Jump to content

How do YOU make IL-2:FB run well??


WWSandMan

Recommended Posts

S!

Call it a tweaking thread... but I'm interested in what folks are doing to get the most out of their hardware, as it pertains to running IL-2:FB.

The tried-and-true baseline for frame-rate measuring (at least it's what the folks at MudMovers use as a test) is to run the "Black Death" track, use the in-game frame counter and record the current, max and min frame rates at the last couple seconds of the track. It ends at 2:35.

To do this, run the track with looping off, view control off (using only the recorded views), time compression off, but keep the inflight messages (if you want to see the elapsed time anyway). Once the track starts, open the console (shift-tab) and type in fps START SHOW then close the console (shift-tab again).

With the base test established, we'll want an idea of how to compare them. Apples to apples, and that kind of thing... Without getting as thorough as Mud Movers did when they were polling the community to submit data for the test, just add a reply with your results and hardware/software configuration.

I'll get the ball rolling, well mostly (I'll edit with my actual numbers from the track)...

::EDIT::

I ran the "Black Death" track three times, with slightly different graphics settings each time. My first run, I also had EAX effects turned on for my Audugy2 sound card. Turning this off made zero difference in performance on a second run with only that change made.

Here's my original graphics options settings:

Visibility Distance = high
Objects Lighting = high
Objects Detail = excellent
Landscape Lighting = excellent
Landscape Detail = excellent
Clouds Detail = detailed clouds

I received an FPS score at 2:35 (end of track) of:

Average: 17
Maximum: 33
Minimum: 3

By backing down a few of those (Objects Detail to "normal"; Landscape Lighting to "high" and Landscape Detail to "high"), I got the following:

Average: 23
Maximum: 36
Minimum: 9

I got the worst drop in FPS at about 1:23 or so, when the view changes to the IL-2 attacking the airfield. It's a complete change in perspective/scenery at that point, and is why -I'm sure- the FPS dropped so much.

Note that dumbing down the eye candy didn't appear to help much. I'll try again with Anti-Aliasing on my vid card set to "application controlled", it's currently set at 4x. (I expect significant positive changes in FPS with that change)

::/EDIT::


Attached are the setup screens as I've got them, both for FB and my video card. (I don't expect everyone to post pictures).

My system basics:

Pentium 4 2.26gHz (running at 2.31gHz)
Gigabyte GA-81EXP mainboard with Intel 845E chipset, 533mHz FSB
512mb PC2100 DDR
160gb Western Digital HDD with 8mb cache
40gb Maxtor HDD with 2mb cache
PNY GeForce 4 Ti-4400, 128mb DDR (NVidia 52.16 drivers) (not over-clocked at all)
Sound Blaster Audigy2 (basic)
(an ancient Dell 1028L 17" CRT monitor)
Win XP Pro
*edit:* DirectX 9.0b

I run IL-2:FB under OpenGL at 1024x768 @32bit color depth, usually with all "excellent" settings in the game's video setup.

Edited by Guest
Link to comment
Share on other sites

So you have >fps START SHOW on for 2nd and 3rd BD runthroughs, or start fps quickly each time it runs?


First thing: leave off the ">" symbol, that's only to access the command from the chat window. In the console (you'll know you've accessed the console when your screen fills with a bunch of semi-tranparent text) use only the "fps START SHOW" command (case sensitive!).

I run a quick QMB mission to turn off the FPS counter after a run of the Black Death track, else it continues counting frames and frame-rate for each track played... counting zero F/R for the downtime, etc., during mission load (which messes up the accuracy). So... Yeah, close the FPS counter (in another mission or however you want) then re-start the FPS counter as soon as the Black Death track starts playing, for consecutive runs of the track after making any changes to your system setup. Edited by Guest
Link to comment
Share on other sites

S! Sandy,

What are your Direct X sound settings?


Umm... from the IL-2 setup program?? It's in the "il2_setup1.jpg" picture above.

Heck... here's the sound section of my config.ini file...

[sound]
SoundUse=1
SoundEngine=1
Speakers=1
Placement=0
SoundFlags.reversestereo=0
RadioFlags.Enabled=0
RadioEngine=2
MusicVolume=0
ObjectVolume=4
MusState.takeoff=0
MusState.inflight=0
MusState.crash=0
MusFlags.play=0
MasterVolume=6
Attenuation=5
SoundMode=2
SamplingRate=2
NumChannels=2
SoundExt.occlusions=1
SoundFlags.hardware=1
SoundFlags.streams=1
SoundFlags.duplex=1
SoundExt.acoustics=1
SoundExt.volumefx=1
SoundFlags.voicemgr=0
SoundFlags.static=1
VoiceVolume=0
Channels=2
SoundExt.extrender=0
SoundFlags.bugscorrect=0
SoundSetupId=9
ActivationLevel=0.02
Preemphasis=0.85
RadioLatency=0.5
AGC=1
PTTMode=1
SoundFlags.UseRadioChatter=0
SoundFlags.AutoActivation=0
RadioFlags.PTTMode=0
RadioFlags.PlayClicks=0
ActLevel=0
MicLevel=0

Link to comment
Share on other sites

No, in your direct X diag program. Sensei gave me a tip that I never would have suspected, and added a 50% + increase in my FR's. The game was nearly unplayable over towns for me even in SP. Changing sound accelleration from full to standard within direct X made all the difference I needed so I didn't have to hurl my machine out the window.

Link to comment
Share on other sites

Okay... made a couple changes to my setup, including dropping the DX sound acceleration from "full" to "standard".

Other changes: Changed my Anti-aliasing from "4x" to "application controlled" and changed the "performance" slider from "balanced" <?? -the middle slot> to "performance". (see the picture in the post above with my video card's settings).

Granted, now that I think of it, I should have run with only the sound change to see what that did... but, the game sounded -to my ears- as good as it did before. The difference in FPS with all the changes above, running the Black Death track was like this:

Average: 29
Maximum: 67
Minimum: 12

Those numbers represent a big spike in maximum FPS (due to the dropping of anti-aliasing mostly, I'm sure)... and average FPS rose by... my math is probably wrong, but I come up with about 24%. A substantial increase...

My in-game video settings were (from the top down):

high
high
normal
high
high
detailed clouds

Link to comment
Share on other sites

Things I've done...

- With Creative based sound cards use the Standard acceleration in DX
- Use Quadrophonic speaker setting instead of headphones
- Disable ansiotropic filtering on ATI cards. It nearly halves the FPS while adding almost zero graphical difference. Trilinear is fine.
- Most 64 meg video cards run 32 bit faster than they run 16 bit. I don't know why. There is a long article at Tom's hardware that explained it almost.

Other settings I have that affect framerates (probably lowering them).

- I run at 6X anti-aliasing. On my 9700 Pro the visual difference at 1024x768x32 is significant while the FPS hit is almost nil (1-2 fps).
- Vertical sync on (The tearing in IL2 with this off is horrible especially with TIR. However, with it off you can gain several fps)

My OpenGL settings from the config menu:

Texture mapping - Trilinear
Texture Compression - None (Oleg's code for handling this is horrible)
Use Vertex Arrays is checked.
Under Extensions:
Multitexture
Combine
Secondary Color
Vertex Array Extension
Clip Hint

are on.

If you use Anisotropic filtering then you have to select Texture Anisotropic Extension. If you use a compression then you have to select Texture Compress ARB Extension.

Sound settings: Horrible bugs with 32 channel playback. Use 16 channel. This has been acknowledged by Oleg. I use a Balanced 3D engine and 44100 sampling rate. Going to play with those since I still don't like my sound...

Link to comment
Share on other sites

Okay... made a couple changes to my setup, including dropping the DX sound acceleration from "full" to "standard".

Other changes: Changed my Anti-aliasing from "4x" to "application controlled" and changed the "performance" slider from "balanced" <?? -the middle slot> to "performance". (see the picture in the post above with my video card's settings).

Granted, now that I think of it, I should have run with only the sound change to see what that did... but, the game sounded -to my ears- as good as it did before. The difference in FPS with all the changes above, running the Black Death track was like this:

Average: 29
Maximum: 67
Minimum: 12

Those numbers represent a big spike in maximum FPS (due to the dropping of anti-aliasing mostly, I'm sure)... and average FPS rose by... my math is probably wrong, but I come up with about 24%. A substantial increase...

My in-game video settings were (from the top down):

high
high
normal
high
high
detailed clouds



I think had you just done the DX change you'd have been startled at the change. I know it makes sense that the vid changes would contribute to your FR increase, but I'll bet most if not all was due to ramping down DX.
Link to comment
Share on other sites

System specs
AMD XP2100+ (Running at 2.08Gig or as XP2600+) [198fsbx10.5 multiplier at 1.625v]
Asus A7N8X Deluxe FSB400
512mb PC3200 DDR 400FSB Corsair RAM (Two 256mb sticks)
60gb Western Digital HDD with 4mb cache
eVGA e-Geforce FX5900 w45.23 Drivers (more on that later) not overclocked. Runs at 450Mhz/850Mhz (Core/Mem)
On board sound (nForce2) Better then Audigy1 or 2 hands down!
NEC950 19” Monitor
Win 98SE
DirectX 9.0b

In Game setup
Visibility Distance = high
Objects Lighting = high
Objects Detail = excellent
Landscape Lighting = excellent
Landscape Detail = excellent
Clouds Detail = detailed clouds


FB Video Setup (external progie)
OpenGL: Custom
Texture mipmap filter: trilinear
Texture Compression: None
Use Vertex Arrays: Checked
Multitexture: Checked
Combine: Checked
Secondary Color: Checked
Vertex Array Extention: Checked
Clip Hint: Checked


FB Sound Setup (external progie)
Playback channels: 16
Speaker type: Headphones
3D engine mode: Balanced
Sampling Rate: 22050
Audio enhancements and hardware acceleration: Checked

Note that I got the Geforce FX5900, the deal was just too good. I could have gotten ATI 9700 plain jane or the FX5900 plane jane and opted for the 5900 as it would (and has proven at least to me) that it would be slightly faster then a 9700 Pro. I could have gotten the ATI 9700 for $240 and the FX5900 for $230, economics and the possibility of a slightly faster card won me over to the Geforce.

I tested three driver sets for the FX5900; 44.03, 45.23, and the 52.16. First off let me start by saying that the 52.16 drivers sucks so bad it nearly made me throw up! Lots of bugs with that one at least for the 98SE version, I struggled with it and finally gave up. Perhaps the XP version is better but I doubt it.

Some benchmarking first to get a general ‘Feel’ as to what I might expect, using 3dMark03 330 I got these scores:

44.03 – 5084
45.23 – 5386
52.16 – 5538

These scores look like there is quiet an improvement from 44.03 to 52.16, but I think this test (and others I’ve done for benchmarking) are meaningless. As you’ll see from my testing below:

I tested two resolutions, 1024x768x32 to match yours Sandy and 1280x960x32 to get (IMHO) the best “Bang” for the buck if you will from the power of the FX5900.

I do NOT use any anisotropic filtering. I found with the FX5900 it was really pretty pointless. Also I never used it for my GF3 when it was working because it was just a card killer for fps. Btw, using the above settings I gave for the external progie setup, I get VERY close to anisotropic smoke/shadows etc anyway, so why bother!

No FSAA 1024x768x32 Done in Average/Max/Min
44.03 – 35/70/8
45.23 – 40/60/12
52.16 – 39/83/11

Here they look pretty even, though the 52.16 driver has a higher peek, it actually scores in the middle for average, also note that the 45.23 drivers seem more “consistent” for frames, even on the low end.

FSAAx4 1024x768x32
44.03 – 32/64/11
45.23 – 37/54/16
52.16 – 30/57/12

Here the 45.23 drivers really seem to shine compared to the other two. Higher average, although the peek is lower then the other two, I’d rather it was ‘Consistent’, and the minimum frames were better then without FSAA. This was something I noticed with all the tests I did, the minimum frames were ALWAYS higher with FSAA then without. Guess the FX cards REALLY like to use FSAA! Also look at the hit I took between no FSAA and FSAAx4 with the 52.16 drivers vs the other two drivers!

Now I tried all kinds of combos on the video (in game and external setup progie) as well as the sound and found that anything other then what I listed above Sandy cost me 1-5! Fps. Sound adjustments made no difference to performance on my end, but I would expect that with the nForce2 sound setup. So Chucks advice to you and your Audigy may very well have a big impact for you. I know my SBLive really seemed to hate IL2 and then FB, not in quality, but in stutter effect. You may also want to look into your driver Sandy. The 52.xx are really for the FX (so is the 45’s for that matter), but recently Oleg did post that the game does BEST with 45.33 (they don’t make that driver for 98SE, but they do for the XP) so I would investigate the driver too! Look at the difference it made in my tests and benchmarks be buggered! I’m using the 45.23 because for me they are the best and visually they even look better then 52.16 and 44.03! Plus I had mucho problems with 52.16 for stability and reliability etc, many programs just didn’t work with 52.16.

Now here are my scores for all drivers but at 1280x960x32

No FSAA
44.03 – 35/79/8
45.23 – 37/54/9
52.16 – 37/79/13

Higher peeks for the 52.16, but same fps average for 45.23 and 52.16, surprisingly the minimum frames seem to be better for the 52.16.

FSAAx4
44.03 – 30/63/11
45.23 – 32/54/12
52.16 – 29/58/8

Here I think something broke for 52.16 they just plane sucked, but again the 45.23 drivers did very well and lost hardly anything going to FSAA, hell although the average is lower, the minimum was higher! The low end score for the 52.16 was the only time in all the testing where it actually lost ground going to FSAA. Just another indication (for me) that something was screwy with the 52.16 drivers.

Btw, just for kicks, I did a test with at 1600x1200x32x4FSAA and got these numbers:
29/59/11!!! I will however use 1280x960x32x4FSAA because the aircraft dots at distance just get WAY to tiny to see well. LOL Though it looks gorgeous.

I also tested using perfect settings, and can do it at 1280x960x32x4FSAA but the average drops to 21fps and the minimum drops to around 8 which is a real noticeable occurrence. Just not worth it in my book for pretty water. I’ll still with high/excellent settings.

I also ran the game at x8FSAA and although like perfect it is incredible looking, the average was around 18 and the minimums were around a 4-5. Real pretty and an indication of where the industry is heading, but I need a faster processor to do it justice. LOL

Anyway I hope some of this helps Sandy, consider trying the other drivers (especially 45.23/45.33) since Oleg has said they work best with FB and my tests seem to bear out his claim!

Link to comment
Share on other sites

Author: Oleg_Maddox
Rank: Creator of IL-2 Sturmovik
Date: 10/30/03 08:15AM





Il-2:FB list of recommended drivers

1C:Maddox Games recommends the following video card drivers to be used with IL-2:Forgotten Battles. The sim may not function properly with other driver versions due to issues outside our control.



NVIDIA
Display Driver Version: 45.33
Release Date: October 9, 2003
WHQL Certified

http://www.nvidia.com/object/pc_winxp-2k_45.33.html


ATI

CATALYST™ 3.7
Driver Build 6.14.10.63787-93-030812a1-010735c


Sorry, we can't fix the things that are not our own.


Oleg Maddox

:shock:

Link to comment
Share on other sites

Author: B16Enk
Rank: Over 500 Postings
Date: 10/30/03 12:45PM




********CAUTION**************

The NVidia drivers specified (possibly )are for the FX series only!!

Consider not installing if you have GeForce 4 or earlier.

Stick with the 43.25 (same for Win9x/ME).

I tried them with my Ti4200 and ended up with 'SafeVGA' mode

Simply removing the VGA device in Device manager, then running 45.23 installation (before rebooting) solves it...

EDITED because I finally read the readme.

States they are Ok for earlier cards but they wouldn't install for me.
INF file reported (to XP) that the drivers were for GE 5xxx cards only

Apologies for any confusion, I didn't want any other hapless people getting into the same difficulties as me.

Link to comment
Share on other sites

I'm not sure that one post made by the one guy who had problems with 45.23 is accurate.

This is what my .inf shows for 45.23, as you can see it supports all nVidia cards from NV5 to NV35.

[strings]
Disk1Name="NVIDIA Compatible Display Driver Disk 1"
Disk2Name="NVIDIA Compatible Display Driver Disk 2"
Disk3Name="NVIDIA Compatible Display Driver Disk 3"
Disk4Name="NVIDIA Compatible Display Driver Disk 4"
Disk5Name="NVIDIA Compatible Display Driver Disk 5"
Disk6Name="NVIDIA Compatible Display Driver Disk 6"
INFName="nvagp.inf"
NVIDIA&DEV_0028.DeviceDesc="NVIDIA RIVA TNT2/TNT2 Pro"
NVIDIA&DEV_0029.DeviceDesc="NVIDIA RIVA TNT2 Ultra"
NVIDIA&DEV_002C.DeviceDesc="NVIDIA Vanta/Vanta LT"
NVIDIA&DEV_002D.DeviceDesc="NVIDIA RIVA TNT2 Model 64/Model 64 Pro"
NVIDIA&DEV_0100.DeviceDesc="NVIDIA GeForce 256"
NVIDIA&DEV_0101.DeviceDesc="NVIDIA GeForce DDR"
NVIDIA&DEV_0103.DeviceDesc="NVIDIA Quadro"
NVIDIA&DEV_0110.DeviceDesc="NVIDIA GeForce2 MX/MX 400"
NVIDIA&DEV_0111.DeviceDesc="NVIDIA GeForce2 MX 100/200"
NVIDIA&DEV_0113.DeviceDesc="NVIDIA Quadro2 MXR/EX"
NVIDIA&DEV_01A0.DeviceDesc="NVIDIA GeForce2 Integrated GPU"
NVIDIA&DEV_0150.DeviceDesc="NVIDIA GeForce2 GTS/GeForce2 Pro"
NVIDIA&DEV_0151.DeviceDesc="NVIDIA GeForce2 Ti"
NVIDIA&DEV_0152.DeviceDesc="NVIDIA GeForce2 Ultra"
NVIDIA&DEV_0153.DeviceDesc="NVIDIA Quadro2 Pro"
NVIDIA&DEV_0170.DeviceDesc="NVIDIA GeForce4 MX 460"
NVIDIA&DEV_0171.DeviceDesc="NVIDIA GeForce4 MX 440"
NVIDIA&DEV_0172.DeviceDesc="NVIDIA GeForce4 MX 420"
NVIDIA&DEV_0173.DeviceDesc="NVIDIA GeForce4 MX 440-SE"
NVIDIA&DEV_0178.DeviceDesc="NVIDIA Quadro4 550 XGL"
NVIDIA&DEV_017A.DeviceDesc="NVIDIA Quadro NVS"
NVIDIA&DEV_0181.DeviceDesc="NVIDIA GeForce4 MX 440 with AGP8X"
NVIDIA&DEV_0182.DeviceDesc="NVIDIA GeForce4 MX 440SE with AGP8X"
NVIDIA&DEV_0188.DeviceDesc="NVIDIA Quadro4 580 XGL"
NVIDIA&DEV_018A.DeviceDesc="NVIDIA Quadro NVS with AGP8X"
NVIDIA&DEV_018B.DeviceDesc="NVIDIA Quadro4 380 XGL"
NVIDIA&DEV_01F0.DeviceDesc="NVIDIA GeForce4 MX Integrated GPU"
NVIDIA&DEV_0200.DeviceDesc="NVIDIA GeForce3"
NVIDIA&DEV_0201.DeviceDesc="NVIDIA GeForce3 Ti 200"
NVIDIA&DEV_0202.DeviceDesc="NVIDIA GeForce3 Ti 500"
NVIDIA&DEV_0203.DeviceDesc="NVIDIA Quadro DCC"
NVIDIA&DEV_0250.DeviceDesc="NVIDIA GeForce4 Ti 4600"
NVIDIA&DEV_0251.DeviceDesc="NVIDIA GeForce4 Ti 4400"
NVIDIA&DEV_0253.DeviceDesc="NVIDIA GeForce4 Ti 4200"
NVIDIA&DEV_0258.DeviceDesc="NVIDIA Quadro4 900 XGL"
NVIDIA&DEV_0259.DeviceDesc="NVIDIA Quadro4 750 XGL"
NVIDIA&DEV_025B.DeviceDesc="NVIDIA Quadro4 700 XGL"
NVIDIA&DEV_0280.DeviceDesc="NVIDIA GeForce4 Ti 4800"
NVIDIA&DEV_0281.DeviceDesc="NVIDIA GeForce4 Ti 4200 with AGP8X"
NVIDIA&DEV_0282.DeviceDesc="NVIDIA GeForce4 Ti 4800 SE"
NVIDIA&DEV_0288.DeviceDesc="NVIDIA Quadro4 980 XGL"
NVIDIA&DEV_0289.DeviceDesc="NVIDIA Quadro4 780 XGL"
NVIDIA&DEV_0301.DeviceDesc="NVIDIA GeForce FX 5800 Ultra"
NVIDIA&DEV_0302.DeviceDesc="NVIDIA GeForce FX 5800"
NVIDIA&DEV_0308.DeviceDesc="NVIDIA Quadro FX 2000"
NVIDIA&DEV_0309.DeviceDesc="NVIDIA Quadro FX 1000"
NVIDIA&DEV_0311.DeviceDesc="NVIDIA GeForce FX 5600 Ultra"
NVIDIA&DEV_0312.DeviceDesc="NVIDIA GeForce FX 5600"
NVIDIA&DEV_0314.DeviceDesc="NVIDIA GeForce FX 5600SE"
NVIDIA&DEV_0321.DeviceDesc="NVIDIA GeForce FX 5200 Ultra"
NVIDIA&DEV_0322.DeviceDesc="NVIDIA GeForce FX 5200"
NVIDIA&DEV_032B.DeviceDesc="NVIDIA Quadro FX 500"
NVIDIA&DEV_0330.DeviceDesc="NVIDIA GeForce FX 5900 Ultra"
NVIDIA&DEV_0331.DeviceDesc="NVIDIA GeForce FX 5900"
NVIDIA&DEV_0338.DeviceDesc="NVIDIA Quadro FX 3000"

Perhaps he got a bad driver, you can easily check your .inf before hand Sandy, maybe the XP is messed up. I used 45.23 with my GF3 without any problems so I would "imagine' it would work with your GF4 okay. But then again, you have XP.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

Please confirm you have read and understand the rules above

Please Sign In or Sign Up