Tech Analysis: Uncharted 3: Gameplay Demo

Just as we were finishing up on our tech analysis of the Uncharted 3 teaser trailer, Naughty Dog unveiled the first live gameplay demo on the Jimmy Fallon show, thus revealing more juicy technical insights into the engine upgrade contained within, along with direct-feed gameplay footage of the title

We lightly touched upon some of the things contained in the gameplay trailer in our last report – the possible use of MLAA, composition of the fire effects, character modelling, and rendering resolution – but due to the compressed nature of the initial teaser trailer, and the late arrival of the gameplay footage, were unable to take a detailed look.

Now, rather than expand our original report, we’ve instead taken the time to asses in greater detail the direct-feed gameplay trailer, looking at more of the game’s in-engine enhancements and referencing it with the original teaser footage from before. Here, you’ll hopefully find a nice companion piece to our previous analysis.

Let’s get on with it then.



In terms of framebuffer resolution, Uncharted 3 appears to be rendering once again in full 720p (1280x720) just like the last two games in the series. However, unlike both of those titles, this time around Naughty Dog seems to have used a different anti-aliasing solution for UC3.

Previously we saw the use of standard 2xMSAA (multi-sampling anti-aliasing) in order to mitigate jagged edges and shimmering artefacts, which actually worked very well all things considering – increased use of shaders, particle/alpha effects etc. But from the gameplay footage shown of Uncharted 3, we can see what looks like evidence of Sony’s custom MLAA solution replacing the more limiting MSAA.

Originally, we thought that those rather smooth looking 720p framebuffer grabs were simply supersampled promo shots, lacking any edge artefacts and appearing all too clean to be in-game. And while we’re still not convinced that they aren’t downsampled bullshots, the direct-feed gameplay footage itself not only shows off a similar smoothing effect, but also shows clear evidence of edge shimmering in areas where supersampling would have effectively dealt with that particular issue. Shader aliasing is also present, another thing which MLAA isn't capable of dealing with in its current implementation.



Take a look at the screenshots above of the window frame and the dislodged wooden beam to the right of it. Ignore for a second the compressed, blurry nature of screenshot and focus on those edges. Clearly, there is a reduced amount of edge smoothing going on here. We can easily see some sub-pixel edge artefacts that would have blended away via supersampling – like in the desert scene from the teaser trailer – but are obviously visible for all to see despite the lack of clarity.

There are also some polygon edges directly situated next to the low res fire effects that suffer from the same problems, though not all. Most likely, this is due to the low resolution alpha buffers interfering with the higher res geometry, and the MLAA being unable to smooth over the affected area. Again, such issues wouldn’t be present to such a degree in a downsampled image and would almost certainly be noticeably worse with 2xMSAA. You can see this below.


The rest of the scene however, benefits from large levels of decent edge smoothing. It's incredibly hard to tell by the compressed nature of the video creating additional artefacts, but we can see many areas getting what appears to be as much as 8xMSAA - some surfaces 16xMSAA, with others having around 4x, which comes as standard when using the technique.

Another benefit of MLAA comes in the form of excellent high contrast edge anti-aliasing, whereby distinct edges - usually too distinct for MSAA to generate good enough samples for - are handled with reletive ease. Although the trailer doesn’t demonstrate this, due to being set in a dark, low contrast environment, I can’t imagine UC3 being any different.

In order to accurately see just how well MLAA is implemented in Uncharted 3, I suggest you download a high bitrate HD 720p video instead of watching the awfully compressed YouTube one we have here. It suffers from shimmering and other artefacting not present in the actual game.


Moving on, and the fire effects have also been expanded over what was present in Uncharted 2. Here we see multiple layers of blended 2D sprites rendered in 1/4 the frambuffer resolution. Obvious artefacts such as pixelation aren’t visible due to the effect being smoothed over via both blending and filtering techniques. However, we can see evidence of jittering and some shimmering, much like with what is happening with shadows in the game. This is most noticeable at the base of the flames.

Whilst being technically quite simple in comparison to various fire effects we’ve seen in the past, the look is still very convincing with the various layers present on screen creating an impressive show via the use of sheer amounts, rather than advanced technical trickery. Also, the increased use of 2D sprite layers allows for a more organic look to be created. The fire in Uncharted 3 is both more animated and has a greater level of depth than in the second game. This is also helped by the scope the effect finds itself in – fire is everywhere, with varying layers spread all across the environment.

Granted Uncharted 3’s use of fire isn’t technically that impressive when compared to the multi-particle, and multi-layered smoke and fire present in the likes of Lost Planet. Although, Naughty Dog’s solution instead fits in with the PS3’s tight bandwidth requirements and the overall engine make-up far more comfortably than Capcom’s alpha-hevy solution ever would. Like in Killzone 2 and 3, the low res nature of alpha is carefully reduced via good use of filtering and layer blending.


In terms of character modelling, details, and shader effects, we can see that the renderings use for gameplay come remarkably close to matching the pre-rendered desert scene footage we covered here, in our first tech analysis of the game. Drake himself, geometry wise looks to be very close with only slightly paired back use of shaders and texturing. His facial features and animations look almost the same, lacking just a little precision in comparison.

Compared to Uncharted 2, and Drake looks to have gained slightly more in the way of detail. Facial details in particular look better – improved texturing and shaders – and his overall facial design has been artistically changed somewhat. He appears to be a bit chubbier this time around, looking older, more rugged. No doubt as a result of his haphazard worldly adventures, and from the development team adapting his polygonal mesh structure.


As to be expected, when up against the desert scene in the original trailer footage skin shaders have also been mildly reduced, as have small texture details and texture resolution. We can also see the usual shadow and self-shadow artefacts on Drake – jittering, some shimmering etc – that was present throughout Uncharted 2 but absent from the pre-rendered parts of the teaser trailer. This can be found on both characters and the environment. The grasses on the floor in particular are noticeably affected.

Whilst there are obvious differences between the teaser trailer’s pre-rendered in-engine footage and the actual real-time gameplay video, there is nothing that drastically separates them as a whole. Sure, the desert scene exposes a polish not possible in a real-time rendered envronment on the PS3 – with better use of shaders, perfect shadowing and polygon clipping. Although, most effects have at least been translated over to the in-game engine instead of being cut.


Performance wise, the short gameplay clip shows off Uncharted 3's solid state at this point. The demo doesn't appear to drop below 30fps - despite plenty of performance sapping alpha effects - and there is no sign of any screen tearing. It's likely that UC3, as with its predecessor is triple buffered: rendering three frames for every one displayed, simply discarding each torn frame until a clean one is found. Tearing should only occur when all three frames are being torn, a scenario only likely to happen in the most heavy load situations.

That said, the demo dosen't feature any extraordinarily large set-pieces like the train crash and helicopter battle from UC2, which would be a real performance indicator. Instead, what we have here is a somewhat more pedestrain scene designed to show off some of the game's additional graphical polish, along with a few new moves for Drake.


All in all, from what we can see Uncharted 3 boats improvements in lighting, texturing, and shaders over Uncharted 2, with more detail being present, smoother animations, and even better use of anti-aliasing. Some of these improvements are mere subtlties, while others are far more noticeable (like the inclusion of MLAA - UC3’s big leap forward, like with SSAO in UC2). And, with just under a year to go Naughty Dog have plenty of time for optimisations, to get those little details that make all the difference down to a fine art.

The first part of our Uncharted 3 trailer analysis can be found here, which focuses on the teaser trailer and also forms a complete look at the tech behind the game so far.

Tech Analysis: Uncharted 3: Teaser Trailer

So, Uncharted 3 has finally been unveiled. And for those who didn’t think this generation of consoles had much more to offer graphically… well, Naughty Dog’s latest – still just under a year a way from completion – definitely looks to silence the critics. Arguably, given the quality of the real-time and in-engine rendering on offer in Uncharted 3, there’s no need to hurry along to meet that five-year hardware lifecycle.

At the recent VGA awards Naughty Dog finally blew the lid off the next title in the globe-hopping, bandit-shooting Uncharted series, with Drake once again looking rather worse for ware, but this time stranded in the desolate Arabian desert.

The trailer showcases a variety of engine improvements, from shadows and lighting, to texturing and skins shaders. All have seen a noticeable upgrade. Some of these look to be suspiciously from in-engine - but not in-game - footage, whereas others look to have taken the actual in-game tech another leap forward (just look at those water effects). Either way, Uncharted 3 at this early stage looks absolutely incredible.



The trailer can be divided up into three distinct parts; the desert scene in which we see Drake walking slowly across some sand dunes, stranded after being involved in a plane crash; a scene in which him and sully are in a darkened room in what looks like a flashback of sorts; and the ending post title footage, which clearly shows off a few short seconds of actual ganeplay.

All three can be seen below for comparison purposes, and it’s here that we can see exactly which parts of the trailer are being rendered in-engine, in real-time, and actual gameplay.




Starting off with the desert scene, and we can see most obviously some of the changes Naughty Dog have implemented in their engine for Uncharted 3. Drake himself has been remodelled, his facial features adjusted to represent an older, rustier, heavily warn adventurer. He looks almost slightly chubby in appearance. But look at his arms and legs through his clothes, and we can see that it’s just his underlying bone structure that has been updated.

Texture detail has been noticeably upped. Even from the highly compressed video footage we can see additional subtleties in the form of wrinkles, stubble, and pores on the skin. Skin shaders too have also seen similar increases in quality, with all those little facial features reacting far more realistically with the game’s environmental lighting.

You could say that it looks almost too good to be in-game… and indeed it is. This second opening scene looks like it has been created using in-engine assets, though not rendered in real time. A few things other than the shaders hint at this. Most notable the composition of lighting and shadowing in the scene as a whole… they’re largely flawless in their execution.


The shadowing model in particular is completely artefact free whilst maintaining an incredibly level of precision and accuracy. Notice how both the environment and Drake’s own self-shadows lack any kind of jittering or mostly any edge shimmering, both of which are present in later parts of the trailer and in Uncharted 2. All shadows, both up close and far away, are perfectly cast without error, carefully adding a great deal of depth to the final image.

The lighting also complements this, with dynamic shadows reacting and changing according to the environment conditions. Shadows are cast where expected, and the shader model delivers reflections and subtle changes usually too computationally heavy to be replicated with such precision either in-game, or in real-time in-engine cut-scenes.


Although saying that, the art assets used in the trailer are all ones that will be used during gameplay according to Naughty Dog; they’ve simply upgraded some of the effects to complement the offline nature of rendering the scene for the trailer. It’s purely an artistic style choice, great for PR screenshots and posters, but not all that far off from what is present in game as it were.

There are other things that also point to this fact, the detailed nature of Drake’s attire for example. Everything from his belt, the bullets situated upon it, his shirt, and his scarf are beautifully rendered. In particular, Drake’s scarf seems to have a soft-cloth simulation of sorts integrated into its animation system, with no polygon clipping or edge and shadow artefacts. It’s like a CGI rendering but using in-engine artwork.



In terms of the actual trailer resolution, it appears that different scenes are rendered in varying original framebuffer sizes before being either upscaled or downscaled to form the final 720p image. Take the opening desert scene for example. Here we have what looks like a 1980x1080 original FB which has been downsampled in a process known as supersampling to deliver large amounts of full-scene anti-aliasing, resulting in very little in the way of jagged lines.

In fact, the aliasing that is present in this scene – shadows and subtle edge aliasing from certain angles, along with texture aliasing – is perhaps more down to shader aliasing and in particular, the lightsoures being used in combination with the resolution of the shadowmaps themselves. Also, there is barely any evidence of subpixel aliasing issues - we can see that thin lines and small pieces of geometry are highly smoothed over – a key component of using supersampling. But there is nothing more than a few edges with ‘soft jaggies’ standing out from the rest of the scene.

On the whole, this centrepiece scene from the trailer comfortably represents the kind of graphical upgrades to be expected throughout the actual game on a baseline level, although shader effects and texturing has obviously been increases slightly beyond levels possible in-game in real-time no less, with additional precision along with more detailed character modelling.


Moving on to the second key scene, and we can see similar engine upgrades taking place, but without the same level of perfection as in the desert portion we’ve discussed above. Both Drake and Sully feature improvements in texturing, normal mapping and shaders, but not to the extent as seen in the ‘in-engine, pre-rendered’ part of the trailer. This scene appears to be rendered not only with in-game assets but also being done in real-time without any of the ultra precise shadowing and lighting.

Obviously like for like comparisons aren’t completely possible – different lighting conditions and the fact that Drake himself looks to be younger, slimmer than in the earlier part of the trailer – although we can still see evidence of visual tweaks and changes while also spotting a few rendering artefacts that reveal the scene’s real-time presence.


Take a close look at the shadowing on Drake for example. Some of the self-shadows evident around his neck clearly show evidence of jittering and some edge shimmering – something larger absent from the desert part of the trailer. In addition Drake’s character model is slightly less detailed, with reduced shader effects and subtle texture details.

Like with the earlier scene, this part of the trailer also appears to be rendered in 1920x1080 before being supersampled down to 720p, which would explain the lack of any noticeable edge shimmering on the geometry. Although, in dark low contrast environments such things rarely manifest themselves.


Onto the actual gameplay portion of the trailer, and this is where things get difficult. The short and chopped up, cropped, and constantly resizing nature of the clips, along with compression induced motion artefacts make it difficult to assertain how close the game holds up to the graphical quality of the cut-scenes in any meaningful way.

However, we can at least see that the quality looks about on par, or close to the scenes in question. Much like in Uncharted 2, the cut scenes do appear to be higher quality renderings, though using in-game assets and running in real-time (assuming the same system from Uncharted 2 is in place, whereby all cut-scenes are real-time and not video recordings of the renderings). I imagine that gameplay will look basically the same to the untrained eye, featuring similar rendering bugs, but with a touch more detail and precision.

This part of the trailer also looks to be rendered in native 720p (1280x720) like with previous Uncharted titles, while anti-aliasing is yet to be determined. Some have said that MLAA looks to have been implemented, though with compression artefacts masking any potential aliasing issues and the clips so short/poor in quality, this may just be wishful thinking at this point. However, we shall be taking a look at the more recently released direct-feed gameplay video to investigate this further.


One thing that does stand out with the gameplay footage, is the noticeable upgrade in the engine’s ability to render water and fire effects. The water in particular looks incredible. It almost looks like a simulation if you don’t pay attention to how it flows and changes in motion.

The main body of the water appears to be more volume-based than particle-based - modelled with large mesh of animated geometry and normal maps - thus avoiding the PS3’s limited available memory bandwidth for alpha effects, although this is indeed backed up with some particles at the front as the water expands and spreads across the environment.

The fire effects also feature more animation than those found in uncharted 2. From the brief few seconds of footage, we can see an increse in the layers of 2D sprites used to form this effect, with greater levels of blending. Although, again… it’s pretty hard to tell, given the quality and duration of the footage.


Instead, a far better example of how Uncharted 3’s gameplay will hold up against both the in-engine pre-rendered parts of the trailer, and with the real-time in-engine cut-scenes can be found here, in the first direct-feed gameplay trailer.

As you can see, there is a noticeable difference between both the character and environment modelling, shader effects, shadows and lighting compared with the trailer. Against the standard cut-scenes too, we can see a slight downgrading - small, but perhaps greater than the differences in Uncharted 2.


Pre-rendered


In-game

We’ll be taking a look at the gameplay trailer in the next day or two at IQGamer in a shorter tech analysis, mainly focusing on the upgraded fire effects and comparison details between pre-rendered in-engine footage and gameplay. Interestingly, these gameplay shots feature very little in the way of aliasing, shimmering edges etc, which could lead to either an MLAA solution being implemented, or perhaps more supersampling - used in creating print and promotional quality bullshots. But we shall see.

In the meantime we can at least see that regardless of how the footage – and indeed the screens – have been enhanced, modified, or created from an offline render, that Uncharted 3 is already delivering a tangible improvement in rendering quality over and above Uncharted 2 and in some respects Killzone 3, while the game still has just under a year to go before it goes gold.

We also haven’t even mentioned that 3D support is also in the pipeline for day one. How this will be implemented – what method: side by line, top to bottom, half res, full res, etc hasn’t been confirmed – but Naughty Dog have stated that the engine is constantly in a state of flux, with new tech and changes being implemented right up until a month before the final crunch to completion. With that said, we won’t likely find out anything concrete for at least a few months. Although details on the 2D rendering engine will certainly surface long before that.


So… Uncharted 3 then, from what little we’ve seen of it, is looking mightily impressive at this point. It’s very early days in the game yet, with many questions still going unanswered, and a few in which the answers are obviously identifiable. The leap between the second and third Uncharted titles doesn’t appear to be as gigantic as the jump from UC1 to UC2 (SSAO was by far the most noticeable upgrade), although there’s plenty of subtle elements that provide additional flair to the engine, along with increased levels of realism to help engross the player further into Drake’s world.

As we’ve already mentioned, a direct-feed gameplay trailer has since been released to complement the highly directed teaser trailer, showing off two minutes of continuous gameplay in which to present the various in-game engine improvements. You can expect another analysis, albeit much shorter based on this in the next few days.

Thanks go out to Nebula for the framebuffer analysis/pixel counting.

Tech Report: A Look At LBP2's Graphical Upgrades

Despite being a platformer with a slightly cutesy disposition, Little Big Planet is no stranger to technical excellence. You might not think that just by looking at it, but under the hood the original LBP was as interesting from a tech point of view as it was from a gameplay perspective.

Little Big Planet 2 then, appears to be much the same in this regard; the characteristic real-world physics of the title so integral to the very core of the game backed up by some impressive, and downright interesting engine enhancements. This is precisely the reason for us to be taking an extended look at the title. What we have here could almost be described as a full tech analysis of sorts, but in reality it’s more of a small glimpse, and dare I say, intriguing update into what media Molecule are doing with this sequel.

The first thing to notice, obviously, when going over the screenshots is the abundance of clean lines and smooth looking edges, or rather the distinct lack of any jaggies spoiling the scene. Now you might be thinking ‘supersampled’ when seeing the quality of the screens, and initially that’s exactly how I felt about the situation. However, this isn’t actually the case as all the screens you see on this page are direct-feed captures rendering in native 720p (1280x720), and are not downsampled from a higher resolution, as is usually the case with most PR shots released to the press.

The reason for the game’s lack of jagged lines and supersampled look then is clear - the use of morphological anti-aliasing has been implemented into the graphics engine.

Although the above anti-aliasing method has been confirmed by Media Molecule themselves you can still see some evidence of sub-pixel based jagged edges, another hint as to the inclusion of MLAA, as with supersampling this just wouldn't occur


MLAA has been featured a few times before here at IQGamer, namely when covering Santa Monica Studio’s God Of War 3, and more recently Guerrilla Games’ Killzone 3. It feels like not a month goes past without some new first-party title using the technique, a technique which is not only cost saving in terms of memory, but also in terms of securing the highest levels of image quality in a console specific release.

From the screens featured on this page it is apparent that the MLAA does far more for the image than what is possible via the more traditional MSAA, only to be beaten by supersampling (SSAA) which isn’t doable, realistically, on consoles due to the additional rendering performance incurred.

In LBP2, like with GOW3 the use of MLAA provides up to 16x MSAA coverage on some surfaces, and better than 4x on most others. The only area in which this form of AA falls down is when dealing with sub-pixel aliasing, where by any polygon edges smaller than the size of a pixel (this is a sub-pixel) receive absolutely no AA coverage at all. The same is true for MSAA as well, but not for supersampling which covers all aspects of the entire image. Regardless, this new form of anti-aliasing is a huge improvement over the 2xMSAA used in the first game.

Of course, the use of MLAA is just one part of many fundamental changes to the underlying graphics engine. For this sequel the developers have also opted for a solely forward rendering approach throughout the entire game, making transparencies and shadowing much easier to do.


Previously the original LBP used a differed solution to rendering lighting and shadowing in the game, which meant issues with displaying certain effects and also a reduction in shadow quality.

By switching to the traditional forward rendering approach this time around, the developers have been able to easily upgrade the shadowing system used in this sequel. Soft shadows are present throughout, which look much nicer than the hard-edged ones used in LPBP1, with shadows being cast for every main light source you can see, now without the need to pre-calculate them as shadowmaps like before. These soft shadows also blend in well with the game’s newly implemented use of screen-space ambient occlusion (SSAO), which is performed in real-time along with most of the lighting and shadowing.

This use of SSAO also gives an even greater depth to the image not found in the last game, complementing the range of visual enhancements on offer.

Seeing as the computational requirements for producing such graphics effects have probably also gone up, the resolution of these shadows is still relatively low compared to the rest of the game. I’m not sure how much resolution loss is occurring, although it is apparent that all shadows look slightly softer than you’d expect them to be, especially compared to if they were rendering in same resolution to match the rest of the game.

Transparencies however, now look to be rendered in a higher resolution than the first game, using proper alpha coverage, whereas before they were rendered in a half-res of sorts using the bandwidth saving alpha-to-coverage, which lead to a slight screen-door effect being present on all objects that featured it.

The true nature of the translucency is a nice touch, and goes well with other enhancements being made to the whole particle/effects system used through the game.


From what we can see LBP2’s use of soft dynamic shadowing in combination with SSAO is undeniably impressive, providing an incredibly life-like appearance to how the whole scene is lit at any given time. The tightly woven nature of the lighting and how it commands the way shadowing occurs in the game cannot be understated, and to this the addition of a global illumination (GI) style solution adds more believability into the mix.

In LBP2 light on some stages looks like it travels down from a main point (the sun, or a singular internal light source) into the world appearing like it reflects off some objects and onto others as it shades and lights the environment. An impressive visual trick, as in reality it isn’t being done in real-time at all, but instead is a pre-calculated simulation that uses something along the lines of a lightmap, moving these around on certain surfaces to create this effect.

The illusion of proper GI, and some cleverly implemented god rays is the focus here, convincingly backing up the original LBP’s realistic lighting model. Alone, these elements are merely hot discussion points. But when all brought together they bring a real sense of naturalness and cohesion to an abstract game world, much like in the way the use of physics provides us with some tangible connectivity between the game and our own experience of reality.

There's still a disconnect for sure, seeing as the whole style and notion of the game’s world is completely off-the-wall, but in it’s own little confines feels completely organic, succinct even. And that’s perhaps Media Molecule’s biggest success, not the visual wonderment that all these effects and improvements provide, but the ability for them to casually sink in and blend together so seamlessly into being part and parcel of the experience.

Low resolution shadows aside, and maybe some stray un-anti-aliased sub-pixel edges, LBP2’s tech has seen some noticeable changes, and some ingenious solutions to problems most lesser developers tend to flake over. The game isn’t visually outstanding from a ‘wow factor’ point of view, but instead has its moment in delivering small subtleties that do more for the overall experience than just for the sheer technical hell of it.

Tech Analysis: Killzone 3 E3 Demo - 2D VS 3D

Killzone 2 is still one of the technical benchmarks for Sony’s PS3. However you might dislike the dark and grainy art style, or the subdued colour palette on offer, the game showed developers (and gamers alike) just what was possible on the system when tightly using the CELL + RSX combo the way is was supposed to be used. And for E3 2010 Sony took Guerrilla Games’ Killzone 3 as their lead technical showcase for not only the PS3, but also for their newly affirmed focus on 3D gaming.

From all the screens and videos released there’s no doubt that Guerrilla’s latest is as visually striking as it is technically brilliant, but what about under the hood? What’s changed? And more importantly how well does the current engine hold up to rendering in 3D, especially with minimal compromises on what’s being pushed around on screen?

Well, in this feature that’s exactly what we’ll be taking a look at, analysing the E3 build of the game in both 3D and 2D, seeing exactly what improvements have been made and what has been paired back in order to get the game working (fully playable I might add) in 3D.

Anyway, before we talk about that in more detail, lets take a look at the game in 2D and see just how it fares at its current point in development.

Like with it’s predecessor Killzone 3 renders in 1280x720, but rather than use quincunx anti-aliasing again the developers have elected to use morphological anti-aliasing instead - a far superior technique for reducing jagged lines whilst maintaining overall image quality.

Looking at the screenshot below, the effect the MLAA has on image quality is obvious. The final image is much sharper and clearer, with fewer jagged lines being present than before, and without any additional blur caused by the use of QAA. The only blur you are seeing in the screens is caused by the various post process, and depth of field effects that Guerrilla are using throughout the game, all of which are artistic choices and not technical compromises. It’s all part of the dark and gritty look of the franchise.


Like with God Of War 3 some surfaces receive as much as 16xMSAA, whilst others more in the range of 4x, or occasionally less in areas with ultra small polygon edges. Ether way the use of MLAA is a marked improvement from the QAA of the first game.

However, unlike in certain games (I’m talking about you Red Dead) the use of QAA in Killzone 2 wasn’t at all detrimental to the overall image. Instead the slightly blurrier looked suited the art style the developers were aiming for, and the image still looked particularly clean and quite sharp. The same could also be said of Insomniac’s Resistance: Fall Of Man, and its sequel - both of which used the infamous QAA.

Switching to MLAA simply allows texture detail to come through unscathed (no blur) with greater levels of edge smoothing at a lower cost. You’re getting a smoother look without making any of the same compromises as before, and potentially saving on memory as well.

So like with God Of War 3 the use of MLAA does much to improve image quality whilst having less of a performance hit than you might think. Although are times in which this new form of anti-aliasing isn’t so effective at dealing with jagged lines, particularly when coming up against sub pixel aliasing - something which does crop up noticeably in parts of Killzone 3. Areas of the game which features loads of thin polygon lines; fences, railings, power cables etc, are all prone to displaying jaggies, and this is something that MLAA can’t really help with.

Below is a clear example of what I mean. In the screenshot it is evident that sub pixel (a triangle smaller in size than a pixel of the rendering resolution) edges receive no AA of any kind, something which would either require a change in how these objects were rendered or a switch to supersampling in order to resolve the problem.


Essentially MLAA works by detecting edges in a scene on a pixel level, finding them and smoothing them over resulting in a highly effective way of dealing with jaggies. This is perfect for high contrast scenes (unlike with MSAA) as edges are clearly detectable thus being easily smoothed over. The problem comes in when the edges you have to deal with are smaller than one pixel of the rendering resolution, and as MLAA works only on pixel size edges anything smaller simply gets no anti-aliasing. Or that is how I understand it. The result is some edge shimmering and noticeable aliasing on objects with lots of sub pixel edges.

Moving on to smoke and particle effects, it is obvious that they are again rendered in a lower resolution than the rest of the game.

Like in Killzone 2 all alpha effect buffers are rendered in 640 x 360 (quarter of the resolution of 720p), a common practice for most PS3 developers due to the system’s lack of available memory bandwidth compared to Microsoft’s 360 with its 10MB EDRAM.

Basically PS3’s GPU, the RSX, features a fairly low pixel fill rate, and this effects how many transparencies can be drawn on screen at any given time. 360 on the other hand through its use of EDRAM provides the GPU with a much higher fill rate enabling not only more transparent objects to be drawn at once, but also to feature transparencies at a matching screen resolution en masse.

In motion the lower resolution of the alpha buffers is hardly visible with the various post processing effects going on – such as depth of field - and they do look rather smooth and well defined. Although, as we can see below in still screens these effects still appear to look softer than the objects around them.


Impressively, it looks like the developers are using volumetric effects for all the smoke in the game (like with Killzone 2), although in reality this is somewhat misleading. Instead of actually rendering 3D volumetric particles, they are using layers of 2D sprites which have been blended together and combined with geometry using something called ‘alpha test’ in order to re-create that volumetric look without the added processing cost of doing it for real.

This blending is also one of the reasons why the smoke and particle effects all look somewhat soft and smoothened, in addition to the AA that they seem to be getting on top of that, and of course the upscaling taking place. It is also noticeable that the higher contrast nature of the stage demoed at E3 seemed to lessen the volumetric look associated with the effects, whilst also diluting the dynamic lighting being used somewhat.


Despite this Killzone 3 still looks visually stunning though, losing nothing along the way from the last game, and the developers may have also seen fit to upgrade the use of ambient occlusion for this latest instalment.

Previously for Killzone 2 Guerrilla were in the process of adding real-time SSAO (screen-space ambient occlusion) to the game but didn’t have time to properly implement the effect, instead using baked AO as a substitute. Now it looks like this could have been changed, and for the first time we are seeing what appears to be proper use of SSAO for Killzone 3.

Although officially unconfirmed at this point, the screenshot below clearly shows some evidence of the effect being present. Just check out the shadows on the floor below the Helgast’s feet, in which we can see that something different is definitely going on. SSAO? Maybe. And it wouldn’t be a surprise to see it given the fact that it was being worked on long before development on KZ3 started.

You can also see the improvements made to texture quality compared to KZ2. Textures are clearer, crisper, and generally more detailed than before, perhaps as a result of no blurring being present from using QAA, but also because texture resolution seems to have been upped for certain objects in the game.


From what we’ve seen so far KZ3 is shaping up to be a clear visual improvement over the last game in 2D, with the MLAA being a particular standout, and the cleaner, sharper look appearing giving the game a more polished feel overall. The sense of scale has been noticeably upped, and the sheer amount of stuff going on at once is undoubtedly impressive.

But how does this compare with the game running in 3D?

Surprisingly, Guerrilla Games have also managed to achieve some of these feats when rendering the game in this mode, like keeping in all the complex smoke and particle effects without cutting back on the amount of stuff on screen at any time. However, the game’s rendering resolution in this mode leaves a lot to be desired.

When rendering in 3D you are essentially doubling up most of your graphics work rendering every frame twice, one for each eye. Now, certain things can be carried over between frames to save on performance, but many things can’t, and this why cutbacks have to be made. And for Killzone 3 there are sizable cutbacks with regards to the games rendering resolution, and the resolution of alpha channel visual effects.

Below are two screenshots showing the game in action. The top one shows the game running in 3D mode, and the bottom the same scene but running in 2D. As you can tell the difference is night and day, with the 3D version looking rather unsightly.


Killzone 3 in 3D


Killzone 3 in 2D

Looking at the above screenshots you can see that image quality has taken a massive hit as a result of the steep drop in both rendering resolution of the main framebuffer, and the alpha channel effects buffers.

For its 3D mode Killzone 3 renders in 640x716 with MLAA, and the effects buffers (which were already rendering in quarter resolution) are again halved down to 320x360 creating an unsightly scene of jagged lines and upscaling artefacts.

The alpha effects in particular seem to suffer the most with this, as when they overlap with opaque geometry they cause aliasing atifacts to appear heightening the games increased jagged appearance. In addition shader and sub pixel aliasing are also magnified as a result.

Having to render twice the amount of geometry on screen at once also causes problems, and various reports of seeing the game running in 3D state that there is noticeably greater levels of pop up compared to running in 2D mode. Even though you are running at half resolution, you still have to render the geometry twice so there is still an impact with performance despite cutbacks in the number of pixels being worked on compared to rendering in full 720p for 2D.

So far it isn’t looking too good for Killzone 3 in 3D, with the current build definitely being a poor representation of how the game should look, although in that respect you simply cannot expect standard 2D levels of performance with current generation console hardware. There just isn’t enough power to handle it, and with optimisations only so much can be done. However seeing the game being displayed with all the intricate particle effects and multiple light sources in 3D is pretty impressive, even if the result isn’t as clean or as smooth as we’d like. I would say that it not only shows promise, but also is a key indication of just how much untapped potential is still left inside the PS3 hardware for games in general.

There is also the opinion that increased levels of jagged edges and upscaling artefacts are less visible when viewing them in 3D compared to seeing the same thing in 2D. How true or accurate this is I don’t know, not actually seeing Killzone 3 running in actual 3D in the flesh – only a 2D version of the game’s 3D rendering mode. But the argument for even having a cut down, lower-res 3D mode is unsurprisingly strong, especially given the marketing potential for this new format.

Either way Guerrilla Games have stated that they are targeting 720p (1280x720) for Killzone 3 in 3D, and it’s likely that they’ll do whatever it takes to reach that milestone without overly compromising the look of the game, optimising where necessary, and cutting back on post processing effects that don’t work so well in 3D (motion blur, depth of field). Full 720p looks to be pretty much out of the equation, realistically. But you know, maybe something like 852x720, which would still provide better image quality than 640x716, but without having to cutback as much on the core graphics make up of the game.

With Killzone 3’s release not until February next year the developers have plenty of time to improve and optimise their engine for both 2D and 3D, so it will be rather interesting to see just how well the game fares a few months down the line. As new videos surface, and information gets drip-fed out we shall no doubt be taking another look at the game and the tech behind it.

Tech Report: Anti-Aliasing In God Of War 3

Looking at any of the full resolution screenshots from the final release code of God Or War 3, you could be forgiven for thinking that what you are seeing is nothing more than supersampled bullshots, shots created for the purpose of making the game look good when blown up several times its original size for magazine printing. However you’d be very wrong indeed, because although some have speculated that these latest direct-feed grabs of GoW3 are downsampled images like most of the screens released for the demo, we at IQGamer can confirm that they in fact showcase the game’s unique Anti-Aliasing method in action.

That method in question is MLAA, or Morphological Anti-Aliasing, another form of edge smoothing done using the SPUs on the PS3 rather than the RSX GPU. We only heard about the game’s somewhat exciting AA solution a couple of weeks ago, after it was revealed in an online interview with one of the developers working on the title. It seems that whilst the original plan was to use the usual 2xMSAA method of jaggies reduction – and that was present in the recent PSN demo – that all changed at some point this year, in which a programmer at Sony’s Santa Monica Studio found a way of performing a much better method of AA on the CELL, whilst in turn saving the use of precious GPU and memory cycles in the process.


Just take a look at the screenshot above. The AA is absolutely amazing, far better than anything the 4xMSAA 360’s GPU can manage, and in some places matching perhaps what you expect from a high end PC gaming rig performing 16xMSAA. Either way, it is the best example of Anti-Aliasing on any console game to date, bar none.

In addition, the developers themselves at Sony’s Santa Monica Studio stated that they managed to increase both performance and quality of the AA used, along with other effects in the game, by offloading tasks normally done on the GPU to the SPUs on CELL.

“AA on the cpu is MLAA Morphological Antialising. We saved 5-6 miliseconds by moving it off the cpu's. Many props to our coder Cedric for making this happen and it looks way better!”

Of course the benefits of having better AA isn’t just dealing with jaggies reduction, but also reducing aliasing caused by certain shaders, such as specular aliasing for reflections and transparencies, whilst also eliminating texture shimmering caused by a lack of AA, bringing a stability and smoothness to the image allowing the art design to really shine. Its importance is usually only ever compromised due to the lack of available memory and bandwidth available on both systems. On 360 it’s having to fit the final framebuffer into the 10MB EDRAM, while on PS3 it’s dealing with just a lack of available bandwidth altogether. The use of MLAA can alleviate some of these issues providing a more even split of resources, which can then be used to improve other graphical aspects of the game.


The above screenshot demonstrates how well this use of MLAA works on gameplay scenarios. Notice that every single pixel in the shot is covered by the AA, without any extensive blurring of the detailed textures, even smoothing off edges that are small and far off into the distance. Everything remains clean and sharp, minus the parts of the scene which feature the engine's depth-of-field effect, used only on specific areas of the scene. You can see this happening on the botton left of the image, in which the scene is slightly blurred, whilst on the right, just above Kratos, all details are sharp and clear as day, even in the distance.

Another recent release, Metro 2033, due out this Friday, also uses a custom form of AA done specifically on the CPU, but this time on Xbox 360. It’s called AAA, Analytical Anti-Aliasing, and works by finding all the pixel edges, rendering samples of them in a higher resolution, before then downsampling those samples for use in the final 720p framebuffer. Effectively the only performance hit you get, is a slight one caused by having to render a small amount of pixels in a higher resolution, and the end result is another AA solution which allows for the same effect as using 16xMSAA. However the developers still managed to save 11MB of memory by using this technique instead of regular MSAA, which was their original choice, memory used for other visual effects taking up precious space in the framebuffer.

So God Of War 3’s impressive use of AA is just one of an increasing amount of games looking for a more flexible AA solution than the out of the box MSAA fix found on both consoles GPUs. By exploring these other areas, and by doing things differently they can ultimately drive forward better image quality, whilst focusing more on things such as art design, which has always been more important that raw technical prowess.

Lighting is another area in which GoW3 proves this notion of thinking.

“We built our engine around being able to use up to 20 dynamic lights per game object. The light can be big or small, it doesn't matter. In the end, I believe we support up to 50 dynamic lights per game object. We are not using a differed lighting scheme. Our lead programmer Vassily came up with this amazing system during pre-production, us artists love it!!! We can place lights in Maya and have them update in realtime in the game on the PS3, its like being able to paint with lights.”


Again, this is Santa Monica Studio thinking up a solution that isn’t right out of the box, but is technically at the forefront whilst also being very much in consideration with just what their artists were looking for. It’s exactly this kind of attitude and understanding that is driving forward the use of alternative AA solutions, at the same time providing increased graphical fidelity, whilst saving on performance, and making development easier for future titles. This is especially important with regards to memory related issues, which generally keep coming up in every console generation, and in which working smarter, not harder is really the only way in pushing forward console performance years after launch.

From a developers point of view, now has never been a more exicting time to be involved in console gaming, in which both tools and innovation are coming together setting the new standard of videogames production. Anti-Aliasing is of course just one part of the picture, but in GoW3's case, it could be its defining one.

IQGamer will be performing our in-depth technical analysis on God Of War 3 this weekend just after the game’s release. Be sure you check it out, as it will be filled with interesting technological reveals about some of the development process behind the game, and the ever more innovative steps that developers are taking to push the visual envelope forward.