proper sims by the simracing community. In the same way, I'm removing Forza Motorsport 2 from the tables. --STenyaK 04:28, 30 May 2007 (UTC)
I don't think there is a documented empirical term that differentiates a racing simulator from a racing arcade game. It appears to be up to subjective opinion. For that reason, I would oppose the removal of titles merely due to personal perception of whether or not it's an emulator. --Darth Borehd (talk) 04:20, 23 December 2007 (UTC)

[edit] GT Legends Connectivity

While GT Legends allows a setting of 36 (not 32!) maximum players, in our league (gtl.ysuprogramming.com) we found that every time there were more than 27 racers on the track something bad has happened. The excess players were unable to race properly due to all kinds of strange errors.

They seem to have 'overcame' this difficulty in the later GTR2 release by restricting to - guess what, exactly 27 players. Ysu 12:39, 27 April 2007 (UTC)

[edit] Wet weather in rFactor

It should be pointed out that wet weather, (and of a more advanced kind than in F1C) does in fact exist in Rfactor but is currently only available in the version taken to promotional events with BMW/Intel.

[edit] RBR Dynamic Lighting

While RBR doesn't change the time of day during a stage, the in-race lighting is different depending on the time of day the stage is run at and also the amount of cloud cover. As such, maybe the Dynamic Lighting column for RBR should be changed to "partial" with a note?

[edit] Updates

Some thoughs:

  • In the Force Feedback section I'd add a column that reads something like "FF generated directly from suspension behavior", to reflect those games with a good physics engine that allows for a simpler and more faithful FF implementation.
  • In the connection quality I'd rather have one column for the max num. of racers and one for the num. of spectators.
  • Does "3 connected differentials for 4wd" may deserve a particular mention among the diff types?

--BMF81 10:58, 30 May 2007 (UTC)

  • I also thought about the "generated from suspension", but then you can see some sims generate them from suspensions. Some from tires. Some from steering wheel (having computed torques and forces through the steering rack). So i think that is the best choice. The thing that differentiates sims the most here is whether they use canned effects or not. By canned i mean that don't derive directly from physics engine torques, but rather from other indirect sources, like "knowing the tire is going over some piece of road that is supposed to produce something similar to a sine-waved torque, with linear dependency to speed". Or knowing that "it's going through some predefined volume codenamed 'water', where we are supposed to double the dampening of the wheel torques". Hope i explained myself well enough :-)
  • I agree with the viewers vs. drivers columns, i'll add them.
  • Since 4wd drive means the 4 wheels get torque, I added it. It may be useful to add other colums specifying whether the torques are split with fixed percentages, or if an active diff is used, or an open one, or torsen or whatnot, but I think it's not that important in the table (at least not yet). I've tried to keep the tables clean from excessive detail, which were beginning to render them useless at times.

--STenyaK 00:26, 31 May 2007 (UTC)

[edit] Eh

I listed this on the racing game page, so I will quote myself:

"Why Racer is listed here as a simulation game, while Gran Turismo as a semi-simulation? Here listed, that GT cannot qualify to a full-time-simulation without damage... So why Racer can? Car damage was in GT2, although only performance was affected."

I suggest to list what simulation elements are present in Gran Turismo, as I checked, there would be many "Yesses" on this list...

And the last thing, why there is no element to what can be tuned and what can't? Most simulated car parts in GT are configurable (reminds of real life), but in many other "simulation" games they are not (reminds of some kart racing arcade, or a NES game)... I think the BIG part of driving simulation is in the tuning of your car. Y'know, the gear ratios, the suspension, brake bias, downforce, ride height, configuration of systems, such as LSD, or yaw control...

P.S. I like the part on "damage" about "driver injuries"... =) GTA has it! =) Echad 00:09, 9 August 2007 (UTC)

GT doesn't correctly compute one of the main and most basic equations of a realistic physics engine: inertia. Rotational inertia is modified so that the car doesn't roll upside down, or lands always on its four wheel after a jump. If something so basic and simple as inertia is done incorrectly on purpose, the purpose of the game is logically not a real simulation. Regardless of the excuses used on some other page to qualify it as an arcade. --STenyaK (talk) 17:11, 31 May 2008 (UTC)

[edit] Moment of inertia

Is there any sim out there that doesn't simulate the moment of inertia? I think that's the first thing you do when programming a physic engine. --Florian —Preceding unsigned comment added by 88.66.3.125 (talk) 20:54, 9 October 2007 (UTC)

[edit] Is there any Sim racing video game which ..

..is aimed at educating its gamers to follow the driving rules (signals, road signs and speeds)? Acidburn24m 18:19, 3 December 2007 (UTC)

[edit] popularity

I would think that popularity would be one of the most important/relevant aspects in comparing racing simulators. Maybe the article should give info on this. Like why would anyone bother with an online sim if nobody plays it, or if a developing game will never be finished? I've no idea how to find info on popularity, but surely some details can be discovered online. Peoplesunionpro (talk) 16:51, 19 December 2007 (UTC)

I don't think that's good idea because there cannot be "facts" on this subject. Only possibly online player counts but that's hard to find out too. There's too much personal opinions affecting "popularity". I think this it's a good to stay just listing what aspects the games simulates and what features they have like the article does currently. Timppis (talk) 18:06, 19 December 2007 (UTC)

[edit] Damage -> Driver injuries

Has any sense to have this column at the tables? It should be removed, I doubt any SIM will simulate driver injuries. —Preceding unsigned comment added by 85.58.89.99 (talk) 21:58, 21 December 2007 (UTC)

Richard Burns Rally does simulate injuries of the driver (both the user and the AI drivers), preventing hospitalized people from finishing the rest of stages of a Rally and similar things. --STenyaK (talk) 17:14, 31 May 2008 (UTC)

[edit] What about popular console simulators

I think that the Gran Turismo Series (beginning on Sony Playstation 1) should at least be mentioned. I don´t play PC games much, and the Gran Turimo (the first) was the first racing game I played that said to be a "racing simulator". The game has been one of the major games for Sony's console. I doubt that a video game owner hasn´t heard of this series. Check, for a start, Gran Turismo (series).

Also, the game Enthusia, for PS2, is a very good simulator. It might be worth compare it to the other games on this article. Check the opening video showing the comparison of a car and the game, on a movie site. I really like this game. It beats Gran Turismo a lot, IMHO.

Finally, there is a racing series for Microsoft's console, XBOX. Is the Forza_Motorsport. I have played it a few times, and it has a few good qualities (of simulation). The game has a sequel for the XBOX 360, and it a concurrent for the Sony's simulator series. —Preceding unsigned comment added by 201.17.132.153 (talk) 11:27, 8 January 2008 (UTC)

I bet everyone who has contributed to this article knows Enthusia, Gran Turismo and Forza series. Reason why they are not included in this article, is that no one has (yet?) included them. Someone from these games' fan base should show up and include them - but one just can't add them and put "yes" to everything, one really needs to know what is included/simulated/calculated and what is not. Timppis (talk) 07:14, 3 February 2008 (UTC)
Forza got taken out. "Please, feel free to add data about them, since they have traditionally been considered as proper sims by the simracing community. In the same way, I'm removing Forza Motorsport 2 from the tables. --STenyaK 04:28, 30 May 2007 (UTC)" from the "Missing from Comparisons" bit on this talk page. Someone decided arbitrarily that Forza isn't actually a "proper sim" as regarded by the "simracing community" who appear to be the self-appointed judges of these things. If you know about it, add it back in and keep doing so until this nerd gets apoplexy from clicking "revert" too often. I'd do it myself but I know nothing about Forza, only some other sims (all featured here) and self-important internet geeks on Wikipedia.88.67.240.234 (talk) 20:57, 25 May 2008 (UTC)

[edit] LFS patch Y changes

I can't confirm the following, but as far as I know, patch Y introduced (to confirm by someone): Data form tyre manufacturer: Avon (suddenly lots of AVON ads got into the game) Racing line affects grip - I'm almost sure about this, even the autor said it in the patch Y video —Preceding unsigned comment added by 194.29.137.43 (talk) 01:12, 24 January 2008 (UTC)

Only changes to tyre physics in Patch Y are that they heat up and wear more quickly. There is no Avon data introduced. Just like there was previously Michelin tyres too - the ad boards on the tracks are just for visuals and tyre sidewalls (you can choose between, Avon, Michelin and other 2 fictional brands) are also just for visuals too currently.
Racing line affects grip? Don't remember what was said in the video but I think you might have misconfused it with the real life test driving what Scawen probably meant? Timppis (talk) 07:07, 3 February 2008 (UTC)
The racing line statement fell as part of the improved AI, it has nothing to do with grip. —Preceding unsigned comment added by 84.132.236.25 (talk) 21:07, 7 February 2008 (UTC)

[edit] Operating System

I'd love to see an Operating System category ie. Yes/No for Windows, Mac OS X, Linux etc. Possibly also system requirements and use of extra controllers (Steering wheels and such). —Preceding unsigned comment added by 59.167.60.62 (talk) 03:36, 25 April 2008 (UTC)

[edit] rFactor tyre model

rFactor's tyre model is stated as being "Pacejka-like", but, according to the rFactor#Physics section, rFactor claim their model "to be much better than the Pacejka model." —Preceding unsigned comment added by Andreas Toth (talk • contribs) 01:55, 8 May 2008 (UTC)


But this doesn't reflect anything. In reality, the tyre model used in ISI engine isn't based on any formula(s) like LFS, NR2k3, or nKPro, especially for single lateral/longitudinal force(e.g. no friction circle theory at this moment), I think. Instead, it is combination of some basic parameters like load sensitivity(both peak and coefficient of friction), drop off function or simply damping/spring rate of tire. But the major problem propably is in the combination of tyre forces, because(according to the analysis done by 'Niels_at_home') ISI's tyre model cannot do a one convincing(realistic) resultant force, thus making the car feel just different/wrong in most cases. But that are only my speculations, of course. Skorpio pl (talk) 14:15, 23 May 2008 (UTC)