Just Cause 2 performance showdown
The demo put through its considerable paces on PS3 and Xbox 360.
Just Cause 2 is a game that pushes all the right buttons for us here at Digital Foundry. Proprietary engine? Check. Scandinavian developer? Check. Colossal open world? Ingenius physics-based gameplay? Epoch-making pyrotechnics? Check, check, check!
The arrival last Thursday of demo code for PlayStation 3 and Xbox 360 gives us our first chance to properly evaluate the technology and performance of the respective console versions within the DF lair, away from the compromises of compressed internet video.
Not that Avalanche's various gameplay vids aren't worth checking out, mind you: they effectively showcase the range of gameplay possibilities made possible by the tech.
It's exciting stuff - not just gamers, but for other game-makers too. On a recent DF visit to a developer currently working on a triple-A title, Avalanche's videos were being showcased at a team meeting on a 50" plasma. Interestingly this presentation wasn't about polygon counts, LOD biases or framebuffers. They were checking out the footage because Just Cause 2 looked like fun to play.
Central to the game's appeal is the inclusion of an extremely long grappling hook. You use it to winch yourself onto just about anywhere within the huge environments, or else to effortlessly hijack speeding vehicles.
Its offensive capabilities are impressive too. Sure, you can use it to pull opponents towards you, but the ingenuity of the grapple-gun comes down to its ability to link enemies and objects together. Attaching your opponent to a barrel that's about to explode has its appeal, as does connecting them to a passing vehicle.
Beyond the grappling, the key appeal of this game is its environments, and this is where the technology Avalanche has been working on comes to the fore. The rendering distance of the gameworld is absolutely colossal, with an impressive level of detail retention even in the most faraway areas. Jump into a chopper to reach stratospheric heights and check out the view: this sort of stunt was impressive in the first Just Cause, but it's been taken to a new level in the sequel.
The overall feel of the game is bolstered by an impressive lighting model, with smooth atmospheric effects. Transparency elements such as particles, clouds and explosions look very smooth indeed. Despite the 34 square kilometres of gameplay world available in the demo, water sources within the playable area are hard to locate (the available play area is so vast, we're still looking). How Avalanche has handled this effect in both of the console builds will be interesting to see.
Early impressions are that PlayStation 3 and Xbox 360 owners have been well-served by Avalanche's code-smiths. However, it is curious to note that while overall the similarities between the console games outweigh the differences by quite a margin, Avalanche has chosen two different approaches to the performance aspect of Just Cause 2. The headline news here is that the Xbox 360 version can drop v-sync while the PS3 game does not.
Let's kick off with a performance analysis of the Xbox 360 version.
Just Cause 2 on Xbox 360 caps frame-rate at 30FPS, but does tear when the engine is put under stress.
Few surprises here. Avalanche has capped frame-rate at 30FPS. In order to maintain fluidity and consistency in response, when the game engine struggles, v-sync is lost and torn frames enter the equation. In the course of this video a mere eight per cent of the 60Hz output of the Xbox 360 consists of torn frames. Cut-scenes apart (which both consoles struggle with in their own way), the game only really seems to suffer when big explosions kick-off.
There's no problem at identifying the core attributes of the framebuffer here. Just Cause 2 on Xbox 360 has the full, requisite 1280x720 and implements 2x multisampling anti-aliasing.
Now let's turn the focus to the PlayStation 3 rendition of the demo.
PlayStation 3 Just Cause is a tear-free zone with a uncapped, variable frame-rate.
Here we see a very different FPS graph owing to a change in the rendering attributes of the Avalanche engine running on PlayStation 3. The game maintains v-sync throughout, making for a higher consistency in image quality as you play. Also note that the PS3 game runs without a frame-rate cap while overall maintaining an average just a notch below 30FPS - just like Xbox 360.
It's not the first time we've seen v-sync/tearing differences like this before, of course. Resident Evil 5 had a similar setup, albeit without the unlocked frame-rate. Why Avalanche has gone with two separate approaches like this is curious: it may well have been the case that performance of both platforms was analysed and the best was chosen for each according to how the engine performed on each architecture.
In terms of image quality, Just Cause 2 appears to use 2x quincunx anti-aliasing, though this is not conclusive owing to the fact it appears to be used relatively sparingly - some edges get smoothed, some do not. Getting a good reading is also made more difficult owing to the way bloom is deployed, and there's plenty of it in Just Cause 2.
Another interesting difference concerns the shadow-rendering. Low-resolution shadows with obvious serrated edges close-up can be seen on both versions of the game, though these serrations are a touch smoother and more refined on PlayStation 3. More obvious is that the inclusion of a camera-based motion blur only appears to have been implemented on the Xbox 360 game
Other elements of the engine appear to be pretty much like-for-like on both console platforms. There's fairly well handled geometry LOD popping on trees and shadows - it usually happens so far away that the eye doesn't really pick up on it. Depth-of-field seems to mask foliage as it transitions into a lower resolution alpha texture.
Bearing in mind how colossal this gameworld is, and how much of it you see, perhaps the biggest question mark remaining concerns just how Just Cause 2 plans to stream its data. Mandatory install on PS3? Lower performance running from DVD on Xbox 360? What about load times?
These are the sorts of thing we'll be looking at in the forthcoming Face-Off. We'll also be factoring in the PC version of the game too. Avalanche is looking to offer plenty of goodies for PC owners with NVIDIA graphics cards: there's 3D Vision compatibility, plus promised support for the GPU-specific programming language, CUDA. And Avalanche is promising an impressive battery of visual effects for CUDA-capable GPUs, which we're looking forward to checking out in the final game.
Just Cause 2 is due out for PC, PS3 and Xbox 360 on 26th March.
You may also like...
-
Digital Foundry: PS3 Skyrim Lag Fixed?
-
Face-Off: Final Fantasy 13-2
-
Face-Off: The Darkness 2
-
Eurogamer.net Podcast #100: Ellie returns! And we filmed it!
-
Eurogamer.net Podcast #99: FF13-2 and Amalur RPG Special
-
Warp Review
-
Why Can't Games Do Sex?
-
The Kickstarter Conundrum
-
Assassin's Creed 3 release date announced
-
Remedy's message to Alan Wake PC pirates: "enjoy the story!"
-
Full-length Far Cry 3 cinematic trailer
-
NCsoft confirms Guild Wars 2 on console
-
Dear Esther Review
-
Alan Wake PC version footage
-
Huge range of PlayStation 2 Classics storm European PlayStation store
-
Japan PlayStation Vita sales at lowest ever weekly total
-
Assassin's Creed Revelations getting Desmond single-player DLC
-
Indie game Dear Esther profitable in less than six hours
-
Bethesda on Skyrim's viewable Morrowind, Cyrodiil: "maybe we'll use it one day"
-
Syndicate launch trailer blasts out the dubstep
-
App of the Day: Tongue Tied!
-
Valve selling a virtual Team Fortress 2 ring for $100
-
Notch can match Schafer's $13m Psychonauts 2 budget valuation
-
Far Cry 3 release date revealed by leaked trailer
-
Solitaire Blitz Preview: Why PopCap's Approach to Facebook Gaming is Anything But Casual









Comments (58) Latest comment 2 years ago
Comments threads automatically close after 30 days, but please feel free to continue chatting on the forum!
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
While the PC version obviously looks the best of all the formats and can run at 60 fps with better AA, I thought the PS3 version was nevertheless impressive given the size of the demo, the smoothness of the framerate and that phenonmenal draw distance. It even has AA!!! Sure, some of the textures were a bit iffy and there's noticeable pop up but they're forgiveable in a game that is extremely well-coded and so much fun to play.
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
all max, 8x AA 16x AS 1650x1050 res
edit: oh and no vsync, cos I hate that stuff. (+60ms lag etc, symptom of having a crappy pc before)
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Wait, Digital Foundry play Dwarf Fortress? Get in! Accelerate those 720p ASCIIs
Comment below viewing threshold Show
Comment below viewing threshold Show
Think this game will become to repetitive fairly quick, the AI really seem dumb - to easy!
Comment below viewing threshold Show
Leading on the PS3 means that the split memory architecture is dealt with up front (360's unified pool means its irrelevant), and with Sony steadily chipping away at the OS''s footprint the 360 no longer has the significant space advantage it used to enjoy.
Going forward, if the MLAA techniques employed by Pandemic and SSM become more widespread, the 360's EDRAM/fillrate advantage should be reduced as the PS3 will have its own "cheap" AA solution... which should remove the final barrier to true platformplatform equality.
Comment below viewing threshold Show
The rat tat shooting was poor and there was a delay evertimeI used the machine gun, it seemed to wait for a second before firing.
Maybe got to play it again, I seem to be missing something ?
Comment below viewing threshold Show
Comment below viewing threshold Show
EDIT: imho of course
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
PS3 version: Video record mode.
360 version: No video record mode.
Given the sandbox nature of the game, that's a fair point. It's a very nice feature to have.
Comment below viewing threshold Show
Comment below viewing threshold Show
Amazing game, I'll be messing around with this demo until my pre-order of the full game arrives.
Comment below viewing threshold Show
Well, in any case it seems like it requires a rather beefy system if you want to run it maxed, more so than any other PC game I've tried running on my current PC.
At 1920x1200 on a Vista x64 Core 2 Quad system with 8 GB RAM and a Geforce GTX 260 (core 192), I initially tried maxing everything - plus 4xAA, 16xAF - but I was getting framerates in the mid teens in the first indoor scene (not sure why exactly), and just barely over 20 when I arrived in the initial camp.
Most of the settings didn't seem to have much of an impact on framerate, but I did eventually end up with a playable 30-40 fps, but with a number of settings off or at medium (eventually I didn't try each one at a time, so I never conclusively determined which ones had the biggest impact - I imagine I could perhaps have found a marginally better looking compromise).
Overall though, with so many settings turned down the PC version didn't really strike me as a whole better looking than the PS3 demo I played yesterday, despite the much higher resolution and 4xAA, 16xAF.
I'm going to check out the PS3 demo again though, because I do suspect at least the textures in the PC demo - which I had at very high - might have been a fair bit sharper.
In any case, since I don't have the powerhouse required to run Just Cause 2 at a consistent 40+ fps on PC at mostly high settings (30 fps just seems smoother on console for some reason, possibly because I'm sitting much further from the screen, and additionally using smoother analogue input), there's no question I'll be buying the PS3 version.
With all the vehicle action I know I'll prefer a gamepad anyway, and though that's also an option in the PC version (didn't try it though as my wired 360 controller wasn't hooked up), I suspect I would feel compelled to mix and match pad and mouse/keyboard use, and that did make for a slightly disjointed experience when I did it in GTA4.
Comment below viewing threshold Show
I'm pretty sure I see more detail in the textures on PC, and the PS3 version has an overall slightly blurrier look to it (but with the heavy depth of field use, even the PC version has a touch of this).
I imagine there were probably differences in things such as shadows as well (I noticed the character shadow was just a "blob" on the PS3 for instance, but I can't even remember how it looked on PC), but I think I've proven to myself with this demo as well as Battlefield: Bad Company 2, where I bought both versions, that unless there are framerate or screen tearing problems (both of which can definitely bug me), I really don't pay much attention to any minor graphical differences.
That's not really a big surprise, but with these two games it's actually the first time I've had that suspicion properly confirmed.
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
I have both 360 and PS3 but nice to see the PS3 matching (or beating) 360 finally
Comment below viewing threshold Show
erm.......heavy rain?
@Demiath
I could swear that is the voodoo lady from the monkey island games!
Comment below viewing threshold Show
Goddamn those controls suck in feel.
Comment below viewing threshold Show
Comment below viewing threshold Show
At 1920x1200 on a Vista x64 Core 2 Quad system with 8 GB RAM and a Geforce GTX 260 (core 192), I initially tried maxing everything - plus 4xAA, 16xAF - but I was getting framerates in the mid teens in the first indoor scene (not sure why exactly), and just barely over 20 when I arrived in the initial camp."
Somethings wrong with your PC as it is underperforming greatly. I myself have an E8400@3.6GHz dual-core CPU, single stock ATI HD4890 1GB, Vista x64. Game runs in range of 35-60fps with avg of 40-45fps. Perfectly smooth with 1680x1050, 4xAA + 4xTSAA (transparency AA), 16xAF, max settings, vsync on, ATI Cat AI optimisation disabled. However Nvidia cards feature CUDA and can enable some extra effects due to this like Bokeh DOF effect and improved water effects. This can be a culprit to your perfomance problems before lowering anything else.
Screenshots with framerate.
[link url=http://img528.imageshack.us/img528/1557/justcau se2demoatidx10.jpg
]http://im g528.imageshack.us/img528/1557/...[/link]
[link url=http://img190.imageshack.us/img190/3563/justca use2demoatidx106.jpg
]http://im g190.imageshack.us/img190/3563/...[/link]
[link url=http://img522.imageshack.us/img522/7536/justca use2demoatidx104.jpg
]http://im g522.imageshack.us/img522/7536/...[/link]
[link url=http://img237.imageshack.us/img237/1770/justca use2demoatidx105.jpg
]http://im g237.imageshack.us/img237/1770/...[/link]
[link url=http://img705.imageshack.us/img705/9618/justca use2demoatidx102.jpg
]http://im g705.imageshack.us/img705/9618/...[/link]
Miths wrote: "I imagine there were probably differences in things such as shadows as well (I noticed the character shadow was just a "blob" on the PS3 for instance, but I can't even remember how it looked on PC), but I think I've proven to myself with this demo as well as Battlefield: Bad Company 2, where I bought both versions, that unless there are framerate or screen tearing problems (both of which can definitely bug me), I really don't pay much attention to any minor graphical differences."
Thing is it is not minor differences which makes me question if you played PC version at all?
From developers themselvess a feature list for PC.
[link url=http://ve3d.ign.com/articles/features/53469 /Just-Cause-2-PC-Interview
]http://ve 3d.ign.com/articles/features/53...[/link]
"Enhanced DirectX10 and 10.1 render path
Certified for Nvidia 3D Vision (Not available on console)
Enhanced for multi-screen setup (Not available on console)
Xbox360 Controller Support
Enhanced GPU Water effect for Nvidia Systems (Not available on console)
Enhanced Bokeh filter for Nvidia Systems (Not available on console)
Increased detail setting (Not available on console)
Soft Shadows (Not available on console)
Soft Particles (Not available on console)
Point Specular (Not available on console)
SSAO (Not available on console)"
The shadows is one huge difference from high detail soft shadows to low detail shadows with no soft edges and low sampling rate. That is one that stands out really big. SSAO gives depth to the whole scene and it certainly cant be missied. Soft particles makes effects look like one smooth entity instead of being able to see particle texture edges colldiding with each other which makes effects look a lot worse. Point specular gives nicer specular shinyness effect for improved surface detail. Then the Nvidia exclusive effects (CUDA) with very nice effect.
And of course detail LOD is a lot higher for PC version... a lot. Distant objects and detail it maintains, gliding down with parachute the trees/objects are pretty much already highest detail LOD when youre at the top at start point of demo where heli drops you off. On consoles you clearly see how they change detail upto 10m infront of you and yet have less detailed altogether and texture res less!
Dont get me wrong, I am not talking down any version but differences are stark and it needs to be clarified as it is not just "minor" and it certainly is visible to eye everywhere and with stark detail increase. The game is an excellent PC version being well optimised and sporting vastly higher graphics visual IQ. Runs great even on old hardware like 8800GTS/E6600 from mid 2006 with max 4xAA/16xAF at 1440x900 and no SSAO, no CUDA extra features.
The developers deserves thanks for making a good PC version pushing both visuals and perfomance. A refreshment from several other multiplatform games where PC hardware is badly utilised due to very unoptimised version and no use of PC special hardware capabilities.
Comment below viewing threshold Show
360-
30FPS max
Drops V-Sync to maintain 30FPS so frames torn
2 x MSAA
Motion blur
PS3-
Variable frame rate
V-Synched throughout
2 x QSAA (In effect 4 x MSAA) Blurred textures and used selectively
Smoother shadows
This is much better. I'm glad to see such multi plat games with no obvious advantages and it's really down to the player to decide which version they'd rather have instead of being forced to pick one due to the obvious inferiority of the other (Bayonetta for example).
Comment below viewing threshold Show
Comment below viewing threshold Show
V-sync Off = increases performace but sacrifices image quality. From an hardware perspective it is less demanding.
V-sync On = decreases performace but gives better image quality. From an hardware perspective it is more demanding.
The difference in performace when it comes to frames per second between On/Off can be around something like 30% or more or less depending on the hardware performance.
/ Ken
Comment below viewing threshold Show
If it can look this good already then why bother going the extra mile on multi-platform games?
Comment below viewing threshold Show
I haven't seen any indications of unusually low framerates in other PC games I've played recently (and I'm starting to split my time at least evenly between PC and PS3 again, after two years with 95% console gaming, so it's been a fair few), which includes Far Cry 2, Battlefield: BC2, Mass Effect 2, iRacing, Age of Conan (the last one still too performance erratic in DX10 mode though, so I've been sticking to DX9) and several others.
I haven't upgraded to the new Nvidia driver that was just released a few days ago though, so perhaps there are some optimizations for Just Cause 2 there. The disappointing performance level in the demo definitely did strike me as a bit odd.
Comment below viewing threshold Show
I love the dicking around aspect of a lot of games, such as GTA, but in you get rewarded for it! And the parachute and sheer scale of things gives you a great sense of freedom. Love the grapple too.
Unfortunately the shooting feels a bit clumsy and inaccurate, and the terrain is a little samey. I would also have liked (X360 pad) the guns/grenades to have been on the LB and RB (due to their seemingly secondary role in the game) and the sprint/grapple controls (which I tend to use all the time) on the LT and RT.
Great demo though. Hopefully the complete game will have good progression and maintains its freshness all the way through.
Oh and on a performance showdown note, although I noticed the tearing on the X360 version I really didn't notice it once when actually playing the game. Its just too fun!
Comment below viewing threshold Show
Your 'V-sync Off' vs. 'V-sync On' analysis is meaningless unless there is a target framerate; you are correct if they were both trying to be 30fps, 'V-sync Off' vs. 'V-sync On' is only less demanding if both are trying to match a particular framerate.
However, 'V-sync Off' for 30fps is likely more demanding than 'V-sync On' for 20fps.
In truth a better summary is:
V-sync Off = torn frames allowed.
V-sync On = no torn frames allowed.
Comment below viewing threshold Show
Comment below viewing threshold Show
Mostly to satisfy my own curiosity, I just did what I didn't last night - played sections of the PC and PS3 demos side by side, walking back and forth between screens to look at the differences with the character in the same locations (some full size PC screenshots below, with framerate overlay from Fraps).
You're right - to an extent at least. Compared side by side the PC version does in fact look a fair bit better, though it seems like it may actually only be down to a few, but important factors (and on PC I forgot to check out the small section of the river you can visit, and based on the video I've seen, water seems to be another area where there's a very big difference).
Here's what I saw - but mind you this is based on "normal" use of my eyes only, not counting the leaves on each palm tree or trying to measure the exact draw distance and sharpness of distant objects.
Textures - the PC version is much sharper here, landscape, characters, buildings, everything.
Object detail - Not quite sure what to look for here as no major differences stood out, but there were certainly cases where I thought the PC version had what I would consider more detail, such as the leaves on bushes (though there's probably an overlap with texture resolution here as well).
Lighting - I'm inclined to say the PC version has a definite edge here, although how much is a bit difficult for me to tell, as my PC monitor and HDTV don't quite match in terms of color, saturation, brightness etc. (and my PC monitor additionally has a very good H-IPS panel, while my HDTV is just a run of the mill Samsung).
Shadows - related to lighting of course, but overall I'll say that the two versions aren't nearly as different here as I initially thought. The "blob" shadow I mentioned last night was due to lighting conditions in a particular location, the PS3 version also has full character shadow. And it seemed to pretty much match the PC for buildings, trees and other objects as well, including soft shadow edges.
Terrain draw distance - I can see just as far in the PS3 demo as in the PC demo, and they look extremely similar (ie. heavy depth of field effect).
Object draw distance - Buildings, trees, grass. Pretty much everything on screen was visible from the same distance in both versions, including some patches of grass that had a pretty close draw-in distance on both systems.
So while I'm sure there are things I've overlooked (but if I didn't notice them when I was actively comparing two versions side by side, I seriously doubt I would during actual gameplay), it seems to boil down to the significantly sharper look and possibly better lighting.
Oh, and I'm still far from satisfied with the PC framerate (down to an unacceptable 20 in the last screenshot, and below 30 in several of the others). As you can see from the settings screenshot below I've turned SSAO off, and playing around with some of the others (haven't touched the resolution though, I really do feel like I should be able to play at native res) I did manage to get the framerate a little higher, but we're still only talking ~35-40 in the best cases.
I have the latest official Nvidia drivers installed (196.21). The new drivers I read about a few days ago must have been a new beta release.
Well, I'm done. It was fairly interesting to try to do my own comparison, but I can't really say I feel like applying for a job at DF - not that I would be remotely qualified anyway
[link url=http://mxmediah ost.com/img4/jc2-1.jpg
]http://mxmediah ost.com/img4/jc2-1.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-2.jpg
]http://mxmediah ost.com/img4/jc2-2.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-3.jpg
]http://mxmediah ost.com/img4/jc2-3.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-4.jpg
]http://mxmediah ost.com/img4/jc2-4.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-5.jpg
]http://mxmediah ost.com/img4/jc2-5.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-6.jpg
]http://mxmediah ost.com/img4/jc2-6.jpg
[/link]
[link url=http://mxmediah ost.com/img4/jc2-7.jpg
]http://mxmediah ost.com/img4/jc2-7.jpg
[/link]
Comment below viewing threshold Show
Comment below viewing threshold Show
I enjoyed the demo on X360 and now intrigued about the performance differences and will be trying out the demo on all platform just to see how much differences there are, ie gulf wide or barely noticeable? Android123 did say X360 edged it a bit for him personally.
Course I dont have same toys to play with as those DF folks, but our own opinion is always the most trustworthy!!
Comment below viewing threshold Show
And it would seem to be quite a lot better than the original.
Comment below viewing threshold Show
Same here, parachuting fast over a lot of guards and explosions, yanking the cemera around...didn't manage to create any tear....usually yanking a camera around does it, but....nope.
I have a feeling it's 720p ....DF probably test everything at 1080p, my xbox is set to 720p. Resident evil was similar, at 1080 It tears, at 720 it doesn't.
Personally i think they should do a test of a lot of games at 1080 vs 720 as i think for 360 at least 720p is the best mode for the console to be set at. It'd be good to know from developers if that's what they set the bar for, as some games seem like they hit 720 perfectly and if you want more resolution, they do it with the engine taking a little v sync hit (i might be miles off of course).
I actually think DF should test under 720p as that's what most people will have set(??) ...will affect the outcome of these face offs for sure...
Comment below viewing threshold Show
but, Nr 1 was fun, and this looks double the fun...ideal sequel
Comment below viewing threshold Show
It is a bout LOD stages. At which distances objects 'Level Of Detail' changes. From say LOD2 to LOD1 and highest LOD version that is LOD0. It's here wher there is a quite dramatical difference. If you watch the DF videos you see how they change to LOD1 very closely and LOD0 at about 10m distance. Not talking about amount of objects but rather their LOD levels and and switching distance and this impacts visuals quite greatly becouse we all dislike pop-ups/LOD switches that are visible.
Miths wrote: "Shadows - related to lighting of course, but overall I'll say that the two versions aren't nearly as different here as I initially thought. The "blob" shadow I mentioned last night was due to lighting conditions in a particular location, the PS3 version also has full character shadow. And it seemed to pretty much match the PC for buildings, trees and other objects as well, including soft shadow edges."
Well PC version sports soft shadows consoles do not, even stated by devs if you follow link I posted on previous comment page.
Here you see soft shadows (PC).
[link url=htt p://img502.imageshack.us/img502/8426/jc2pc.jpg
]http://im g502.imageshack.us/img502/8426/...[/link]
[link url=ht tp://img708.imageshack.us/img708/8007/jc2pc2.jpg
]http://im g708.imageshack.us/img708/8007/...[/link]
Here you so non-soft shadows (PS3 version).
[link url=http://openpandora.info/hosting/upload/files/2010 -03/2ddea0.png
]http://op enpandora.info/hosting/upload/f...[/link]
[link url=http://openpandora.info/hosting/upload/files/2010 -03/114592.png
]http://op enpandora.info/hosting/upload/f...[/link]
In motion it stands out even more between non-soft and soft shadows as the shadow pixelation due to low-res and low sampling rate makes the edges 'crawl' and 'flicker'.
Also SSAO makes great difference giving depth and making grass cast 'shadows' and objects look grounded to scenery in a more convincing way. This alone is one large difference. Just compare grass, bushes shading, objects connections/intersections to terrain/object parts.
About your perfomance issues I cant relate to them but you should disable Bokeh DOF effect and the CUDA water shader as I read these can impact perfomance with upto 40% on Nvidia hardware (not available for ATI hardware). Also you are running 1920x1200 4xAA, max/near max details and 16xAF with a GTX260? How could that possibly be worse than fluctuable 20-30fps 1280x720 0-2xAA (temporal?) 0xAF and quite a lot lower IQ?
Anwyay either version delivers extremely fun sandbox gameplay that is the most important aspect!
Comment below viewing threshold Show
My eyes seem to be more inclined to see PC framerates around and especially slightly below 30 as less than smooth, than the case is with console games running at the same framerates.
I'm positive this has nothing to do with the actual game version or the platform though, but rather the distance to and size of my respective screens (eyes around 60 cm from a 24" monitor for PC gaming, ~2.30 m from a 40" HDTV for PS3 games).
I tried a few screen resolution changes on the PC version just after my last post - 1680x1050 and 1280x800, but even the lowest (which looked quite ugly on my monitor, almost as bad as putting my face much closer than usual to my HDTV running the PS3 version) only gave approximately 10-15 extra frames per second, with the rest of the graphics setting the same.
... I decided to give it one last try before hitting send on this post, seeing as I really do like the sharper look of the PC version so much that I would like to consider it as a possible option when the game arrives.
It seems like the magic trick to better framerates was first and foremost two settings - setting shadows quality to medium instead of high, and turning off GPU water simulation, as that seemed to have a large negative impact even with the nearest water miles away.
I experimented some more with the various on/off settings as well. Many of them seem to be situational, so I might be looking in one direction and see no difference in framerate, but look in another and there could be a 5 fps difference or more between on and off.
In any case I've been able to arrive at some 1920x1200 (2xAA - 4xAA seems like it may even be possible, 8xAF) settings that stay above 40 fps most of the time, and regularly go into the 50s or even 60s - and still look sharper than the PS3 version (I still think a whole lot of that is thanks to textures and the much higher screen resolution though).
Comment below viewing threshold Show
Richard Leadbetter has just got his ass handed to him, he's been proven to be a dirty fanboy.
<a href="http://www.games onsmash.com/?p=15375
">http://www.games onsmash.com/?p=15375
</a>
Comment below viewing threshold Show
/Yawn
Comment below viewing threshold Show
I am mostly hovering between 80-100 fps with everything on max, 4x AA and 1920x1080. When I fly around and blow up a lot of stuff it sometimes momentarily sinks to 60-something fps. I don't have the Bokeh and GPU water options though since I'm using ATI.
Doesn't feel unoptimized to me.
Comment below viewing threshold Show
I am on a Q9550 at 3.6Ghz, ATI 5850 and 8Gb Ram and I Just switched everything to top at 1080p (SSAO HIGH, 8x AA, everything at the top). And it seems to run perfectly, never had a problem so an actual FPS figure would be great.
Comment below viewing threshold Show
Comment below viewing threshold Show
Comment below viewing threshold Show
Loved the demo by the way. I didn't like the original much, but this looks like a major improvement.
Comment below viewing threshold Show