Battlefield Bad Company 2 - Another Console port (from PC perspective)

[Cigar]

First off, I hope you read this entirely and not just skim over it. Second off,
I was not out to prove anything before researching this. I just wanted to see
the results. From the results, I have drawn a conclusion.

My Rig:
Phenom x4 9850 (64 bit) overclocked to 3.0 ghz
4 Gigs DDR2 1066mhz (under-clocked to 800mhz, running at 128 bit instead of 64 bit)
Samsung 7200rpm 500gb HDD - C: Windows installed on this drive
Samsung 7200rpm 500gb HDD - D: Battlefield Bad Company 2 installed on this drive
Samsung 7200rpm 500gb HDD - E: Windows page file on this drive
ATI Radeon 4870 1gb DDR5 - 785mhz core clock; 925mhz mem clock; 3700mhz mem speed
(750mhz core 900mhz mem clocks un-overclocked)
ASUS M3A79T-Deluxe Motherboard
24" Sceptre NAGA LCD. 2ms response time. Nat.Res. 1920x1200 (my normal monitor)
18" Microtec 815C LCD. 30ms response time. Nat. Res. 1280x1024 (used in one test)



Operating system and processes:
Windows Vista 64 Ultimate
41 processes running (without any non essential programs running)
5 of those are my G15 keyboard, 2 belong to my G9 mouse, 2 are my video card drivers,
2 are video card overclocking/monitoring processes and one is my HD surround sound drivers.

Drivers:
Newer (most stable) Video Card drivers and MoBo bios.

CPU/Mem Usage:
CPU usage: 0% Physical Memory (ram usage): 25%
Running Battlefield Bad Company 2 in 1920x1200 resolution
CPU usage: 75% Physical Memory (ram usage): 47%



The Data:

For the first test (1280x1024), I disconnected my normal 24 inch monitor and connected
an old 18 inch monitor.

All tests were done with the desktop resolution matching the in-game resolution.
I used medium settings, with 1x Anti-Aliasing and 1x Anisotropic filtering. (I personally don't use anti-aliasing, ever, but I normally have my Anisotropic Filtering at 4x or higher to help in long range sniping) Vertical sync and HBAO were off.

I recorded the following by loading into battlefield with FRAPS running. I ran around, keeping my view on the battlefield (not at the sky or ground). I made sure to get into heavy combat, stand near explosions, drive several vehicles and fire my weapon often. I paid attention to the FPS shown by fraps, and recorded the highest and lowest FPS as I played (often getting me killed).

My findings were as follows:


Resolution Settings FPS Battle FPS (with smoke, taking fire, in vehicles)
1280x1024 Medium AA1x AF1x 60-83 fps 64
1440x900 Medium AA1x AF1x 54-84 fps 54
1680x1050 Medium AA1X AF1X 53-67 fps 55
1920x1080 Medium AA1X AF1x 53-80 fps 57
1920x1200 Medium AA1x AF1x 43-73 fps 43

A little number crunching:

1280x1024 10.9% increase in FPS; 58% decrease in resolution size compaired to 1920x1080
1440x900 15.6% DECREASE in FPS; 1.1% DECREASE in resolution size from 1280x1024
1680x1050 1.8% increase in FPS; 26.5% increase in resolution from 1440x900
=========
1920x1080 (1080p HD) 3.6% increase in FPS; 17.6% increase in resolution from 1680x1050


=========
1920x1200 24.6% DECREASE in FPS; 11.1% increase in resolution from 1920x1080

My conclusions:

.

If you look at the data starting at 1440x900, you will see as I went up in resolution, the
performance increased. That is exactly the opposite of what would happen on game designed
for the PC. As you increase resolution and increase graphical settings, performance on games
designed for PC goes down, not up. If you look at the data, you will see that as the resolution
increase towards 1080p, FPS goes up as well. There was also less Vertical Sync tearing.

1080p (1920x1080) was resolution in which game was at it's best:
no vertical tears (which can be corrected by vsync), a high resolution and good
frame rate (close to 60fps, the maximum for normal human eyes).
Going up or down in resolution from 1080p caused a decrease in performance. The most
drastic of these was shown by the test in 1920x1200, my monitor's natural resolution,
where the FPS dropped a whopping 24.6% after only increasing the screen size 11%.

The only thing I can come up with to explain that performance drop is that the game was designed
for 1080p, which is High Definition. All Next-Gen consoles (xbox360, PS3) are themselves also
designed to output at 1080p.

Now, I must point out, that the _best_ frame rate and overall performance was
in 1280x1024. However I must point out, as I have shown previously, that 1280x1024 is a 58% decrease

in resolution from 1080p, while the gain in FPS was only 10.9%. However, the majority of PC gamers use 1280x1024 (21% of all steam users use it, out of 12 resolutions sampled -
http://en.wikipedia.org/wiki/Display_resolution ). They simply optimized the
code to run in 1280x1024 as well. That is all fine and dandy, but advanced and expert PC
gamers don't spend 2,000 dollars building a machine (3-5 grand from a company like dell)
to run games in 1280x1024. While there are far more casual schlubs out there who buy 700
dollar machines and try to game on them than there are those who spend several thousand
on gaming PCs, all the gamers I know who have an passion for PC gaming do spend more.

I therefor conclude:

This game IS a port from a console designed game.

Without optimizing the code for Gaming-Grade computer resolutions, and by their action of optimizing

it for HD-TVs, the facts point to this game being a port.

 

[Irridium]

The engine Bad Company 2 is using was specifically designed for consoles.
It was explained in the Bad Company 1 dev diaries. They set out to make an engine tailor made for the 360/PS3.

So of course its going to feel like a port, because it is.

That is why its optimized oddly compared to other PC games.

This is probably why they are having a beta for it, to see how well it works and what they need to fix.

 

[Swaki]

wow, much of this was just gibberish in form of numbers and data for me, but i got the last part, but since i have limited space in my apartment im changing my monitor out with a hd tv tomorrow (atm i got my consoles hooked up to the monitor and its not pretty) so the problem seems to become moot in my case, i do however no think that just because its not my problem that its not a problem.

battlefield really tried sucking up to the pc community once some rather unpleasant news about CoD was revealed so for them to compensate on their coding and really how the game is going to be viewed is crazy, thats like shooting yourself in the foot with a really advanced gun.

 

[Cigar]

Actually, the developers have told us many times that the game is _not_ a port from console to PC, as modern warfare was, but was designed for PC along side console, which is a bold lie.

 

[RyePunk]

Uhm, you are basing all those wonderful figures and data upon a public beta build.
Yea good job. Come back in march when the game is released.
I'm not apologizing for the sorry state of the game (which I dont see as being much advanced beyond BF2) but honestly unless you have the final product you can't bash the engine the game is running on.
They've got a month of optimizations to make, and god knows how far past the beta build they already are?
Give them a chance. If it sucks in March I'll bash them with ya.

 

[TZer0]

You've got it wrong.

First of all, testing environments might have been different. More/less players, more items on the ground, more trees, less trees etc. Conditions internally in the PC might also have been different, maybe your AV went on a small stroll through the system or something else started lagging the system.

The next thing.. why did you disconnect your 24" to run the game in 1280x1024?

Besides, I don't know if FRAPS actually might have more of an effect when you run on higher resolutions, there's more data to store.
Edit: disregard that, I thought you were recording.

When it comes to that FPS-drop on the two last resolutions, I've experienced that myself in TF2. 1680x1050 vs. 1400x1050. First one runs bad, stutters, laggs, second ones runs just fine. 1400x1050 nice online and offline. Tested in sterile environments (local server, no players), multiple times in succession.

 

[silverbullet1989]

i dont care if it is a port or not, at least they have taken the time to impliment dedi servers, and not a cheap console match making system. How ever i cannot really comment on this as my bloody beta key has not turned up yet and i can only play the console version, which is good but im hoping its better on the pc =/

 

[Sixties Spidey]

It is a bit sad that a franchise (battlefield, not just Bad Company) that got it's start on the PC has went for the porting option, but they still have time to redeem themselves.

Well, maybe it has something to do with time limitations? Revamping an engine like the Frostbite engine to include all of those new graphical extras in them would have to have taken up a large portion of the development cycle along with bug-testing the engine, developing the game and the multiplayer, bug-testing the game, etc.

It is to my understanding that devs port their games to consoles if they want to bring it to more platforms, but lack the time allotted to them, so maybe that's the reason.

 

[Liberaliter]

Well it was a beta... the point of a beta is to iron things out for when the game goes live to the mass public.

 

[Eggsnham]

It was made to be like that, the only reason it's for the PC is because all the PC gamers felt that they were missing out on the first one (don't know why, though. It wasn't THAT great).

 

[chances]

Guys im new to the pc gaming set-ups,,, so please bear with me.... i downloaded bf2 on steam but when the game loads to main menu the mouse is not in sync with the tabs.... ive tried reloading it but still this happens,....any ideas on who i can fix this as im stressed to my tits lol ..... your help would be much appreciate tks

 

[Scruffy]

As already said by others your test method wasn't best for comparison so I guess different scenes had the most impact on your results. Apart from that, how should one even optimize for a specific resolution? They'd have to cut somewhere else, a bigger resolution will just need more power.

Some other thoughts that arise:
- Why would they optimize for 1080p when the game won't run on that resolution on a console? Maybe it's 720p or even lower that is rendered, nearly no game really does 1080p.
- How does the FOV differ on the different aspect ratios?
- What about LODs making a comparison harder? Especially together with a different FOV.

 

[mawt]

This is the biggest load of crap I've read in years. Lets put an end to it now.

Almost all AAA studio games are written on top of cross-platform middleware and engines. Just because an engine debutes on a console doesn't mean the game is any less PC. You have to bare in mind that the vast majority of games are prototyped on a PC as a PC game product, and then differentiated from then on - the end game might be console only, but it was still a PC-developed game.

What counts today is kind of resources a developer puts into individual platforms. The PC tends to lack focus today because its volume of sales is much lower than that of a console - there are fewer PC gamers out there. I'm not going to say why or even allude to "casual gaming" because everyone's been wrong about everything for the last 10 years (there are 80 million FarmVille players *right now* - who said PC gaming was dead?).

A clear "console port" is where a studio gives the game to another shop to port and/or tune for a specific platform. In most cases they redesign the input and menu system or code their own off a more abstract design contract. Whether a PC game is a "console port" or not depends on many more factors than "it used to be on the console". Is COD a console game since COD3? No. You want to look at the effort put into supporting larger ranges of hardware - is audio offloading used where there's a high performance soundcard, can I specify shader, LOD or dithering options to increase performance, is it moddable? How else have to directly considered the PC in the fundamental design?

The BC2 PC interface was done inhouse. It's as bad as BF1 was - remember that full screen FMV with the godawful crash-prone menu on top? - and no better than most other games out there through the years (most early 3D menu systems used more FPS than the game itself).

The stats at the top, for different resolutions are a load of ignorant bullshit. Lets start by thinking about what a resolution is? A resolution is just a set of numbers that define a screen area - it's quite misleading because it has nothing to do with an actual "resolution", which would involve determining the smallest distinguishable measurement of a given distance or time and using that to determine a relative ratio.

What the resolution doesn't tell us is how hard (or not so hard) a GFX chip is working and what it is doing. Just because, get this, a resolution doesn't scale linearly (or even positively) with performance doesn't mean a game isn't a "PC game" or is any more obviously a PC port.

Because different resolutions mean different things to different cards and families and different implementations of different APIs/middleware, I can't really explain everything here, but ill answer the two common quesetions.

Why does game performance drop when I increase my resolution? There are more pixels on the screen, the graphics card has to do more work to render each pixel and the computer has to do more work shifting the data around?

Why does the fps not seem very consistent at different resolutions? Alignment and powers. A modern computer relies on data buses of specific widths (in bits and bytes). Usually data n bits long on an n-bit data bus takes 1 cycle to propagate. Data that is n-bits+1 requries two cycles to convey all the data, etc. And then you have assets, such textures at one aspect ratio and many sizes which all require more time to shift - the bigger the resolution of the end product, the bigger the resolution of your assets need to be so to avoid pixelation. Bigger bitmaps require more memory. When they don't easily fit into an internal buffer or structure, they have to be resized, taking time (huge kill with unaligned FBO and BVOs)... and lots more reasons along that lines.

So why does performance get better at a higher res? That could be your drivers, that the internal hardware of your graphics card deals with that size/alignment better or it could be the neat engine has started to cull the very small things you can't see - reducing time for rasterization and memory bandwidth required (culls & scissors) to draw.

So... stop it! I've had enough. Go get your own MSc!

 

[Horticulture]

snip

This is the best first post.

 

[clutch-monkey]

It is a bit sad that a franchise (battlefield, not just Bad Company) that got it's start on the PC has went for the porting option, but they still have time to redeem themselves.

not really... i guess it's where the money is.
god knows the PC community doesn't help.

 

[campdude]

If the demo/beta doesnt preform good... and the developers "promise" better preformance from the actual game... Dont buy that.. its a money making scheme. Can they promise 25% more preformance: the answer is no.

For a PC game to sell well along with the console versions... PC version should contain something that makes it look better on PC vs console.

Back when PC gaming was king... The PC version was always better Graphics than the xbox 1/PS2 version. Now we are gonna be stuck at the 1080 p for a while so ramping up the resolution on games is not the quick fix to make a pc game better.

PC gamers were at HD before they even labeled it "HD".

Im a PC gamer. I only purchase PC games when... either of the below happens:

The Game is Better looking on PC than Console.
The Game is Desgined for PC ground up.
The Game is optimized well using a 3d Engine made for pc.

PC gamers tend to punish companys for releasing crappy console ports.

For instance. Dirt 1 was crappy console port but they redeemed themselves with DIRT 2. (all the dx 11 effects)

I still didnt buy Dirt 2 cause of Dirt 1. But will buy Dirt 3 because they redeemed themselves with Dirt 2.

Console People aren?t so elitist and have no personal rules like that.