I agree, GTR2 is a significant improvement over GTR and especially rFactor and joins GTL as the only ISI-engined games I can actually play without thinking 'god, this is crap'. I'll almost certainly pick it up when it hits the stores.
I am really looking forward to see the quality of games SimBin/Blimey! are capable of creating once they finally distance themselves from ISI and develop their own sim engine.
User LOD: 1.00
Dust LOD: 1.00
Mirror LOD: 0.30
Screen width affects LOD: yes
Track LOD reduction: yes
Mip Bias: 0.00
Shadow type: high res
Simple wheels: mirror
Disable trees: mirror
Disable flags: mirror
Disable rubber: mirror
Enable dither: yes
Enable mip filter: yes
Use compressed skins: no
Half texture size: no
Hardware vertex shading: on
Z-buffer depth: 24
Simple track: off
Haze effect: yes
Adjusting the various in-game settings results in a ~5fps difference either side of the average i.e. no changes significant enough to account for a 20-30fps shortfall.
Now that I have unlocked the other cars in the demo, installed a handful of the better GTL add-on tracks, added some tweaked tyre curves and reduced some of the artificial feeling canned force feedback effects, my impressions of the game are much better.
Rain, headlights and the Motec display are still framerate hogs but as long as the framerate stays above 30-35fps to avoid the controller latency which plagues ISI engined games, it's a lot of fun. Threading the Saleen or the Murcielago through the Lime Rock Park or Laguna Seca circuits at night is a blast.
Athlon64 3500+, 2GB DDR, GeForce 6800GS. I'm running the demo at 1280x960x32 with DirectX 8 shaders and 8xS AA, 8x AF. I might need to reduce AA and AF and drop the resolution down to 1024x768 and even drop the colour depth to 16 bits to see if I can get it to run smoothly.
[edit]: just turned off AA and AF and reduced the resolution to 1024x768x16. In the rain at night with headlights on I'm getting 25-35 fps on track by myself and the 10-15fps on the starting grid with 15 AI cars. It looks like crap at those settings too. I was getting similar framerates when I was running at a higher resolution and with AA and AF enabled so they don't seem to be the limiting factor.
Lots of people over at RSC are saying that they get good performance on relatively low spec systems though, so either my expectations are too high, their expectations are too low or my system has a bottleneck that GTR2 doesn't like. Best to give it a go on your own system and see what the performance is like.
Second impressions after spending a couple of hours with it:
- this thing is a resource hog. Turning hotlaps by myself around the track I usually see framerates around the 40-50fps mark. Add some AI cars into the mix and the framerate drops down to 25-30fps. Add headlights and it drops even more. Add rain and the framerate drops through the floor. Unfortunately, the controller response noticeably degrades when the framerate falls below 25fps or so and the car handling goes all to hell. I've tried reducing graphics options, reducing screen resolution and running DirectX 8 shaders but still can't get the game to run consistently smoothly on my system (Athlon64 3500+, 2GB DDR, GeForce 6800GS).
- lovely lighting effects; the gradual transition from day to dusk to night back and back to dawn is gorgeous. Headlights are a resource hog but look fantastic.
- the handling is significantly better than other ISI engined games but I still don't feel connected to the car. There's still something missing. It's hard to describe but the handling feels artificial, synthetic. It's missing that visceral, dynamic element that makes a car feel alive.
- the artificial force feedback effects are annoying and tell me nothing about the car. I suppose they're intended to improve the driving experience and enhance immersion but they really do nothing for me.
After playing with the demo for a couple of hours, I loaded up LFS and took a couple of the GTR style cars out for some laps around Aston. I was immediately struck by how much smoother LFS runs, how clean the graphics are and how responsive and alive the cars feel. If only SimBin had used the LFS engine instead of that fricking ISI gMotor engine ...
Just gave the demo a shot and my first impressions are mixed.
The controller latency which seems to plague ISI engined games - and especially rFactor - seems to have been addressed and is much improved. I get a much better feel for the dynamics of the car. I can finally sense it starting to slide and break loose and can countersteer and use the throttle to recover. Very encouraging.
Force feedback is the same mixed bag found in other ISI and SimBin titles. Bleh.
Graphics are disappointing which is a surprise since I expected them to be a highlight. They're not much different than GTR and GTL and manage to look both washed out and vaguely cartoony at the same time. The track could have been lifted straight from GTR. Rain effects are nice and atmospheric though I would have preferred much reduced visibility through the windscreen. Unhappily, it also murders my framerate. Lined up at the start with a full-grid and no rain, I get around 27fps. With rain, that plummets to 9fps, which is unplayable. Once the field spreads out a bit and there are only a handful of cars in the same section of track, the framerate creeps back up to around 17fps with rain and 45fps without. I suspect the novelty of rain will quickly wear off.
Sound is on par with GTR and GTL. Good but not great.
All in all, it feels almost like a slightly newer version of the original GTR with some refinements and a bit of spit and polish. A bit of a disappointment though the improved feel and handling of the cars is encouraging.
I use a 60 degree FOV but I also use TrackIR which means that I can look around very easily so the relatively constrained field of view isn't a problem.
I agree that the tracks seem quite wide in places; like you say, sometimes it feels as though you're driving down a freeway/motorway instead of a race track. If the tracks were a couple of metres narrower and the railing, fencing and other trackside objects closer to the edge of the track, the sensation of speed would be better. More trackside objects help too: the sensation of speed at Aston, for instance, has been improved by the increased number of trackside objects.
Thanks for double-checking, Kev, it's very much appreciated. I am glad to hear that someone else is seeing the same thing, as it means I don't have to futz around with my controllers.
I've double- and triple-checked my axis settings and calibration using DxTweak2 and in-game, I've used the virtual pedals to check for background brake input and an external check of the brake lights shows that they are not engaged (I know the granularity of these last two tests isn't very accurate).
One possible factor here is that I use a clutch pedal and manual shifting but that shouldn't affect non-driven wheels. Yet, when I flip a car onto its back, the non-driven wheels slow to a stop by themselves and the driven wheels do the same if I disengage the clutch (there is some RPM even when the engine is idling so if the clutch is engaged the driven wheels will spin).
Can one or two of you guys do me a favour and do a quick double-check on your own systems to see if the non-driven wheels on a flipped car will continue to spin indefinitely if you don't apply brakes or whether they gradually slow down and stop? I'm using non-driven wheels to test since driven wheels have more confounding factors to consider. If so, I'm going to have to do some thorough troubleshooting of my controllers.
Thanks.
edit: I just did another couple of flip tests and this time one took over 3 minutes for the non-driven wheel to come to a complete stop. Three minutes is much longer than the other times but even so the wheels did stop eventually.
edit: it happens even when the car is on its back with both sets of wheels free to spin and it happens to both front and rear wheels. Since in the UFR the rear wheels are non-driven and are not connected via a diff, perhaps friction is modeled in the running gear after all.
Last night I noticed one of those little touches that drive home how detailed the underlying physics model is in LFS. I flipped my UF GTR onto its side and noticed in an external view that wheels which are freewheeling will gradually slow down and stop, presumably meaning that friction in the running gear is modeled. It looked very realistic. Who knows, maybe it happens in other sims too, I've just never noticed.
This thread is embarassing. I'm in the middle of trying to convince a handful of people to try out LFS and it turns out that if they come here to the official forums one of the top threads is about how to play cops and robbers? Gad, they'll think LFS is full of 10 year olds.
Does anyone have working links to high-quality versions of Steven Kranhold's (Kraniwani) LFS movies "Just Driving" and "Szwei"? The links previously posted to Kraniwani's website no longer work. The movies are up on YouTube but the quality of the compressed movies is poor.
Write my own articles in order to see some LFS coverage? Why should I have to? These are supposedly professional or aspiring-to-be-professional magazines, they have advertizing, marketing and corporate relations staff, they have their own cadres of in-house editors and staff writers who write the overwhelming bulk of articles and other content and their objective is to "cover the exciting sport and hobby of simulated racing". Yet apparently their staff, who are able to churn out seemingly endless copy about FILSCA and rFactor, aren't able to keep up with what's going on in LFS and/or don't think there's anything worth reporting on. That's a crock. There's plenty of material to cover. There's just no interest in their part in doing so. You can practically smell the politics; it reeks.
And on that note, I'm done. You're welcome to the last word if you want it.
I'm really not sure what to make of this. The track doesn't resemble LFS's Blackwood rallycross very much apart from its general layout e.g. crest here, right turn onto dirt here etc. Distances, elevations, track widths, bumps and dips etc are all quite different. The look and feel of the place is quite different too. It looks to these tired old eyes like something out of a console racer. In fact, if I didn't already know that it was supposed to be Blackwood it probably would have taken a while to realize.
I'm a bit surprised that Scavier agreed to this conversion. On the one hand, I suppose it increases LFS's visibility. On the other hand, it isn't a very good representation of the quality of tracks in LFS and I wonder if people who are unfamiliar with LFS might come away with the wrong impression?
There's a lot going on in LFS that touches on the sim-racing world at large right now. Just off the top of my head, here are three subjects that any journalist worth his salt should be able to construct an article or two around:
1. the advent of the wrecker barricade to combat online wreckers. This touches on one of the biggest downsides to online racing events and online gaming in general: intentional wreckers. The topic can be spun out to a larger article easily. Discussion of the wrecker barricade could spur other user groups to do something similar with their game of choice, or even collaborate to compile a cross-game database of known wreckers. Lots of material in there;
2. the Intel Racing Tour 2006, where people compete against one another in virtual races using LFS with a BMW M6 as the grand prize. This leads to a discussion of increasing corporate sponsorship and increasing visibility of racing sims in the mainstream. You could also touch on other teams, series and leagues that have corporate sponsorship and more mainstream coverage e.g. ESL Germany which has LFS races broadcast on GIGA TV and the net, the upcoming ESL UK series, etc;
3. the recent release of the BMW Sauber F1 car for LFS (and possibly rFactor) leads to a discussion of IP licensing issues in racing sims /gaming/hobbies as a whole which could be an entire series of articles in its own right, with obvious references back to the corporate aspects of item 2 above;
There's plenty of material to report on, you just need to stop relying on members of the community to send you material and instead start doing some real journalism.
You don't get it, do you? They already have nine people on staff. If, between them, they are unable to keep their finger sufficiently on the pulse of what's going on in LFS to provide coverage, that's a sign of an apathetic editorial staff. To then turn around and blame lack of coverage on the userbase is beyond the pale. It's all moot anyway, as I don't see much point in these e-zines; almost all of the pertinent information in them is old news and has been hashed and rehashed several times over by the time they're released and the rest is filler.