Go Back   PC Gamer Forums > Games

    Thread Tools Display Modes
      #1   Report Post  
    Old 08-30-2012, 08:04 AM
    mason.s mason.s is offline
    Junior Member
    Join Date: Jul 2011
    Posts: 24
    Default What Can We Expect from Half Life 3?

    The Half Life universe has begun to occupy quite a bit of my thought lately. In fact, that's a bit of an understatement. If I came back here in a couple of days and told you that I had gotten a divorce because of it, just don't be surprised. But it won't be because my wife got so bored with my constant banter that she finally decided to throw me out. Oh no. It would be because I got so tired of her falling asleep in the middle of me telling her my thoughts on the subject and snoring away through all of the good parts!

    As an addendum to this current work, it may also interest the gentle reader that I have penned a thread dedicated to why I think it's taking Valve so cursedly long to finally grace us with the game that will end the world of gaming as we know it. That can be found here: [URL="http://www.pcgamer.com/forum/showthread.php?t=12625"]http://www.pcgamer.com/forum/showthread.php?t=12625[/URL]

    So, to begin. Recently I read back through the story so far on an incredibly well put together site which you can read here: [URL="http://members.shaw.ca/halflifestory/"]http://members.shaw.ca/halflifestory/[/URL]. It's a fantastic read and really puts all of the events of the games into perspective with a lot of the events that we don't see in the games but either are mentioned therein, or mentioned by the man that penned the HL stories, Marc Laidlaw. Since then what's been occupying my mind is a superficially and thus, quite deceptively simple question; just what can we expect when the long wait for Half Life 3 (or Half Life 2: Episode 3, or whatever. Hell, I'm starting to think the numeral "3" doesn't exist within the quantum confines of the space that Valve headquarters occupies. Maybe they'll just pull an Apple and call it "The New Half Life." Hell, as long as it's got the Freeman in it, a crowbar, and something to slay therewith, I'm cool man. I am cool. By the way, this may be the longest parenthetical statement in the history of mankind, so just to prepare you, the rest of the actual sentence picks back up.... now) is finally punctuated by the progress bar finishing its work after patiently ushering us through the download of the game and the "Play Now" button finally goes live? Well, I'll share with you some of the ideas my wandering mind got up to over the last couple weeks and maybe, if you're still with me at that point, you'd be so kind as to share with me some of yours. Mmmkay?

    A Better SOURCE

    Sometimes we jaded gamers get so wrapped up in petty things such as storyline, or character development, or ya know, like gameplay, that we begin to forget what really matters; dem grafix! Amiright? So yeah, I'll just leave this here ---> /s

    But on a more serious note, Valve has crafted something so great in so many, multifaceted ways with the Half Life games. They've given us a rich story to enjoy, an engaging gameworld with a sense of place that few are able to imbue such a digital space with, and then filled it with characters that we truly, truly care about. On top of all of that, they also managed to push graphics technology in numerous ways with the original source engine. To enumerate a few:

    Better facial animations than had ever before been seen in games.

    DX9 shaders and lighting models.

    Freaking drinkable-looking water in which everything from soda cans to dead bodies floated and acted exactly the way you'd expect.

    A robust and nearly pervasive physics engine that not only powered death animations, but turned the gameworld into a puzzle to be solved and made every animation move and flow like poetry.

    Doing all of that but still enabling the game engine to be run on hardware that was considered obsolete by every day computing standards and horrifically aged by gaming standards (anyone here remember the TNT2?)

    There are a number of other notable examples (3D skyboxes were a huge part of the immersion. Think of the constant presence of the Citadel off in the distance), but for the sake of this treatise, I will constrain myself to those mentioned above. If Valve were only to improve on those areas, without even going beyond just those, we would still have a revolutionary game engine. And this is how they could be doing it:

    "All right Mr. DeMille, I'm ready for my close up."

    There have been major advances in facial animation technology since Half Life 2 set the bar in 2004. The gold standard currently must be LA Noire. The techniques used to capture and reproduce an actor's facial movements which Team Bondi employed represented a quantum leap in technology in that crucial area. MotionScan, as the technology is called, records an actor's performance with 32 high resolution cameras, strategically placed to create a fully 3D, 360 degree view of the performance. That data is then translated onto the in-game model whereon the reproduction of the performance is rendered quite magnificently.

    So that's in then. We've achieved near perfection in facial animation. That's all she wrote, right? Well, not according to Valve man Jason Mitchell. He is quoted by develop-online.net, in a right spiffing article (to be found here: [URL="http://www.develop-online.net/news/37635/Valve-applauds-LA-Noire-facial-animation"]http://www.develop-online.net/news/37635/Valve-applauds-LA-Noire-facial-animation[/URL]) as saying, "Theres pros and cons to that LA Noire approach." While he lauds the tech, saying, Its impressive technically, for sure, and continuing, Im sure its something people will be referring to for years to come on the history of interactive facial tech," he's not convinced that it truly fits with what Valve are trying to accomplish. The system is based on a playback of a performance, which can go against how we like to think about characters interacting with our players.

    We like our performances to be far more reactive to what the player does, and not to something pre-acted on a sound stage. Its not completely obvious how [the MotionScan] tech would integrate into our work.

    That's the conundrum then. MotionScan and other motion capture techniques have come a very, very long way and we have seen some impressive acting performances acted out through the digital characters of this generation. But it's not enough. Valve need their tech to be dynamic. To react to the player and the gameworld in a natural way. Simply put, they need their digital people to act more like real people, and not just digital puppets.

    While developing the animation system for the original Source, Valve needed something flexible and compelling. Not finding such a technology among those on offer at the time, they set about creating a new one with an interesting basis; "...we found a method outside of traditional computer graphics," says Ken Birdwell in the book, Half Life 2: Raising the Bar. What they found was research from the '70s by one Dr. Paul Ekman who had dedicated himself to diagnosing mental illness based on facial expressions. "What he came up with was a set of about 40 or so 'Facial Action Units,'" says Birdwell. These became the basis for the facial expression system in place in original Source. This way, the animators and scripters had a library of accurate expressions to draw on that all naturally flowed from one to the other. It was dynamic and it was compelling. Further I expect Valve to take a similar approach as they push further than anyone else in the area of compelling digital acting.

    One company is working on a technology that could give us a preview as to how Valve may be approaching facial tech. Remedy, the makers of Alan Wake, are using motion capture that is accurate to within .02 inches to capture the 64 facial poses that they believe form the basis for all human facial expression (according to this article: [URL="http://www.joystiq.com/2011/04/11/remedy-la-noire-raises-the-bar-for-facial-animation/"]http://www.joystiq.com/2011/04/11/remedy-la-noire-raises-the-bar-for-facial-animation/[/URL]). From the article: "Animators can manipulate faces in real-time rather than rely on canned recordings." This bodes well for a future where facial expressions can be as dynamic and context driven as they are in real life.

    *Continued in next post*
    Reply With Quote
      #2   Report Post  
    Old 08-30-2012, 08:05 AM
    mason.s mason.s is offline
    Junior Member
    Join Date: Jul 2011
    Posts: 24

    Still I expect Valve to go even further by taking out so much need for animator intervention in the first place. The idea is that it truly is context that drives expression, rather than an animator having to step in and say "I think the character would react this way," and have to then manipulate sliders and such to get the right effect.

    Just as with original Source, I believe that we can expect Valve to come up with something completely new to fit their unique set of demands for player immersion.

    Let there be light(ing)

    Half Life 2 really saw the birth of many modern lighting paradigms. HDR lighting, per-pixel shading on a massive scale including diffuse and specular mapping, detailed reflections and refraction, many of these things which are now household names in gaming graphics were not seen generally in games prior to Half Life 2. It was part of a new generation of games which at the time included also Far Cry and Doom 3. After the release of these games a rapid influx of new graphics technology followed, culminating is such revolutionary engines as UE3, CryEngine 2, Frostbyte and others over the next couple of years. At this point, graphics technology advances at, at best, an iterative pace. When was the last real revolution? Frostbyte 2? Please. It's amazing, don't get me wrong, but it's not a huge departure from original Frostbyte and doesn't improve much on CryEngine 2 either, though the latter saw its first action with Crysis in 2007. UE4? Yeah, it's the nicest dev kit anyone's ever put together in all likelihood, but nothing I've seen from it really screams the same kind of revolution that occurred around 2004-2005.

    So when will we see the next real breakthrough? And what will it look like?

    Speculation only gets us so far, of course, but the current APIs in existence have been far from maxed out by modern graphic artists. Largely this is due to the low overhead, or utter lack of modern rendering technology that they have to work with when it comes to current hardware on the console side of things. But if we were to explore rendering without those limits, we could expect at least the following:

    Global Illumination - The Holy Grail of lighting. The thing current lighting models can only really simulate. A true global illumination model, simply put, is one in which all light acts like it does in real life, which is to say, being reflected and refracted correctly by objects in the world and creating both direct and indirect lighting situations (direct lighting: the object which is first contacted by the light-source. Indirect lighting: any subsequent objects that are lit by reflection off of or refraction through the directly lit object). Currently, "Global Illumination" techniques are a hodgepodge of various techniques which together simulate the effects of actual global illumination. Combining various types of light sources, simulating radiosity and ambient occlusion, altogether with hybrid shader effects and emissive textures and other such gobbledygook, we can approach the effect of a truly global ilumination model. But true global illumination system would be one in which light simply acts naturally, flowing, reflecting and being refracted in a natural way without manipulation through simulated techniques. When achieved, this will result not only in better looking lighting effects, but also in less work for artists to get a scene lit correctly and looking the way it should.

    More Geometry - While the faces in Half Life 2 were fantastic, character models were woefully short on geometry. This is one of the few places where Half Life 2 absolutely does not hold up well with modern games. However, in today's world of more and more memory in our computers and right on our graphics cards, and API support for tessellation, we should see a massive increase in the geometric complexity of characters and environments.

    Bigger Textures - Once again, with more memory overhead, combined with better and better compression algorithms, extremely high fidelity textures are right around the corner. Expect Valve to be at the tip of the spear here. They set great store by the immersiveness of their game worlds.

    Extreme Performance and Scalability - Half Life 2 was famous for a lot of things, but one of the things that most set it apart was the fact that it ran ridiculously efficiently on current-gen hardware of the time and could be scaled back to run on almost anything. Like GladOS running on a potato. Put it this way; there will be basically no one on the planet that will be excluded from playing Half Life 3 due to the constraints of their hardware.

    Water water everywhere...

    There is considerable reason to believe that water simulation tech will be huge push for Valve as they continue to craft Half Life 3, and any other games, going forward. Half Life 2 and the Source engine have received great laud for the water effects Valve was able to accomplish. The water was simply beautiful, and everything that interacted with the water behaved exactly how you would expect.

    Portal 2 continued to show Valve's emphasis on liquid simulation and it was a major triumph in that regard.

    As for Half Life 3, it can be expected that it will continue according to its pedigree. But more so even than that, water in particular will likely be the star of a lot of a show that will revolve largely around a mysterious, derelict, sea-faring vessel; namely the Borealis (even if it is just suspended in ice somewhere).

    We have so much further to push in this regard. Physics of water; volume, density, flow, almost none of those things have even been touched, let alone in any significant way. And of course, Valve will do it the Valve way, which is to make their water simulation to be not just a spiffy visual effect, but to have actual gameplay implications. It's time we move past Nathan Drake's wet t-shirt being known as the coolest thing someone has done with liquid simulation in a game. I mean, seriously? That's what we've got so far?

    What goes up must come down, etc. etc.

    If there's one thing that impacted me more, from a visual sense, than the graphics in Half Life 2, it was clearly and indelibly the physics of it. Valve put physics front and center. For the first time, real world physics became a part of how you interacted with the game world, and for that reason the game world began to feel more like a real world. Not only was it totally immersive to know that things in the game world would react to your presence and your actions in an accurate way, but it made puzzles that for the first time felt like part of the game world and not just part of a game. The puzzles in Half Life 2 are logical, intuitive, and never ever take you out of the game since all you're doing to solve the puzzle is interacting in a natural way with the game world. That barrel is in the way, so move it. That thing needs to float in the water, so stuff some of these empty, sealed jugs underneath it. It was blissful and immersive.

    But the physics in original Source are quite limited. Some objects have physical properties, others don't. Buildings, walls and terrain are notably almost purely devoid of any sort of physical property in the game world that we would attribute to such in real life. You can break a pallet with a crowbar, but no amount of rocket bombardment will ever bring down a brick wall in Half Life 2.

    Further, breakability of objects tends to be limited to pre-set portions of an object and lacks any sort of procedural dynamics. Ergo, hit a pallet with a crowbar and it pretty much breaks the same way, every time. Hit those pieces and they break into a predictable number of other pieces with predictable shapes.

    I fully expect Half Life 3 and the new Source engine to fully break into a more pervasive and global manner of physics simulation. I hope to see, and on some level am confident we will see, a far more procedural and dynamic approach to physics. Destructable materials with specific properties that govern how it can break, not necessarily how it will. Add to that better tracking of the disparate parts of the broken materials (such as individual bricks from a shattered wall, or the bits of wood from a broken pallet), deformable terrain, and accurate simulation of a material's structural integrity and the result could be not only visually compelling, but really change the way players approach different situations tactically.

    Well that's it. If you've stuck with me this far then, well, let's be honest, maybe we should both take a serious look at what we're doing with our lives. Also, good on ya. And thank you for listening to my ramblings for a bit. I assure you, it was a therapeutic experience for me indeed. And so will it be for you, gentle reader, as you post your own speculation below. There's a lot more to cover, including story elements, and many other technology discussions possible as I've only just covered a few of the visual advancements I believe we'll see. So, go ahead. Don't be shy. Hit up that there reply button and let'er rip.

    Last edited by mason.s; 08-30-2012 at 08:23 AM.
    Reply With Quote
      #3   Report Post  
    Old 08-30-2012, 10:14 AM
    General Armchair General Armchair is offline
    Senior Member
    Join Date: Dec 2011
    Location: Vienna
    Posts: 141

    Junior Member
    Reply With Quote
      #4   Report Post  
    Old 08-30-2012, 10:25 AM
    Scaramoosh Scaramoosh is offline
    Senior Member
    Join Date: Jul 2010
    Posts: 1,185

    What we can expect is a console focused FPS just like Source 2 to be console focused and all the latest Valve games have been apart from dota 2.

    Left 4 Dead's potential was clearly ruined by consoles, the menu design was built for consoles and they said themselves the level design was too. It was originally going to be more open and lack of physics gameplay and lobbies, all something done for consoles.

    Portal 2 again the puzzles dumbed down for consoles so people with controllers can do them. All the twitched based mid air puzzles were gone and it felt so dumbed down to the point where often you'd only have one small part of the level that you could fire a portal onto.

    CS:GO is clearly just a console CS, all the changes to it were so it would work on consoles and with a controller, the new radial style buy menu is the perfect example of this. Valve said that menu would change for the PC and it never did....

    So when we see the next HL I would be surprised if it was another dumbed down console game.
    Reply With Quote
      #5   Report Post  
    Old 08-30-2012, 12:36 PM
    Kan Hassen Kan Hassen is offline
    Senior Member
    Join Date: Nov 2010
    Location: USA
    Posts: 1,152

    Unlike Scaramoosh here, I don't give a shit about console mechanics. I don't give a shit about physics puzzles. I just want some closure.

    Bring back all the guns, fine. I don't need any new ones. Just give me some closure.
    Processor-Intel Q9300 Core 2 Quad 2.50 GHz
    RAM-4 GB DDR2
    Video Card-EVGA Geforce 9800GT 512 MB
    Motherboard-Asus P5KPL-VM
    Case-Nzxt TEMPEST
    Reply With Quote
      #6   Report Post  
    Old 08-30-2012, 12:51 PM
    SAeN SAeN is offline
    Grand Inquisitor
    Join Date: Jun 2010
    Location: Dundee
    Posts: 2,957
    Send a message via Skype™ to SAeN

    I'm not quite sure what the point of all this is.
    Reply With Quote
      #7   Report Post  
    Old 08-30-2012, 01:57 PM
    Aervalis Aervalis is offline
    Join Date: Mar 2011
    Posts: 97

    Expect Headcrabs.
    Reply With Quote
      #8   Report Post  
    Old 08-30-2012, 02:36 PM
    Tykh02 Tykh02 is offline
    Senior Member
    Join Date: Feb 2011
    Location: United Kingdom
    Posts: 962

    I imagine a potato powered crowbar may make an appearance...
    Reply With Quote
      #9   Report Post  
    Old 08-30-2012, 02:51 PM
    Megazell Megazell is offline
    Senior Member
    Join Date: Jan 2011
    Location: Where the shadow of Fort Apache linger.
    Posts: 3,155
    Send a message via AIM to Megazell Send a message via MSN to Megazell Send a message via Skype™ to Megazell

    Return of the Snarks!

    Reply With Quote
      #10   Report Post  
    Old 08-30-2012, 05:14 PM
    Freddie Freddie is offline
    Senior Member
    Join Date: Jun 2010
    Location: England
    Posts: 660

    Perhaps all the characters wearing silly hats?
    Reply With Quote


    half, half-life, life, source, valve

    Thread Tools
    Display Modes

    Posting Rules
    You may not post new threads
    You may not post replies
    You may not post attachments
    You may not edit your posts

    BB code is On
    Smilies are On
    [IMG] code is On
    HTML code is Off

    Forum Jump