Quake3World

o'dium wrote:What, just because I didnt piss myself with excitement that id software are planning another game based on graphics and not gameplay...?

You don't know this. The same as you've posted only rendering tech videos/screenies they posted only rendering tech videos/screenies. The gameplay of overdose and rage look equally good to me from what I've been able to see.

I wonder how this will pan out for the engine choices of games companies. This kind of engine means presumably that there's no longer such a thing as an FPS engine or a flight sim engine, it should scale well between almost any game engine tasks. The level of sophistication involved also means that in-house engines will surely become rarer or cease all together (I'm surprised at the number that are still made) and companies will license from one of a few engine providers.

As a comment on the engine itself, sorry obsidian but I've got to agree with O'dium's POV: so what? Give me gameplay or give me graphical style, high-powered engines seem to be mainly concerned with vomiting detail onto every surface, regardless of how that looks.

Hopefully improved work-flow can lead to a bit more creativity and focus on gameplay, to me that's the main positive of this tech.

Here's Olick's Siggraph presentation slides. It's pretty indepth, most of it deals with optimizations and parallel processing techniques. He starts to bring up SVOs in the second half of the notes (pg. 126) after pointing out some of the limitations of traditional polygon based rendering when it comes down to parallelization. Make sure you mouseover his notes on the top-left corner.

Also Wikipedia article on voxel. What amazes me is that in the traditional sense, voxels give you that idea of a super pixelated "Lego block" structure since they were on a uniform grid, not this current scalar and smooth appearance as demonstrated in the video.

From what I understand, it seems as if SVOs are raycasted for occlusion culling, so like Castle says, it could be "fool-proof". It'll probably throw out most of the need for portaling (probably still necessary for dynamic polygonal stuff).

As far as workflow goes, I think this kind of streamlining will eventually improve all aspects of the game. We spend much too much time optimizing maps to make sure that everything runs at smooth frame rates, optimizing polys, rendering normal maps, setting up portals, cutting out entire sections or ideas because they are too taxing on the system. This could be one of those things that really makes game design simpler by tossing out all the traditional technical limitations that a normal artist or level designer has to worry about so that they can focus on the important things, making the overall visual experience better, making better gameplay.

pjw wrote:Y'know, I may get slammed a bit too, but I honestly agree 100% with everything in o'dium's initial post. In fact, this:

o'dium wrote:I just want a dev company now to say "you know what? Games look good enough. Lets move onto what makes the gameplay better..."

is going into my quote file.

(...)

Talent and passion will never be handicapped. People with talent and passion learn and adapt. A great artist is a great artist, no matter the tools and techniques.

Actually, I find the tech less and less important lately. All recent engines can look great on their own ways, so, despite their fights for being the state of the art in rendering, the artists and the gameplay are more and more important now than ever, imo. Perfect proof is HL2 source engine, which can be considered by all meanings outdated, and all the games based on it look superb compared to anything, and play better than most other games.

BTW, I wonder what was the file size of that player model SVO data. Storing the raw 3D volume has to be insane.

BTW, I wonder what was the file size of that player model SVO data. Storing the raw 3D volume has to be insane.

From what I understood, the whole SVO thing also results in surprisingly small file sizes (relative to what you would think) because only physical data is stored (empty space between objects isn't stored). That may have been the voxel hash table article by Microsoft I was reading though. Im pretty sure it was the SVO one.

obsidian wrote:Talking to o'dium is like talking to an infinite loop.

and listening to you is like listening to one

he made a good point, which a lot of people would agree with (mostly because it's true) and which is entirely relevant to this thread. this is nice tech, but getting all moist over it is like raving over the in-flight movie selection of a plane with no wings

carmack should just hand off the tech to valve and go back to building spaceships; that way we'd get a decent game out of it

So we're back to the bash Carmack fashion. Doesn't matter we all stick to his work again and again. He sux. Of course!

Who or what is valve?

O_o

pjw wrote:....Talent and passion will never be handicapped. People with talent and passion learn and adapt. A great artist is a great artist, no matter the tools and techniques.

Right, but look at how many of them there are; look at current tech and how studios are gasping for talent to produce content for normal mapped engines, there aren't enough people to go around becasue artists haven't caught up with the tech, still, 4+ years on and in an environment where you can 'cheat' by using converted 2D artwork.

Don't get me wrong, this is absolutely fascinating tech, and although years away still, and looking at it from an amateurs point of view, the outlook from a production point of view doesn't look rosey.

Of course odium has a point about the need for good gameplay, but I don't really see how that's relevant in this thread.
It's like saying there's no need for Bluray, because the focus should be on making good movies. Yeah.

This just seems like an exciting 'new' technique that increases model fidelity.
From what I understand about it (which, admittedly, isn't much), it shouldn't dramatically increase workload.

Isn't it the case now that modellers create high poly models that then get bogged down and simplified into low poly ones for in-game use?
Sounds great to me if that simplification isn't necessary anymore, and models can be used in all their original, detailed glory.
Bring it.

The engine wars have ended many years ago and have remained so for quite some time. I would not say that things have become entirely stagnant, I see engines doing amazing things these days.

However its hard for me to not take notice when the only people still acting as if the engine wars are still going full force are generally only competing with themselves like hamsters on wheels.. lol..

I think that it could be a good thing for a para dime shift in the age of graphics. Perhaps we will be lucky and see it offspring possibly be new genre kings! Who knows.

And one of the things I find oddly amusing about the criticisms of Id for producing a simple quake style first person shooter is that it has been so long now I believe that we NEED a game like that again! My friend refers to most modern FPS games as the age of the assault rifle.

Most of the new players on consoles these days don't even know the concept of the simple cyclic, and gamey find key open door mechanics that we once scoffed at LOL..

If Id made a new engine and basically cloned quake 1 and stuck to their guns considering modern FPS staples such as advancement through leveling and perks as seen in COD4 with achievements ect with a purely oldschool approach it would almost be like a brand new idea to a huge portion of the market!..

You mean no assault rifles and really tight E-sport 1 vs 1? Say it aint so!
Its bell bottoms man we can come full circle now I swear!

corrected my englerch a bit ><

Explain to me how Donkey Kong Country 2 was a better game than Quake 4, and *MAYBE* you will have a point.

The point is, Quake 4 was great to look at (If you get over the fact it looks nothing like Q2), but basic in, well, everything. See something in front of you? Well just hold "Fire" until it goes away. Different weapons just mean you hold for longer or shorter times.

The point is, you could be able to shoot a monster in the face and see his head explode in the most amazing way, with blood coming out in the correct place and brains can rot over time... But whats the point when the monster just runs at you in the first place..?

I think we need to leave the graphics behind now, and move onto some kind of dynamic AI. Everything praises a game when it ships with great BOT AI, but why cant we combine the two? Thingsl ike cover, weapon use, using the world against you etc etc...

The simple fact is in all those years since Donkey Kong Country, what exactly have we advanced with, game play wise...?

The problem is, on the flip side of the coin, Mirrors Edge took the leap and tried to do something new, and fuck... It sold what, 30k copies in its first week I heard? Really? I dunno...

o'dium wrote:Explain to me how Donkey Kong Country 2 was a better game than Quake 4, and *MAYBE* you will have a point.

The point is, Quake 4 was great to look at (If you get over the fact it looks nothing like Q2), but basic in, well, everything. See something in front of you? Well just hold "Fire" until it goes away. Different weapons just mean you hold for longer or shorter times.

The point is, you could be able to shoot a monster in the face and see his head explode in the most amazing way, with blood coming out in the correct place and brains can rot over time... But whats the point when the monster just runs at you in the first place..?

I think we need to leave the graphics behind now, and move onto some kind of dynamic AI. Everything praises a game when it ships with great BOT AI, but why cant we combine the two? Thingsl ike cover, weapon use, using the world against you etc etc...

The simple fact is in all those years since Donkey Kong Country, what exactly have we advanced with, game play wise...?

The problem is, on the flip side of the coin, Mirrors Edge took the leap and tried to do something new, and fuck... It sold what, 30k copies in its first week I heard? Really? I dunno...

Why..
Why Donkey Kong Country?
um Ok I see what you are saying..

First of all Quake 4 had other problems. Namely copying the Valve Half-life formula with out creating enough truly unique situations and way too much focus on creating un-compelling scripted sequences. A shark jump of sorts exactly what happened with Doom 3. Two games originaly based around a more cyclic design formula where progression is much more akin to clearing a level or finding your next goal (Heritic/Hexen) with a more open ended progression. You cannot shoe horn the half life formula into these designs. Its folly!

They are both fraught with sensory overload in places it does not belong with absolutely no interest in sticking to the original formulas that made the games good in the first place. Replaced with a poor replicated Half-life formula with out really paying enough attention to what made Half-Life successful in the first place.

Then repackaged it in a sci-fi fantasy universe with humans with assault rifles vs the borg but world war 2 instead.. Trust me Quake 4 would have been less trite had I been forced to progress single player by grabbing giant floating spinning key cards and every team mate was DOA for no apparent reason.

The problem with most modern Id games is that they are not sticking to their guns and creating half baked design formulas around what is currently popular. You cant take the half life formula then try to make large levels, blow all your pacing on poor prototyping because the levels are large, and eventually forgo the entire concept of pacing at all and expect people to love it!

What you are experiencing isn't a inability to progress.. Its blind progression with out proper understanding of the end result. These are entirely different topics almost. Graphics are irrelevant in this discussion because the number one most important thing you have to know when you attempt to replicate a successful formula is knowing EXACTLY why that formula was successful in the first place.

Id CANNOT do half life. They fail miserably. They need to make DOOM or QUAKE and progress in a more logical direction on their own terms using those ideas as a base.

And on that note games have progressed a hell of a lot since Donkey Kong Country. Conceptual wisdom finds its basis just as much with in modern design concepts as it does with the latest rehash of an old idea. The market only wants its innovation in safe and familiar packages anyway. On that same coin would you criticize a novel because it is still using chapters?

And for that matter many genres exist today that were arguably not present when Doom and Quake were the genre kings of Ye old days. This isn't about one step forward for all of mankind anymore. Each dev house has to buckle down and pay attention to where these steps are taking them. Logical progression has long since moved from pushing the boundaries of what can be done in video games to selling a complete product with a well thought design philosophy.

Edit: Case and point. You mentioned Mirrors edge? Mirrors edge follows the Half-life formula but breaks tedium of progression by making the act of navigation itself more interesting while also allowing for some sequences to include combat with disposable weapons. It is a very well paced game both because and in spite of its simplicity.

If the new graphics can come at a small cost, and provides more simplistic pipelines it could be of great value. I LOVE fool-proof systems that just work. Bring it on.

and really tight E-sport 1 vs 1?

Blizzard might pull it off, no-one else will.

ix-ir wrote:

and really tight E-sport 1 vs 1?

Blizzard might pull it off, no-one else will.

Actually blizzard is very unlikely to pull it off considering the natural tension between specialized class based mechanics and 1 vs 1 match ups. Considering that 1 vs 1 is the most ideal set up for a true E sport setting because the number of errors your opponent can make before it has an effect on the match is at its lowest common denominator I feel personally that it is too important to omit. This is not an option for Blizzard.

Id software is FAR more likely to pull this off because there original system does not involve such a class hierarchy or grind based progression.

All id needs to do is recognize the defining differences between an E-Sport environment, and open ended war environment and accommodate those differences allowing for casual players and hardcore players what they want at the same time.

For example E-Sports and Grind do not mix. Open war and PVE mix very well with grind however and offer casual players a tangible feeling of progression. Its a retention carrot for casual players. Hardcore PVP players just want to see the better man always win and care only about each individuals talent and skill.

Id has been, and will always be in a better position to pull this off, over Blizzard.

WoW is not the only game Blizzard have made and in any case the design process is what matters. Blizzard set out to produce an MMO with WoW, not a specialized E-Sports game so I don't see it as all that relevant, other than that they were phenomenally successful at meeting their design goals.

Id don't have the knowledge to make a good 1v1 E-sports game any longer. They've lost their touch or focus and the goal posts have moved on while Blizzard have a very strong track-record of community driven tweaking to produce highly polished games with Diablo, Warcraft 3, Starcraft and WoW. Another indication- the QuakeLive beta isn't that impressive so far, Id have some very talented people but not quite the right meeting of minds or power in the wrong hands to produce what we're talking about.

this tech is interesting... but when do you ever have time to stare point blank at a finger or any other part of a high detail model... unless its dead. this new super high detail is only good for cut scenes and screenshots. little else. i agree that ai needs more work... and focus should be fully on gameplay.. not screenshots to sell a product.

ix-ir wrote:WoW is not the only game Blizzard have made and in any case the design process is what matters. Blizzard set out to produce an MMO with WoW, not a specialized E-Sports game so I don't see it as all that relevant, other than that they were phenomenally successful at meeting their design goals.

Id don't have the knowledge to make a good 1v1 E-sports game any longer. They've lost their touch or focus and the goal posts have moved on while Blizzard have a very strong track-record of community driven tweaking to produce highly polished games with Diablo, Warcraft 3, Starcraft and WoW. Another indication- the QuakeLive beta isn't that impressive so far, Id have some very talented people but not quite the right meeting of minds or power in the wrong hands to produce what we're talking about.

AH yes
Warcraft and Starcraft are very much an E-sport platform. Yeah and I have to agree that I am fairly convinced that if Blizzard says they will do something they will most likely get it right. However Id already has proven it can create 1 vs 1 E-Sport in an FPS.

The only problem with basing your game entirely on a PVP E-Sport set up is that you immediately shrink your potential player base if you don't offer some kind of PVE or casual single player/coop content.

Anyway with that said.
Quake live not being impressive does not surprise me terribly.
Its not as if the distribution was the main bottleneck that was suddenly relieved with a new business model. If anything, there are fundamental flaws brewing about under the hood of this off beat idea. Mod support is an afterthought?
My head hurts thinking about that considering the history of the game..

One thing I wonder about is how will this level of graphical detail play into the hands of the third party mod/mapping communities? I mean, right now it seems like anything at Id Tech 4/Source and beyond is just an incredible amount of work to do right by the hands of the average player, and bumping up the quality even MORE seems like it would make things more difficult without adding a whole lot of gameplay to go along with it.

I just know that trying to find third party maps for Quake 3 was as simple as going to LvL. For newer games... I'm not seeing it as much now at all.

WHAT!! wrote: I just know that trying to find third party maps for Quake 3 was as simple as going to LvL. For newer games... I'm not seeing it as much now at all.

But that's not because the added level of detail. Making maps for Quake 4 doesn't require any more work than for Quake 3 unless you want to create your own texture set. The difference is that Quake 3 at its time had a huge comunity of both players and creators, which of course produced huge amounts of everything.

If you want to see that for newer games, just look at Valve's direction and you'll find it.

ix-ir wrote:... Id don't have the knowledge to make a good 1v1 E-sports game any longer... Id have some very talented people but not quite the right meeting of minds or power in the wrong hands to produce what we're talking about

It's got nothing to do with that, it makes absolutely no finacial sense to make games that cater to just one (rather small) section of the wider gaming community.

jal_ wrote:
If you want to see that for newer games, just look at Valve's direction and you'll find it.

I've been looking for decent third party stuff from Source games, and I'm just not seeing the same level of quality that I saw in quake 3. Maybe mappers are just more spread out since once you're into Unreal, you don't really go back to Quake, and vice versa.. that kinda thing.

Yeah, Valve promotes the creation of mods for source games, so content creators are more grouped and don't distribute their own works alone as much as it happened at Quake. The amount of texture sets or maps for source is probably smaller than at q3, but the amount of total conversion mods is much bigger.