Thats with AA set to 16x (although i can tell the difference between 4x and 16x, and 4x doesn't need an uber PC to use) and AF to 16x (which most modern cards can do quite easily), Hi-Res skins (2048's) and some new textures.
Ok,
But if I`m not using Hi-Res skins, would I be able to have a game looking similar good like that with AA and AF maxed?
Sorry for going off topic btw.
No, I'm not happy at all to upgrade, it is real pain of arse, also does cost few grands too no matter how to do it so that I could delay unavoidable as long as possible ;D
LFS runs just and just ok (constant 60fps in all situations) with my home machine now, but I can't use all details, needed to reduce LOD already, so upgrade is needed for that too, it is just not a must upgrade yet, but for rtractor it is needed so I could do more testing for that praised dx9.
Using computer for playing modern games does cost around 1000 euros / year at current rate + costs that comes when you jump into crazy train of upgrades which is around 2000-3000 euros depending a bit how soon you like to upgrade your machine.
Also this fast ageing process products loads of problematic waste as older parts are no good for anything and gets thrown away and again this stresses our planet in bigger scale. So if you like from green bandwagon, then check how much trouble this consuming madness causes and think if it is CO2 or wastes that is going to be a problem sooner in future.
@Jakg, holy screenshot Batman, need to scroll sideways for text now, but it sure is pretty and eyecandy, however when racing I doubt that I could see difference at all. Btw, does that yellow line float above track there as it does in my computer?
Personally I don't have too much of a problem upgrading every 3 or 4 years or so, because my main interest is computers, but I agree that the waste factor is a big problem, I wish more were possible to help cut down on computer and tech waste in general. Green computing was an idea which was tossed around years ago, but I don't hear anything about it these days (making components from less toxic materials and designing things to be more recyclable).
I did repair computers for living over 4 years and I have been teaching how to use them few years, so I'm surely interested from them, but 'child of shoemaker has no shoes' seem to be true, I don't like to be on edge with my own computer, I would prefer set and forget method with it
I don't see why we need so many sockets and why always there is need to get new motherboard, why not create standard board where manufacturers could add what they like but connectors and sockets would be what they are so new board is not needed every time, maybe such board that you could get upgraded without replacing whole board. It would be a challenge, but that would help a lot already.
I did upgrade gfx card drivers yesterday as I had bit of trouble with old ones, got bit of performance increase too which is good, but now computer uses 60MB more ram than before and only because of driver likes to keep parts of control centre in memory so that if I happen to start it someday it will start faster, old one had that too but took lot less memory, so If I shut down those processes I get 80MB more available ram!! Such resource wasting should be penalized with something that hurts and hurts really long time, if we take 10 different programs or drivers and if all waste resources like this, it is no surprise that we need 3Ghz cpu and 2GB of ram to run windows itself.
Using DX10 or similar is bit of same if features that it's offer is not really needed for anything, best is to make software so that it is as light and fast as possible and only parts that really are needed are used from 'outside' where programmers may have had bit different approach, like with direct x it is eye candy over performance in many cases.
Yes, also MSI, Abit, etc. We don't see more environmental friendly computer manufacturing until something forces them to make parts differently and when that happens we see how few big companies shuts down, enviromentally good products are not those that will make loads of money, because instead of 2 years they tend to last 6 years or more.
Bit of egg or chicken problem again, also if magically consumers starts to demand products that last longer (which will never happen at least as long as air is free) prices will go up quite lot, so you don't save money actually at all, but those are current realities, it is not possible to save our economy and planet at same time.
But if software makers do better programs, then pace when you need to upgrade is at least bit reduced.
I couldn't care less about environmental issues with hardware manufacturing but if that makes prices drop or more streamlined socket upgrade solutions that makes more sense, then I'm all for waving Greenpeace flag.
DX5 is just way too old now. LFS began to be coded from the time that DX8 came in- they were using the present technology of the time. I can't think of any game that's compatible from DX5 through to DX8 (but perhaps there are some), and considering that many people are looking forward to potential new features made possible by DX9/10, going back to the older DX5 just seems really strange!
Obviously LFS works on your computer now, otherwise you wouldn't be playing it
In almost all games (and I suppose LFS too), during frame update, physics engine does not run at all, and later, it runs several times in a raw to keep the pace. For example, when you have 25 fps and 100 hz physics engine (like LFS), physics engine will run 4 times in a raw after every frame, 3 times to simulate what happened during the frame update, and 1 time for next step.
It does not cause any inaccuracy but....The problem is that physics engine does not know how the controller input varied during frame update, and it will run 4 times with the same controller input.
Conclusion:
If you have 25fps, this is an average, and you can expect that sometimes the framerate drops to 10fps for a few frames...
If you have X fps, physics engine see your input (controller) and the network at Xfps. When it reaches 10fps it causes big problem because the engine sees your moves a lot more brutal than they really are (stroboscopic view) and misses quick moves you make.
And there is a one frame duration delay before the physics engine knows about your move, and another frame duration delay before you receive FFB corresponding to your move. At 25fps you have 0.08s delay between move and FFB, this is enough to cause crazy wheel oscillations, prevent the wheel from countersteering effectively...etc
Same for network activity...60+ fps is a LOT better than 25fps.
Nice, this is already a great improvement for smoothing problems, but it does not solve the problem I am talking about. I hope I will manage to explain it...
This improvement in X version means that every time the physics engine updates, it queries the controller state. So controller query, like physics engine, runs *in average* 100 times per second.
But don't forget that if you have 25 frames per second, frame is displayed and then physics/controller query runs 4 times in a raw to keep the 100hz pace.
3 times to simulate what happened during frame update, and 1 time for the new update. And every time it queries for controller state.
So, it means that the 3 physics updates simulating what happened during frame update use controller state AFTER the frame update, not controller state DURING the frame update. Changes in controller state during the frame update are lost in most games (unless in LFS they use another trick, but I doubt controller inputs can be acquired during frame update...but maybe they manage to do it?)->stroboscopic vision.
They did this change (query controller input every physics update) to have consistant smoothing on all computers. rFactor did it too. Simbin games (GTL/GTR/GTR2/Race) still have a framerate dependant controller smoothing...you have to change smoothing coefficient depending on your framerate to get the same feeling on different computers (tested it)
Maybe in LFS they manage to query controller in the middle of frame update...then it is perfect...but until now most games don't do that as it is slower, and then higher framerate in game gives a better responsiveness, less controller lag, better interaction with other network players.
your vision of one motherboard is next to impossible. it would be too hard to have standard sockets. Jst because the CPU and other chips are almost binary compatible dosent mean there design are anywhere near similar. infact there radically different.
thats jst plane wrong
Personally i never felt the grahics in terms of candy really bothered me for LFS. sorta viewed it as a game where the physics are the concentration. I can ignore the Gfx because its not what the game is about.
its really the trend of todays games expsecially PC games to jst use the extra power they get with newer hardware, to do what they were doing before and not to innovate.
IE writing more and more complex shaders for surfaces.
I feel LFS should apply it to something more contructive in a sim. few examples Pixel Shaders -
Used to visually simulate Tire Wear, Graining, dirt
Procedural Track wetness
Really should step it up and be able to have Different parts of the track Dry and different speeds and have it visually represented
Windshield Wipers-
with the advent of Force cockpit view it would be nice to have water on the windsheild visually represented and have the wipers effect it
all of those ad to the overall simulation aspect and prob would look sick as well. on top of that it would be a first for a sim i think i would could marketing and bragging points