DX3 reviews?

[paladin181]

I still can't bring myself to shell out cash monies for a brick that will only, only play games. At least with a PC I can kinda justify that I'm also going to use it for pornwork.

Plus modding (ahahaha).

And of course, it's not like I'm burgeoning with excess cash at the mo anyway (though soon to have first payday in almost a year!!), so hey.

Aaaanyway: regarding QTEs, I think the only ones I really have experience with are probably those from deadspace, and maybe..arkham asylum? And both of those were chiefly of the MASH TEH BHUTAN TO NOT DIE FROM GRIBBLY variety. And I fucking hated them. As I recall, farenheit had things like "press this button here, then this button here, then alternate these two buttons for this bit" etc etc stuff, which sounds preferable (though hey, 'things preferable to <press X to not die>' is a looong list, including some forms of invasive dentistry).

It's kinda odd that the whole multimapping of context-specific functions to a button 'press E to climb/run/open_door/fellate_necromorph/read_magazine' is slowly essentially turning into a QTE by itself.

And of course, I fucking hate that (and have ranted about it many times before).

I can kinda see why they might want to do this: complex behaviours are getting easier to incorporate, but control systems are not matching them for complexity, and indeed the level of complexity required would produce one motherfucking ugly controller (like..N64 levels of bad)..but still, I wish they'd at least try to do something beyond "press X to not die" more frequently. The tentacle drag sequences in deadspace, for instance, were kinda QTE-esque but were really nicely implemented, since it felt like a natural extension of what you were doing anyway (i.e. shootfighting), rather than arbitrary (pressing use now shakes off gribblies, for some reason).

I almost never use my PS3 for gaming, actually. As much as I play, I more often use it to stream movies from my 2TB HDD or Netflix, and occasionally to browse the web.

I don't dislike QTEs as long as they're not over used. In Heavy Rain, I thought it was ok. They were mostly intuitive and pretty solid, plus it was the ONLY method of interaction. In God of War, they were an annoyance that required switching game modes.. "Hammer Square, square, Triangle, square, trian-- follow the controller sequence with the right joystick? FUCK! Botched it because I was still pressing Triangle and square!!!" Annoying is hardly the word to describe that. It's down right inconvenient (and yes, I am a huge GoW fan despite this).

I disliked the N64 controller simply because of the inaccessibility of the buttons. You're using EITHER Z or Left trigger, you're using EITHER the Joystick OR the D-Pad. And then there were the games that were poorly conceptualized to abuse both alternate sets at the same time. "What do you mean hold Z and left trigger at the same time?!"

[AEmer]

Finally remembered what I wanted to say before, in response to something Jonas said:

And I think it makes the Deus Ex series more viable in the long run, because games like Hitman take ridiculous ass-loads of resources to make, with all its special-case scripting and context-sensitive animation, even compared to Deus Ex.

I question this, I really do. I think the biggest problem is that there's a requirement of a cross-section of skills - that the level designers are also capable game designers and programmers, and have a good sense of art design too. When I consider the opera level in Hitman: Blood Money, there's a _ton_ of gameplay in that level when compared with any level of similar size in Mass Effect or Gears of War. Sure, it might much have more impressive artwork than the Mordin recruitment mission on Omega (for example), but I bet that less effort was invested on the graphical fidelity overall. I'm fairly sure that the amount of unique animations should be in a similar ballpark too.

But the opera level has incredible depth, and where they started and how they ended up where they got? I envision either a sublimely talented dude doing most of the work, or at least 5 guys where 2 effectively do nothing but manage overhead and half of the 3rd ones work gets tossed due to changes.

I shouldn't be surprised if IO really did have a sublimely talented dude on this one. Of course, Jonas had a weeks internship there once, so he probably knows.

Anyway, however things got put together, like I said, I question whether the opera level was truly that much more expensive to make than the Mordin Solus reqruitment mission...I really want to know the answer, to know if Jonas is right, but I have no idea how I would find out

______

On the footsteps sound issue, because I wanted to say something about that too but I didn't remember:

I don't think designing a system for more accurate footstep simulation is necessarily hard or prohibitively expensive on resources. Computergames are made up of smoke and mirrors, is how I like to put it: Things are never done in a fashion that simulates a world, they're done in a fashion such that they behave in a reasonable fashion. Big difference, because simulating things will sometimes give the correct results (though it's prohibitively resource intensive and sometimes inadvertently buggy when run in real time), but reproducing reasonable behaviour, now that tends to function.

In this situation, you can do a bit of tomfoolery: Assume that guardsmen can't hear 2 seperate pairs of footsteps at once (apart from his own) if one of them is loud and is supposed to be coming from a fellow guard. Then assume that the guardsmen near to Jensen can hear exactly what Jensen can hear. Then put down the footsteps of each guard as a unique resource that needs to be locked as it's used.

When Jensen generates noise (probably something done once a second or so), and enemies are close enough, the AI will determine whether they can hear Jensen or not. If a footstep resource that is not the pinged' guards footstep resource is currently locked, then triangulate the distance from the guard that hears jensen to the guard whose footsteps are also going off, and if the number is too low, ignore the noise.

Since AI isn't mission critical (doesn't need to be known immediately), this little bit of logic can be handled in a distinct thread asymchronously, so the extra check and triangulation won't delay a single frame by more than a negligible amount.

Similarly, you can put in a check to see if the unamplified volume on the FX sounds channel is too high, in which case the noise should also be ignored.

The effect would be that guards would never hear Jensens footsteps if one of their friends was too close and Jensen could hear said friend. If one was somehow too close, but the player can't hear him, the player will never figure out that he shouldn't have been heard in the first place, so immersion is maintained.

Ok, so I made several assumptions about the underlying systems of HR, but my point is, a good programmer can find a non-resource-intensive "trick" to manipulate the player into thinking the game is smarter than it actually is, and that might very well be applicable here.

A more pertinent example would be how snow and rain are implemented in TNM. The foolish players never realize that it's only _them_ that has the bad weather, not anybody else , which is one of the most connieving tricks I've yet seen

[Dragon]

A more pertinent example would be how snow and rain are implemented in TNM. The foolish players never realize that it's only _them_ that has the bad weather, not anybody else , which is one of the most connieving tricks I've yet seen

What do you mean now with "only them has bad weather not anybody else"?

[Jonas]

I question this, I really do. I think the biggest problem is that there's a requirement of a cross-section of skills - that the level designers are also capable game designers and programmers, and have a good sense of art design too. When I consider the opera level in Hitman: Blood Money, there's a _ton_ of gameplay in that level when compared with any level of similar size in Mass Effect or Gears of War. Sure, it might much have more impressive artwork than the Mordin recruitment mission on Omega (for example), but I bet that less effort was invested on the graphical fidelity overall. I'm fairly sure that the amount of unique animations should be in a similar ballpark too.

But the opera level has incredible depth, and where they started and how they ended up where they got? I envision either a sublimely talented dude doing most of the work, or at least 5 guys where 2 effectively do nothing but manage overhead and half of the 3rd ones work gets tossed due to changes.

There are a couple of points to make here.

Mass Effect and especially Gears of War are highly modular games. No matter where in the galaxy of Mass Effect you go, you'll see the roughly same enemies, the same aperture, the same technical equipment, the same crates, and the same building types. In Gears of War, it's even more so - there are some bosses and minibosses, but mostly you have a general pool of enemies that are slowly introduced throughout the game, but previous enemies never go away and they rarely have to change behaviour depending on the circumstances. More importantly you'll do precisely the same things in both games - the weapons aren't unique to each level, there are no fancy environmental interactions that are only used in one particular place in the entire game, at most there's some fancy mission scripting when the sun irradiates you in Mass Effect 2 or the rain is deadly in Gears of War 2.

I don't have access to the particular project management knowledge of any of these games so I'm extrapolating from my own knowledge and what I've picked up around the corners of the Internet. Basically, Hitman: Blood Money is insanely expensive to make because an incredible amount of stuff in it is only used once. Every mission has a set of unique "accident" kills that require unique programming and animation. Animation, by the way, is probably the most expensive asset type in game development, possibly followed by AI behaviour. And of course every level has a fair amount of unique AI behaviour as well - even if they've just coded a really flexible AI system, I'm certain it requires meticulously setting up nodes and scripts throughout the levels (including defining for every area of the game what kind of outfit will allow 47 to enter without causing a fuss). There are some character models that are used in multiple levels, but a shocking amount of them are made for and used in just one mission. Same goes for weapons and other instruments of death.

That said, Mass Effect's animated conversation system is pretty fucking outrageous too, and I'm sure they pour an assload of resources into setting up all those conversations as well, with cinematic design, lip sync, facial expression, gestures, voice acting, custom music cues, and even the occasional walk-and-talk sequence. But notice one thing about that, almost every player will see almost every conversation in Mass Effect 2 - yeah there's a fair bit of optional stuff, but nothing compared to how many incidental environmental or character details in Blood Money will go unnoticed by all but the most attentive player. How many times did you have to play through the suburbian mission before you realised you could get rid of a body by throwing it in the garbage truck? I'm not confident that most players played each of Blood Money's missions more than once.

I shouldn't be surprised if IO really did have a sublimely talented dude on this one. Of course, Jonas had a weeks internship there once, so he probably knows.

That was during the development of Silent Assassin though, it wasn't as big of a problem back then because assets were cheaper to produce and the levels hadn't yet reached quite the insane level of multiple solutions as in Blood Money.

It's mostly based on the few contacts I have within IO though. That sounds kinda badass but it's sadly not quite as direct as I'd like - one of my business partners has friends in IO, and one of my university supervisors (plus a former teacher) work there. I have heard mention that the cost/benefit of the Hitman games was the reason they tried to develop some more straight-forward franchises such as Kane & Lynch and Mini Ninjas, and now that it didn't work, IOI are apparently looking very seriously at how they're building the Hitman games in order to try to bring down the overhead so the series can remain commercially viable. It's second-hand information, so don't completely take my word for it, but it does make a lot of sense

What do you mean now with "only them has bad weather not anybody else"?

It's a localised snow storm, it exists in a relatively short radius around the player, and nowhere else. When the player enters a shelter such as the highway tunnel, the cloud is turned off and precipitation nodes at each exit are turned on instead, so it still looks like it's snowing everywhere outside though it's only snowing a couple of feet outside the tunnel exits. The rain in Forum City works exactly the same way.

[AEmer]

What should be noted about most such tricks is that they take far longer to implement than the "naive" solutions, but that they're often several, even many sometimes, orders of magnitude less demanding.

Anyway, Jonas, I hear you...and the AI system is usually done by one or at most two programmers, it's not actually that bad.

Hitmans system does require a lot of additional scripting though; area markup by the map makers, essentially setting up various "zones", to which the AI will then respond dynamically.

It's nothing that can't be set up in a matter of 1-2 days for any one level though, so I'm pretty sure that's not where the expenses are. And animation...well, there's a ton of reuse in animation, and motion captured (bad) animation is pretty cheap to crank out...perhaps the biggest animation expense is rigging the character models in the first place, so far as I know.

There's tools like inverse kinematics that you can apply to a rigged model which will let you prototype keyframes real fast.

But you're fundamentally right, good-looking animation can be quite expensive.

The thing is, the price for animation and ai scripting should have, if anything, gone down, whereas the price for market-level 3d models should've gone up, right? The tools for the first two have improved vastly, whereas the demands for them remain relatively the same as they were in (say) Hitman 2... I mean, yeah, bloodmoney used _more_ animation and AI scripting, but each bit of work, each AI script or each keyframe, oughta be cheaper today than back then.

Perhaps the exception is that now, we can do full-digit animation rather than hand-animation and the more advanced 3d models mean there needs to be more advanced animation to keep up.

Do you know what the budget for blood money was?

[m3rc1l3ss]

I don't know if any of you have seen this article, but http://www.next-gen.biz/news/eidos-mont ... boss-error .

Nice to see a developer actually step up and say that something was a mistake for a change...

[fantsu]

OK, I at last finished HR.

What an bad game, nothing to remember, no reason to go back.
I guess thats it.
See you in DX4.

[Jonas]

What an bad game, nothing to remember, no reason to go back.
I guess thats it.
See you in DX4.

You are completely wrong.

That is all.

[gamer0004]

What's remarkable is that on a forum full of Deus Ex fans, only a few months after release nobody is talking about it or discussing it. I think that says enough.

[Jonas]

If they put out an SDK we'd talk about it again.

[EER]

It's like how we're not talking about IW. No SDK = no reason to talk about it after the fact. I wish Eidos would see that, but maybe they don't want us to talk about it.

[gamer0004]

I'm not so sure... Many discussions about DX were about its characters and its setting which is unrelated to there being an SDK or not. There were even more discussions about IW than about HR.

[Jonas]

But most of the discussions about IW were about how bad it was.

I think it's true though, DXHR isn't as groundbreaking or momentous as DX1. It's a sequel, and it's not staunchly trying to reinvent the wheel the way IW did. It's happy just doing what DX1 did, slightly better. That's not a thing that makes you want to keep talking about it or discussing the implications of its existence for years.

It's also a more focused game than DX1, not as full of incidental stuff largely unrelated to the central plot, which means there's less random threads to unearth and discuss. DX1 and 2 were about conspiracies and secrets, the high-level concept of those games fostered very complex background narrative with side threads running out of it and trailing off in every direction. DXHR is about one central technological/ethical discussion, and it mostly sticks to that. There are far fewer loose threads, so it doesn't encourage embellishment and speculation to the same degree.

That doesn't make it a worse game or a less interesting game. It's just more self-contained.

[gamer0004]

But most of the discussions about IW were about how bad it was.

I think it's true though, DXHR isn't as groundbreaking or momentous as DX1. It's a sequel, and it's not staunchly trying to reinvent the wheel the way IW did. It's happy just doing what DX1 did, slightly better. That's not a thing that makes you want to keep talking about it or discussing the implications of its existence for years.

Yeah I really have to disagree there. IW took DX and tried to make it better, more focused. The result was that it didn't really work, because the random crap was what made DX work (irrelevent dialogue, the large amount of characters which resulted in less character development than otherwise would've been possible, skills which had the same effects as augs, multitools and lockpicking and hacking &c.) by making it more immersive and interesting. This is all in line with Harvey Smith's ideas about what he wanted to do with IW and how it turned out (at one point he said something like, in Deus ex "I can be like aquaman by picking the right skills and augs", which wasn't possible in IW because there were fewer options. He also said that by having so many characters not all of them were fully fleshed out, like Bob Page being a geek which doesn't really show in the game).

It's also a more focused game than DX1, not as full of incidental stuff largely unrelated to the central plot, which means there's less random threads to unearth and discuss. DX1 and 2 were about conspiracies and secrets, the high-level concept of those games fostered very complex background narrative with side threads running out of it and trailing off in every direction. DXHR is about one central technological/ethical discussion, and it mostly sticks to that. There are far fewer loose threads, so it doesn't encourage embellishment and speculation to the same degree.

That doesn't make it a worse game or a less interesting game. It's just more self-contained.

This is related to my point as well. IW took DX and tried to make it better (better meaning "more focused", both in terms of gameplay and things like character development). HR took DX (that is, the name of the franchise and its world) and made a different game with only superficial similarities.* Which is exactly why it was both more focused while still being fun/good (to many people), but it also mean they shouldn't call it Deus Ex.

*Bold statement, I know. But I think IW showed why this is true: very similar to DX, yet dropping complexity/random crap = fail, HR, dropping complexity/random crap = success. A true Deus Ex game cannot be successful when dropping the unrelated stuff because it's all about an interesting, believable and immersive world. Dropping complexity or the unrelated stuff results in a setting which is certainly not really immersive (though the game might be while playing it) or believable, and, but that's a personal preference, dull.

[Jonas]

HR took DX (that is, the name of the franchise and its world) and made a different game with only superficial similarities. Bold statement, I know. But I think IW showed why this is true: very similar to DX, yet dropping complexity/random crap = fail, HR, dropping complexity/random crap = success.

I don't even