User comment history
< Newer [ 1 ] Older >
| News Comments > Sunday Tech Bits |
 |
| 24. |
Re: Source vs Doom |
Jan 30, 2005, 21:18 |
CyberA(rtist) |
|
|
Couple things, some already stated here.
BSP - Yes, nearly all popular 3d engines in the past couple of years have used BSP in some form. Both the Doom 3 Engine and Unreal Engine 2.0 use a form of portals for actual occlusion. I believe the Source engine does as well. Both the Quake engines and the Unreal Engine used nearly all geometry as occluders (to hide other geometry in the world, as well as some basic portal-like zoning of areas to make the calculations faster. When games started using higher polygon counts and 3d cards got so fast it no longer became efficient to use every world poly as occlusion, and so they both moved to portal systems. BSP is now mainly used to store the basic polygon data as well as world collision data for all 3 engines.
Lighting - Doom 3 uses unified lighting, that's easy enough. Your normal distance / angle fall off calculations that nearly every game uses with normal mapping or bump mapping, per pixel highlights, other OpenGL shader routines and volume shadows. The only thing new there is the volume shadows, and even that's been done before. Shadow volume shadows have existed for decades in the form of science papers, etc. The only thing new about them is Carmack just looked at the math and realized it could be simplified so that it was faster and in doing so fix some of the issues it's had in the past implemenations.
Half Life 2 uses lightmaps with normal maps or bump maps with spherical harmonic lighting data compressed in to the lightmap information. On higher end video cards it means that walls that would normally look flat actually use normal maps just like Doom 3 does with correctly lit directional lighting. Also means the flashlight and other dynamic lighting looks pretty damn good. Half Life 2's dynamic lights are nearly identical to Doom 3's, just minus the real time shadow casting (which is a rather large portion of it, to be sure). Not entire sure how their character lighting is done, but it looks like it's just your normal point + directional lighting with nice shader effects (and highlighting), and maybe a little of the spherical harmonic lighting information to smooth things out.
Physics - Doom 3 has it's own system, writen specifically for the game. It's pretty good too. They just don't use it often apart from the random box, ragdoll, and falling wall panel. Half Life 2 uses Havok, which is pretty damn good too. Both are probably about as good, and it really just comes down to how well they're implemented by the team, how the joints are positioned, how the weights are assigned, what the gravity and forces are set to. Havok might also be slightly better with online play.
Open Enviroments - With Doom 3 they decided not to add terrain tools, all the outdoor enviroments are modelled in some 3d package and imported in just like most of the world props. Half Life 2 does have terrain tools, and so it's a lot easier to make large open areas quickly. At the same time there are two main reasons why Doom 3 doesn't have outdoor areas. One, it wasn't part of their design for the game and two, Doom 3 doesn't do well with large numbers of lights in a scene; each light in a scene, and the ammount of area each covers (and overlaps) is much more expensive due to the fact that they area all real time. Since nearly all of Half Life 2's lighting is "baked" in to the world, and those few dynamic lights are marginal in it's effect on performance in comparison, doing large outdoor areas isn't a big deal. Doom 3's poly counts are also a lot lower overall than Half Life 2 on similar hardware since the higher the poly count, the slower volume shadows are.
Networking - Doom 3's networking "issue" has nothing to do with the graphics code and everything to do with physics. Basically all of Doom 3's game physics are handled through their realistic physics simulation, which is very slow. To keep everything synced together, the game is already locked at running 60 fps because of the physics (the game may actually try to run the physics at 60 "fps" even when the rendered framerate is lower). For multiplayer all the different peers need to be perfectly synced so that the physics on each person's computer shows up the same.
Half Life 2 has two separate physics systems in place during the game, the "game" physics and the physics simulation. Game physics in general tend to be far less exacting than "real" physics, meaning a lot less data is being processed, and in turn needs to be sent over the network during a multiplayer game. The realistic physics simulation is then done on the server side, and the client side. The two are checked against each other every few frames to make sure they're still the same, and the server will force the client side if they don't quite match up. Half Life 2's implementation will allow them to sometimes show up in slightly different positions from time to time, where as Doom 3's will never allow them to get even a fraction off.
Collision - Doom 3 appears to use per-poly collision on their characters, as well as the world. Many games do per-poly on static objects, but few do it on skeletal animated or dynamic objects. It's slower, and not always needed (or wanted). In Doom 3 multiplayer it's possible to shoot a rocket between an enemy's legs where as most other games it would simply hit their collision box, cylinder, capsule and register as a hit. Some games have used multiple cylinders, or closest bone to the point hit on the cylinder for location damage, but most new games have started using collision boxes, cylinders, capsules, spheres, etc. that are linked to the body giving a semi accurate version of the character's body. Half Life 2 does this, so it is also possible to shoot a rocket through someone's legs. Doom 3's is definatly more accurate, but considering the fact that Counterstrike Source's gameplay is a lot more about trying to hit exactly the right location that Doom 3's is, and that they use hit boxes to aproximate the models vs. per poly, it seems to work just as well and is far less expensive CPU and network wise. When it comes to their realistic physics models, both use simplified shapes (though probably allow for per poly) because it's a lot faster. In that article he talks about Doom 3's amazing per-pixel hit detection, which just makes him sound like more of an idiot, since while Doom 3 does do per pixel hit detection, it does it for the menu systems and the in-game video screens, etc. which every menu system for every mouse driven game, ever, uses. Doing per pixel hit detection for things like shooting is just dumb and hasn't been done since the days of 2D since it's A) slower and B) less accurate than just using the polygons that exist in the scene to start with. |
|
|
|
|
|
|
|
| |
| News Comments > EA Buys Criterion |
|
| 17. |
No subject |
Jul 28, 2004, 18:33 |
CyberA(rtist) |
|
|
"This is a great fit" for making EA a crap load of money in engine licenses.
Here's a small list of games that have or will shorly come out that use Renderware.
http://www.renderware.com/clients_titles.asp
The last 2 Grand Theft Auto games, as mentioned before, also used Renderware, as well as hundreds more over the last 4-5 years. Of the handfull of games actually listed on the Renderware site, it's interesting to note that none of them are EA games. By buying Criterion and with it Renderware is going to give EA an extremely compitent team of 3d engine engineers with which they can use for most or all of their 3d games (meaning EA never has to spend money licensing external engines again). It also means that any other publisher who wants to release a game with Renderware has to pay EA, effectively stretching EA's proverbal arms around more of the video game money pot. It also should be curious to see if EA forces people who are using the Renderware engine in upcoming games to put the EA logo on the boxes along with the actual publisher. That could be a huge coup for EA either way. |
|
|
|
|
|
|
|
| |
| News Comments > Ships Ahoy - Catwoman |
|
| 12. |
Re: halle berry |
Jul 23, 2004, 14:34 |
CyberA(rtist) |
|
|
| Hmm... actually it could make a good movie, just not one that'd appear in "main stream" theaters. You know the type, the ones Paul Reubens frequent on occation... |
|
|
|
|
|
|
|
| |
| News Comments > Syberia II Demo |
|
| 6. |
Stunningly short demo |
Mar 1, 2004, 13:39 |
CyberA(rtist) |
|
|
While I loved Syberia, that was possibly the shortest demo I've ever played. Granted it took me an hour to finish, but there's only two puzzles and the difficulty was more in finding where the solution was hidden than in the puzzle itself.
The main problem comes from an issue I have with prerendered adventure games in general. If there's an area where the player can zoom in on, if it appear in two or more different "angles", you're often able to select it from one. Also the general problem of not being able to easily tell what you're able to click on and what you are. You spend a lot of time scanning each area with your mouse hoping your cursor will change.
Oddly, the use of prerendered or better matched real-time rendering of objects makes the game harder. In Monkey Island 3, Grim Fandango, even The Longest Journey, the objects the player could move or pick up usually stuck out from the rest of the world like a sore thumb. Badly aliased edges, flat lightly or badly matched levels actually in some ways made the game better so more of it was spent figuring out puzzles instead of finding puzzles.
I guess if you loved the first one, this one doesn't appear to be any different apart from some nice extra visual touches (scrolling scenes, foreground elements, etc.) |
|
|
|
|
|
|
|
| |
| News Comments > Uru Demo |
|
| 21. |
Impressive visuals |
Nov 12, 2003, 08:56 |
CyberA(rtist) |
|
|
I ran this on my laptop with a Radeon Mobility 9000, ran surprisingly well and looking stunning even with many of the settings turned down / off.
If people get really annoyed with the controls, even in first person, move your mouse cursor to the lower right of the screen and go to the options. Turn off "move to point clicked" or whatever. I found after that it was much easier to control in first person view. |
|
|
|
|
|
|
|
| |
| News Comments > LotR: Return of the King Demo |
|
| 61. |
Re: Plays like the first game |
Sep 22, 2003, 23:33 |
CyberA(rtist) |
|
|
Actually, I agree that there is room on the PC for shallow, instant-gratification games ... it's called Bejewelled and Sticky Balls ( ftp://ftp.zedtwo.com/sticky.zip ). PC gamers tend to expect more deep gameplay from a game they're paying $50+ for, even if those games usually degrade to braindead tactics since they always seem to be the most effective regardless of what high gameplay techniques are in place (but that's another discussion).
As for the utter lack of deep gameplay in PC games, that's our fault as gamers as much as it is the fault of the publishers and developers. GBA is extremely easy to develop for as far as overall risk. Quick development times, simple graphics, and low cost (plus massive installed user base) make GBA games more likely to try new things. PC games and "triple A" console games tend to stick to the rule "if it ain't broke, don't fix it" which usually turns in to "lets do the same game we did last time, but with less energy in making that gameplay seem unique." We crave super duper graphics in our games, which takes a long time to produce and eats up the budget for what would otherwise be used for playing around with new gameplay ideas. Most games can't afford to be in production for longer than a year or two at most, so often the artwork side of things starts as soon as the project does leaving little time for anything but cut and dry games. Only those few "big" names that have the money to try something new ever do, and those games usually bomb bad (Sims being one of the few exceptions) which lead to those big names just retreating back to what got them big in the first place.
As gamers, we've been spoiled by such amazingly polished games that every time one comes out that doesn't look like freshly polished crystal we instantly disregard it as being a shoddy product. Those games that try and dig deep and do something new and exciting tend to be those games.
Ah well, now I'm just venting my frustration of the current state of the gaming industry. I'll stop now. |
|
|
|
|
|
|
|
| |
| News Comments > LotR: Return of the King Demo |
|
| 56. |
Plays like the first game |
Sep 22, 2003, 08:55 |
CyberA(rtist) |
|
|
I'm assuming most of the people here haven't played, or likey don't even acknowledge the existance of the first game. Concepts like the parry, physical attack (odd name for what is otherwise the "knock down" button) and any combos (which, if they exist, eluded me beyond the usual fast, fast, fast, and strong, strong combos) are some of the more difficult moves in the first game to use well. The only real benefit from their use was extra experience points which were used to upgrade your character.
This demo doesn't let you do that, nor does it take any time to explain what or how to go about doing anything. Also, the camera issues that've been mentioned before I can only hope are a mistake by the developers. The first game had some really great camera work for the most part, with the exception of some of the more enclosed areas like this one. I can only assume they chose this level on the fact that it was easily self contained to make for a smaller file size. I also wonder if the default camera is the one intended for coop play and was mistakenly set to be used by default instead of the other, while still horribly flawed, obviously superior camera setup.
I also question EA's reasoning in putting this game on the PC. It so definatly isn't a PC game, all I can figure is EA saw how Atari managed to sell a bajillion copies of Enter the Matrix and figured they'd cash in on the PC market in the same way. I dare say they're likely to sell tons of copies of this game for the console, but it might of been a mistake to release this demo to the generally "better informed" PC market who are likely to of gotten a foul taste from this.
The first game had a good mix of brain dead button smashing and advanced combat techniques that made it extremely easy to pick up, but painfully difficult to complete when they started introducing enemies that required those more advanced skills to defeat. Anyone who thought this demo sucked and own a console, I'd urge you to go out and rent the first game and make your judgement on that. If nothing else, it'll help explain what all the random crap on the screen means. This version appears to be more of the same but with more dynamic in-level events and coop play.
I think #50 said it well... it can be quite theraputic to go and slash into swarms of orcs for no apparent reason. Sometimes it feels like PC gamers these days look too hard for "deep" gameplay where none is needed. |
|
|
|
|
|
|
|
| |
| News Comments > DOOM 3 Hardware Requirements |
|
| 80. |
Re: Real-time shadows |
Jul 18, 2003, 16:45 |
CyberA(rtist) |
|
|
| Every surface in Doom3 uses shaders. All lighting and shadowing is done via shaders. In that respect, Doom3 has a higher percentage of shader-using polys than Half-Life 2. At the same time, I agree that Half-Life probably has more shaders total, but that's like saying Everquest has more textures than Quake 3, the scope of the projects are much different, as are their intended player experience. |
|
|
|
|
|
|
|
| |
| News Comments > Longest Journey 2 Q&A |
|
| 9. |
Full 3d, fixed cameras .... |
May 25, 2003, 19:26 |
CyberA(rtist) |
|
|
| ... and cameras on rails. I have to wonder if they've been playing Ico or one of the Devil May Cry games. If so, this could be good news. Ico and Devil May Cry feel like a natural progression of the classic prerendered adventure games. I'd guess this game will probably resemble the more recent Resident Evil games which combine full 3d with mostly fixed camera systems, and that doesn't sound like a bad thing to me. You get the cinematic visuals of the prerendered images, but the dynamic element of full 3d. The first TLJ had some amazing visuals, and if they can get those happening with the stunning camera work in Ico or Devil May Cry, I'll be extremely happy ... even though I'll buy it even if it doesn't since the first one was so good. |
|
|
|
|
|
|
|
| |
| News Comments > Tech Bits |
|
| 11. |
IBM 75GXP |
Oct 25, 2001, 03:45 |
CyberA(rtist) |
|
|
I had my 75GXP 45Gig for several months now. For the first couple months of use it worked like a dream. After moving to a new location the hard drive started making the dreaded click of death as well as refusing to even spin up at times. After about a month and a half of this, along with several surface tests with what can only be described as ghost bad sectors, it stopped having problems and appears to be running fine once again.
As soon as I get a chance, I'll definitely be changing to the 60GXP, which has had a far better track record. I feel like I'm running on borrowed time for now, and I'm about to move again.
-------
-CyberA(rtist) |
|
|
|
|
|
|
|
| |
13 Comments. 1 pages. Viewing page 1.
< Newer [ 1 ] Older > |
|
Twitter 
