The online racing simulator
Quote from Linsen :Yeah, I see what you mean. But how about taking the lfs-benchmark approach and applying it to your situation? Meaning: Post a couple of mpr's, each on a different track, with a full grid and state the max-settings used and the min-settings used (you could even post the corresponding cfg-files) as well as the car one should switch to. Everybody who wants to compare could then dl the mpr's and run them at the same settings (once on max, once on min). Than we'd have something to compare and see if those 40 fps which seem to be prevalent on the start are down to coincidence or in fact might be an LFS-inherent issue. The way I see it, it is very well possible that many people adjust their graphic settings to a level that just leads to those 40 fps accidentally.

Edit: This was to Zyber's post #21 of course.

Done file is included

Make a backup of your cfg.txt file or simply rename it. (main lfs folder)
Place one of the cfg.txt's included in this zip there. ('/min' and '/max' folder)
Copy the FPS_TEST.mpr file to your 'lfs/data/mpr' folder.
Start LFS and view the replay.. Select the car Thorburn use the in car camera
do this for both Max and Min and report it on the forum

Only check fps at start as soon as cars start to spread out fps will raise

CPU: AMD 64 3200+
MEM: 3 GB DDR 400 (Pared)
GFX: GF 7600 GT (256)
OS: Windows XP (32 bit)

Max: 140
min: 37

EDIT:
Test it without any AA or AF as i did here.. anyways i get 37 if i have 8xAA and 16xAF on aswell
Attached files
FPS_test.zip - 1.1 MB - 103 views
#27 - Jakg
Quote from BigDave2967 :Athlon 35000+ 64 Bit (2.2ghz) 6800GT 256MB AGPX8 2 PC3200 DDR RAM FPS 35-40. With 8sAA Alone: 80FPS with 8sAA

you sure? i would of thought that the 3500 runs at 2.0, seeing as my 3700 defaults to 2.2
Is the "full grid" in question a full set of AI or full online race? I find online I never really drop below 50, but starting 12th in a full AI race I will be anywhere from 39-50 depending on the car, and how close the starting grid is (the lx's seem to get the least fps for me). Note, If I'm mid-pack or at the head of the grid, I get 60fps, (I capped at 60, my monitor refresh rate).

I think the lowest fps depends on 4 things (not in order of magnitude):

1) In Game Settings, AA/AF, LOD, Dust etc...
2) Monitor Resolution
3) CPU
4) GPU

Therefore:
1) 8xAA Temporal and 16xAF, all other settings set to max quality, Max LOD everything in-game
2) 1680x1050 32 bit
3) AMD X2 4800 at 2618MHz (FX-60 speed)
4) ATi X850XT 256MB

I'm being held back at the start a little bit by these 4 things. The graphics settings when turned way down lead to increased FPS. Also my CPU isn't as good as a newer C2D. If my GPU was an 8-series Nvidia, the Resolution and AA/AF wouldn't impact my FPS as much.

Interesting topic. The Four (4) things impacting FPS I believe everyone should list. Unless you only have 128MB of RAM at DDR166 running 5-5-5-15 2T timings, I don't believe its a deciding factor. Opinions welcome,

Stu
Quote from Zyber :Done file is included

...

I'll have a go when I get home tonight and let you know my results.

Edit: My fps should be higher than yours, I guess, as I have a E6300 oc'ed to 2.4 GHz and also a 7600 GT (but the overclocked XFX 7600GT XXX), so it will be interesting to see what results I get.
3500+ (@ 2,55 GHz)
6800 gt (@ ultra)
2gb ram (@ 255 MHz)
no AA, nor AF, lfs settings maxxed
60-70fps at full grid start (online! AI would make my weak cpu struttle like hell..), otherwise mostly at the limit (90fps)

edit: after noticing that even high-end cpu's perform worse on the full grid start, i really have to check my full grid fps again
Here are the results (min and max cfg-files are in the wrong folders btw):

min-settings 8xAA and 16xAF: ca. 140 fps
min-settings no AA and no AF: ca. 180 fps

max-settings 8xAA and 16xAF: ca. 48 fps
max-settings no AA and no AF: ca. 48 (maybe 49) fps

System:
E6300@2.4GHz
XFX 7600GT XXX
2 GB CL 5 RAM 5-5-5-15 @ 343 MHz

Hope that helps .

Edit: btw, the conclusion would be that with max-settings the CPU is the limiting factor as AA and AF settings hardly make any difference. Correct?
...at.. the...at a full grid... :turtle2: ...:snail: ... just 10 fps with shift+N...I know :worried: :tombstone .

nvm. got something mixed up.
System spec:

Intel Core2Duo 6600 (2 x 2.4 GHz)
2048MB DDR2 RAM (800MHz)
512MB DDR3 GeForce 7900 GTO

max: 51 FPS on grid, dropping to a minimum of 44 FPS just after the start
min: 183 FPS on grid, dropping to a minimum of 179 FPS just after the start

Interestingly, I tried the 'max' CFG file with 16xAF, 8xS AA, all settings maxed in the nVidia control panel and I got exactly the same frame rates.
Quote from StewartFisher :Interestingly, I tried the 'max' CFG file with 16xAF, 8xS AA, all settings maxed in the nVidia control panel and I got exactly the same frame rates.

Yeah, as I said, it's the same for me and logically that would point to the conclusion that with max-settings the CPU would be the limiting factor, which seems a little odd, as I would expect the gfx-card to be stressed more when gfx-settings ingame are set to full. Maybe someone with more insight to these things could explain?
#36 - Davo
There's a lot of physics to calculate for 20 cars and it's all done by the CPU so that's the limiting factor. The only thing that would stress the GPU is AA, AF and high res.

Oh yeh I forgot.

I get about 40fps with a full grid, not that it happens often

AMD X2 3800+ @ 2.5ghz
Nvidia 7900GS
All lfs settings maxed - 8xSAA, 16xAF, 1680x1050@60Hz
Quote from Linsen :Here are the results (min and max cfg-files are in the wrong folders btw):

min-settings 8xAA and 16xAF: ca. 140 fps
min-settings no AA and no AF: ca. 180 fps

max-settings 8xAA and 16xAF: ca. 48 fps
max-settings no AA and no AF: ca. 48 (maybe 49) fps

System:
E6300@2.4GHz
XFX 7600GT XXX
2 GB CL 5 RAM 5-5-5-15 @ 343 MHz

Hope that helps .

Edit: btw, the conclusion would be that with max-settings the CPU is the limiting factor as AA and AF settings hardly make any difference. Correct?

hehe thats exacly what i meen.. AA AF doesnt have any impact on fps unless ur alone on track.. anyways.. your cpu is ALOT faster than mine.. and u get 48 fps.. im around 40

my cpu is a old amd 64.. no x2 or dual core just plain old x64.. accually so old it isnt sold anymore.. its a 2 ghz..

i still think something is wrong here.. do u have a G15 keyboard? if so how much is your cpu load running lfs? mine is 75% wich tells me that my cpu isnt even being used 100 %
You lucky DUcK's :P i got 15 fps on a full grid, 50 in hotlapping (single player).
I rarely see over 30 in a race, with more than 8 cars.
That's with low graphics and evrything
Quote from Zyber :hehe thats exacly what i meen.. AA AF doesnt have any impact on fps unless ur alone on track.. anyways.. your cpu is ALOT faster than mine.. and u get 48 fps.. im around 40

my cpu is a old amd 64.. no x2 or dual core just plain old x64.. accually so old it isnt sold anymore.. its a 2 ghz..

i still think something is wrong here.. do u have a G15 keyboard? if so how much is your cpu load running lfs? mine is 75% wich tells me that my cpu isnt even being used 100 %

I thought the 3200+ ran at 2.2 GHz, which would sort of explain the small difference, as there's really not much of an advantage when using a dual core processor, since LFS (like almost all other games) does not support it. So my E6300 should be fairly comparable to a single core cpu running at 2.4 GHz.

@Davo: Wouldn't the physics to be calculated be the same no matter what gfx-settings you use? It's always 20 cars, I thought. Or does LFS change physics calculation with lower (ingame) gfx-settings?

Edit: No, I'm not using a G15, just a simple MS keyborard. As for CPU-load, I don't know, I'd have to check when I'm at home.
#40 - Davo
Even with a dual core LFS only utilises one core so you can't directly compare the two. Limiting frame rate helps with cpu utilisation so that it's not pumping out data to the graphics card all the time and can do more physics calculations.

As I understand it, the physics are calcualted per car since they're all doing something different. In patch V the car infront on the start grid is calculated in hi res which would use even more cpu. I don't think you can change physics calculations otherwise the game would play different, I'm only basing this off my own experience and what I've read on the forum.
I get around 40-50fps with the full grid at 1920x1200 4xAA 8xAF and adaptive anti-aliasing enabled.

E6400 @ 3.0GHz
X1900XT
2GB ddr2-800
Athlon64 4000+
X1650 Pro
512 DDR400 (I know my systems needs more )
80GB S-ATA

Offline i have like 100fps, online around 60fps, with a full grid never lower than 45fps
we need dual core support
i got a dell xps all custom 3.4 p4 ht with 2gigs of ram and a 10,000 rpm hard drive and a x1650 agp (512) and the fps drop to 10 to 24 full 19 odd cars being the last car i dont think the cpu maxes and i know the card is fine b/c normal any were else after start is around 70 -80 some time as high as 120 -130

dual core and hyper threading support is what we need
After looking at this thread I soooo need a new cpu least 939s are cheap now
Pentium 4 1.6 GHz
nVidia GeForce 2 MX100/200
5-10 FPS with full grid

Yeah, my comp sucks.
2.0Ghz Intel Core 2 Duo
2 GB RAM
ATi Mobility Radeon X1400 (here comes something I didn't know) 900MB!
Dx10 baby.

25 FPS at 1440x900 in windowed mode with full set of AI in FE Green.
Highest settings, no AA or AF (but who needs it anyway on a laptop)?

Online I achieve well over 100 in full Autox tracks so I think I'm good for now.
#48 - Jakg
Quote from gotspeed :we need dual core support
i got a dell xps all custom 3.4 p4 ht with 2gigs of ram and a 10,000 rpm hard drive and a x1650 agp (512) and the fps drop to 10 to 24 full 19 odd cars being the last car i dont think the cpu maxes and i know the card is fine b/c normal any were else after start is around 70 -80 some time as high as 120 -130

dual core and hyper threading support is what we need

HT support is pointless seeing as out of all the CPU's, only Presshots had it, and the "AMD Takeup" in LFS is a little higher than in other games (imo - it seems more Geeks play LFS, and they tend to have AMD's because they dominated Presshots at the time).

Multi-Core support would be useful, especially in the future, and i'm hoping we get it for S3
Jakg,

3500+ has default 2.2Ghz.Your CPu has as well 2.2Ghz but rated at 3700+ because you have 1MB L2 cache.

3200+ which is mine CPU runs at default 2Ghz.
I get 30 FPS +-3FPS when there is FULL grid of 32 players.Its almost at the limit where the gameplay is smooth so I guess why now there are a lot of crashes at T1 because with slower CPU it must be a pain.

I have 7900GT 8xAA+16AF and no matter if I change resolution the FPS is still same.I get same FPS testing with 6600 running at 400/800Mhz and 2xAA+16AF.

I think people dont compare here apples to apples as some mean full grid like 23 players some more.

Also LFS is not anymore low PC friendly as it was if you play on servers with more then 25drivers.LFS is now heavy CPU limited game.
Intel 640 3.2 ghz (single core)
1 gig of corsair XMS ddr400
ati x850 XT PE

Max graphics inside LFS outside High quality AF at 8x
AA at 4x + some high res level textures.

Game sits at 99 fps dropping down to 60 min at any point but a full grid My fps can drop down to 27/25 fps the lowest ive seen as we close up into t1. Soons t1's out the way the fps gradually start climbing up to 99.

FGED GREDG RDFGDR GSFDG