BWX232, almost every post I have ever seen you make about LFS is a criticism or is negative in some way. Is there anything about LFS you actually like? Frankly, if I disliked something as much as you appear to dislike LFS, I would have deleted it and moved on to bigger and better things long ago. Life is too short to waste your free time on things you don't enjoy.
The only cars in rFactor that I come back to are the VW Lupo mod and the stock 2006 Rhez which was included in the 1150 update. Most of the others leave me cold.
I see it as a strength. It means that their vision does not become diluted and ensures that there is a coherence and consistency to the end-product. That's the upside. The downsides are that product updates are less frequent than they might be otherwise and the devs run the risk of becoming burnt out or bored with no one able to pick up the slack if and when they falter. I'd much rather wait a couple of years for S3 than have the devs sacrifice quality for the sake of getting product out the door.
It's not a race. LFS is being developed at its own pace. Sure, I'd like to see more frequent updates but the end product produced by Scawen, Erik and Victor is so obviously better than the work produced by other dev teams that their approach clearly works. If you get bored of LFS, move on to something else. You'll come back to LFS eventually.
Just a coincidence I'm sure but RMachucaA's first post, which was more or less LFS's first exposure to the gaming world at large, was made on August 18, 2002, exactly 4 years ago.
"The Canadian's absence from the Hungarian Grand Prix was seen by many as the first crack in what has appeared to be an improved relationship this season, but Jacques Villeneuve will not be back in the F1.06 after both parties confirmed a termination of their relationship."
Time for Villeneuve to finally turn his back on F1 for good and move on. I hope he lands a seat in a series like DTM or ALMS instead of the rumoured bid for a seat in NASCAR though.
There are none so blind as those who will not see.
Forbin, do pay attention, I will say this only once more. Either you get it or you don't and if you don't get it by the end of this post, you never will.
With those images fresh in your mind, let's move on.
With ATI, there is no need to use a negative mipmap LOD bias since ATI's anisotropic filtering does not produce texture smearing artefacts (see the screenshot above). Read that sentence over again until you completely understand it.
nVidia's anisotropic filtering does produce texture smearing artefacts (see the screenshot above) so you are forced to either live with the smearing or use a negative mipmap LOD bias and put up with the resulting texture shimmering.
You said yourself that you see texture smearing on your nVidia card when the bias is set to 'Clamp', right? You chose to use a negative LOD bias setting instead which means that you end up with some texture shimmering. Well, guess what? If you had an ATI card, you wouldn't see any texture smearing in the first place so you wouldn't have to use a negative LOD bias and therefore wouldn't get any texture shimmering either.
It's quite simple really: nVidia forces you into a compromise. ATI doesn't.
And on that note, I'm done. Enjoy your nVidia card.
Your screenshots show problems typical of using a negative mipmap LOD bias. I've highlighted some areas where the problems are particularly noticeable below:
These areas probably seem to flicker a bit when you're driving. This is texture shimmering.
You can get rid of texture shimmering by setting the Negative LOD bias setting in your video card driver to 'Clamp' as shown in the following screenshot:
The downside to doing this with nVidia cards is that you may start to notice smearing of textures in the mid-distance.
One other thing I noticed is that there is a lot of aliasing in your second screenshot, the one taken in-cockpit. Are you sure your AA was set to 8xS for that screenshot?
Texture shimmering is not seen in static screenshots, it only becomes apparent when the scene is in motion. Don't take my word for it, read the section about Negative LOD Bias in the nHancer link I provided previously "...[b]ut if the scene is moving, the textures start to shimmer." So, all your test shows is that mipmap LOD bias settings affect texture sharpness. Yes, that's exactly what it's supposed to do.
The point that you seem to be taking pains to ignore is that the reason you have to adjust mipmap LOD bias settings with nVidia cards in the first place is that nVidia's anisotropic filtering results in texture smearing. Unfortunately, sharpening textures through negative mipmap LOD biases results in texture shimmer when the scene is in motion. So, you're forced to find a compromise between smearing and shimmer and that compromise varies from game to game and even track to track sometimes.
With ATI, you're not forced into making a compromise because their anisotropic filtering does not result in smeared textures. This means that you don't have to adjust mipmap LOD bias settings at all. You get crisp, detailed textures with no smearing and no shimmering at the same time.
From an image quality standpoint, the choice between ATI and nVidia is a no-brainer.
I encourage anyone considering a new video card to do some careful research into the differences between ATI and nVidia image quality before commiting to any card or any brand. Unless you're looking to buy one of these (http://pcgames.gwn.com/news/story.php/id/10042/) in which case, nVidia all the way baby!
Guys, for those of you who are having problems with steering feel and force feedback, try adjusting these settings in your controller.ini configuration file, which is located in the UserData/<USERNAME> folder.
Fixed Rate Inputs="1" // Forces the game to poll input at a constant 100Hz
Steer Ratio Speed="0.00000" // Removes annoying speed sensitive steering
Speed Sensitive Steering="0.00000" // Removes annoying speed sensitive steering
Steering Wheel Range="540" // Sets the in-game steering wheel range
FFB Effects Level="1" // Gets rid of some canned effects
FFB steer force neutral range="0.00000" // Removes force feedback center deadzone
FFB steer force input max="-11500.00000" // Need to use negative values for Logitech wheels
FFB steer force grip factor="0.10000" // Reduces lost force feedback during understeer
FFB steer front grip fract="0.00000" // Reduces lost force feedback during understeer
I find these settings dramatically improve the steering feel; it's still nowhere near as good as LFS but still much better than stock.
If you had bothered to read my post instead of leaping to the conclusion that I have some agenda against nVidia, you would have seen that I set the bias to Clamp because when it is set to Allow there is major texture shimmering when the scene is in motion. If you think I'm making that up, check the Negative LOD bias section here: http://www.nhancer.com/help/optimizations.php (nHancer is a tool for tweaking nVidia driver settings, by the way). Here's the relevant text:
"Negative LOD BIAS
The so called Level of Detail BIAS(LOD BIAS) controls which at which distance from the viewer the switch to lower resolution mip maps takes place (see here for more details about mip maps). The standard value of the LOD BIAS is 0.0 (zero). If you lower the LOD BIAS below zero, the mip map levels are moved farther away, resulting in seemingly sharper textures. But if the scene is moving, the textures start to shimmer.
Because of this, it's not a good idea to use a lower LOD BIAS to improve the sharpness of the image. It's better to use an Anisotropic Filter instead.
Some games force a negative LOD BIAS nevertheless. The result is heavy texture shimmering. To avoid this, the driver can clamp the LOD BIAS to zero. That means that the LOD BIAS can still be raised above zero, but it cannot set lower than zero."
Check out the road markings and rumble strips at South City and the track lines, rumble strips and main start-finish straight grandstands at Kyoto Ring for examples of bad texture shimmer in LFS. Other games such as GTR2, rFactor, GTL etc all exhibit very heavy texture shimmering.
So, with nVidia, I am forced to compromise between texture smearing or texture shimmering. That's like having to choose between a chocolate flavoured pile of crap and a vanilla flavoured pile of crap. With ATI, on the other hand, you are not forced into a compromise and can have crisp detailed textures without texture shimmer. Hence my choice to absorb the cost of the nVidia card, call it an object lesson and move back to ATI. If you're happy slurping down variously flavoured piles of crap, more power to you, just don't try to convince others that what you're eating tastes great and that they should have some too.
To those who are interested in the differences in image quality between ATI and nVidia cards, I suggest doing a google search on the keywords "image quality" nVidia ATI, do some research, see what the general concensus of opinion is for yourselves and don't rely on a handful of posts from any one forum.
That is the direct opposite of my experience and the exact opposite of the image quality differences seen in the screenshots I posted above. I'll believe concrete proof and the evidence of my own eyes.
Here are some comparison screenshots that I made several weeks ago when discussing ATI vs nVidia image quality on another board. The images were captured in rFactor since that was the game being discussed but the same image quality differences are seen in LFS too.
All images are captured at 1280x960x32 with 4xAA and 8xAF enabled in the drivers. The ATI card is a 9800Pro and the nVidia card is a 6800GS. The cards were swapped into the same system so all other components are identical. I spent several hours trying different nVidia drivers (currently running ForceWare 91.28) and tweaking settings to get the best possible image quality vs performance that I could.
The moiré pattern effect seen in the nVidia image causes texture shimmering in-game when moving. You can also see some texture smearing of the white lines in the nVidia image that isn't present in the ATI image.
The texture smearing seen above is due to having Negative LOD Bias set to Clamp rather than Allow in the drivers. Negative LOD bias sharpens the textures but also causes terrible texture shimmering when the scene is in motion. The texture shimmering is bad enough to be an eyesore so I leave my setting at Clamp as a compromise.
The thing is, with ATI there's no need to compromise: you can have clean lines and no shimmering at the same time.
Believe it or not, the steering lag in the new version of rFactor is a significant improvement over previous versions. Previous versions polled the controllers once every frame so if your framerate was 20fps the controllers were polled at only 20Hz. The new version uses a separate thread which polls the controllers at a constant 100Hz which makes the steering much more responsive. If you think there's lag now, be thankful you didn't try it before!
There's a lot about this latest version of rFactor that I like and which has been improved upon but there's still a lot about it that I don't like. GTR2 delivers a better driving experience in my opinion than rFactor -- especially when using axus' tweaked tyre curves -- which is strange considering that they both use similar game engines. I suspect ISI aren't very good at creating content while SimBin/Blimey! are able to wring out the most possible from ISI's game engine.
All in all though, in my opinion LFS still provides the best, most natural feeling and most dynamic driving experience of all the games currently on the market. However, the gap is narrowing and the amount of quality content available for GTL/GTR2/rFactor is steadily increasing. I have been spending quite a bit of time recently running the GTR2 Saleen around some of the better GTL add-on tracks (Limerock Park, Nordschleife, Croft, Assen, Brands Hatch, Watkins Glen, etc) and have been having a lot of fun. I just wish the LFS dynamics and driving feel could be merged somehow with the GTR2/GTL content ...
I moved from an ATI 9800Pro to an nVidia 6800GS a couple of months ago. The nVidia image quality is significantly worse; using the same resolution, AA and AF settings, the nVidia card exhibits anisotropic filtering artefacts (texture smearing, texture shimmering, moire patterns) and anti-aliasing artefacts (poor smoothing, loss of fine detail) that did not happen with the ATI card. I eventually decided to dump the 6800GS and replace it with an ATI X850XT, which should arrive next week.
All of the recent ISI engined games (GTR, GTL, GTR2, rFactor) are like that and I agree that it looks terrible. There is some debate about the cause of the differences between the perceived dynamics of the car in internal vs external views. Some people argue that the external view give a true representation of the underlying physics and that the internal view uses effects (cockpit shake, eyepoint/head movement, sound and force feedback effects) to mask poor dynamics. Other people argue that the external view uses a simplified visual model which doesn't properly depict the underlying physics. The new Rhez 2006 car in the most recent rFactor patch does move around quite realistically in external view, so I'm not sure which camp is right.
A new version of the Drivers Republic alpha has been released with improved netcode and additional content (a clone of the Limerock Park track).
The new version also includes an invite function which allows each of the 400 original testers to invite 5 more people to take part in the testing.
If anyone wants to participate, PM me with your name and email address (required by the invite procedure) and I'll send out an invite. [edit: sorry folks, all my invites have now been used. Maybe someone else has some to spare?]
Thanks axus, the new Pirelli tyres feel great. Just did a few sessions around Lime Rock Park with the modified tyre definitions and the cars feel much more lively and dynamic. Nice job!
I agree with ShannonN, the forum is getting out of hand. I am trying to persuade several people to try LFS and am frankly embarassed by the state of the forum and what they must think of the LFS 'community' as a result.
Many of the useless posts come from the same handful of people; several are prolific posters and most seem to be fairly young. I suspect that they're on summer break from school and are bored.
Perhaps if there was a way to limit the number of posts per day to, say, 5, that might serve to reduce the amount of noise on the forums?
The fact that the database server for the multiplayer lobby has been down for several days and is still down even after kunos' miraculous reappearance should provide some indication of how actively kunos et al are supporting their product. If they can't even be bothered to bring the database server back online, what reason is there to think that they're doing any work on netKar PRO at all?