Quote:Film directors have been all over this stuff, for a long time now. They've gotten really good at it. It's not going to translate to the video game camera without a bit of work, but isn't it time we started listening to them?
no as someone pointed out games!=movies, whilst both are entertainment, the main thing in a game is for it to be fun/easy to play, doing arty farty camera shots whilst may look good are gonna piss off a huge segment of the population.
its sort of like deciding what fluffy dice u want in your car whilst having no wheels take this screenshot (choosen cause ive seen heaps of ppl say oh wow etc) http://screenshots.teamxbox.com/screen-hires/44719/Unreal-Tournament-2007/ whilst the individual parts look good the overall picture looks crap, it looks completely 2d, ie the cars + person look like theyve been glued over the background picture. gotta get wheels on the car afore u worry about the dice
Quote:Original post by zedzeek no as someone pointed out games!=movies, whilst both are entertainment, the main thing in a game is for it to be fun/easy to play, doing arty farty camera shots whilst may look good are gonna piss off a huge segment of the population.
Not all games are Tetris. Would Resident Evil have been Resident Evil without its distracting-at-times camera angles? Would System Shock 2 have been as affecting if they hadn't made it dark and difficult to see? Would Anachronox's bizarre geometry have been as interesting without the swaying, serpentine camera motions introducing new areas? If nothing else, the proliferation of cinematics shows that people DO want "arty farty" stuff.
Perhaps the problem is not the limitations of an in-game camera, but the limitations of gaming genres which are all about literal and literalized action. The majority of games devolve into simple navigations of spatial relationships between objects, raised in difficulty by increasing the number of targets or shrinking the response time window and so forth.
For a video game cinematography to be practical and profitable, the objective and very mechanism of the game must become the manipulation of abstract quantities, such that the specifics of space become secondary at best. The obvious interpretation is an advocacy of "dramatic games" in which the player navigates social relationships, seeks and peddles influence and so on, with the camera being programmed to display or reveal portions of the scene (or overall narrative) in response to variations in the quantities being manipulated - show the face of a soldier who lowers his gun in response to reason, the eyes of a scorned lover, etc.
While the evolution of dramatic games is both necessary and inevitable, there are genres that are available to us today which can be divorced from the strictly spatially literal, if we are so bold as to tamper with convention. The majority of the resources employed in an RTS or TBS can be or are inherently abstracted: troop movements and activities can be reduced to orders issued, and represented in aggregate as the Chain of Command; munitions, provisions and reinforcements can be viewed as the Supply Lines; enemy positions, strengths, research operations and so forth constitute Intelligence; and the relationships between the various entities that run a war - the War Machine - are inherently abstract. The generals who run the military, the politicians who use the war to achieve their own goals, the intrigues within and between both groups as individuals strive for power and influence, the economic ramifications for military and civilian alike, the internal political consequences of war, and specifically of casualties...
For game genres that rely on strict spatial reasoning, cinematography is useful only as an after-effect (replay) or in limited doses. For games that are based on the manipulation of abstract quantities, though, preferrably minimally represented as "influence bars" or "poll percentages," and particularly for games that are built on the interaction between the gamer and NPCs, cinematography represents a significant design opportunity.
Oluseyi: You make some really interesting points. I guess it ties into what I was saying about needing to refine the control scheme, but you've taken it much further. Your proposal for RTSes, in particular, strikes me as very promising. I can imagine a nineteenth century warfare game where instead of somehow having a godlike top-down view of everything happening everywhere you've explored, you receive information from reports and personal observation. The camera would then become a subtle indicator of troop morale, the level of action, relative military strengths, etc.
Quote:Original post by takingsometime The primary issue with incompetence in a camera system in a game is due to the methods used to derive the camera position. They're computationally efficient, but ghastly at providing good control. In order to rectify current camera systems, I personally believe that you need to look beyond the polar/spherical coordinate methods with damping, quaternion interpolation, steering, and proportional controller methods, and look at something with better expressive power. Of course, with expressive power comes computational cost :(. Applying cinematography rules (they're more guidelines than rules) will not resolve (m)any of the issues that currently plague games.
Ah, but that's exactly why they could resolve many of those issues. You said it yourself: we're talking about cameras in the wrong language, using terms like "quaternions" and "splines". We need another layer of abstraction, so that we can use these guidelines directly, instead of just keeping them in mind while coding up our PD controllers. For instance, most chase-style TPCs are afraid of cuts. They'll only do them when absolutely necessary, or not at all. But cuts, we remember from a century ago, are not inherently disorienting. They're only disorienting in the hands of a purely utilitarian camera which cuts to get itself out of a bad situation--and thus doesn't have many options open--rather than a cinematographic one which cuts as a matter of course.
Quote:Not all games are Tetris. Would Resident Evil have been Resident Evil without its distracting-at-times camera angles? Would System Shock 2 have been as affecting if they hadn't made it dark and difficult to see? Would Anachronox's bizarre geometry have been as interesting without the swaying, serpentine camera motions introducing new areas? If nothing else, the proliferation of cinematics shows that people DO want "arty farty" stuff.
of those games i only know re (never seen it in the flesh though), but i was under the impression that the camera was pretty static ( i heard u didnt have much control over anything much, a bit like dragons lair ) i dont know System Shock 2 although mentioning lighting (which i agree is very important), its a totally differ mother to camera movement which is what your post was about. on the subject of lighting though, alpthough i dont know ss2 but i do know doom3 which is also "dark and difficult to see" and im sure youre aware the majoritiy of comments about that game are ppl moaning about the darkness :)
Quote:As the anonymous poster said: games != movies
great minds think alike, oh wait gold coast, hmm scratch that ;)
I don't think we'd have to go that far to effectively use camera's the right way to enhance that game eperience. When you look at the Homeworld series for example, there the 'cutscenes' where done pretty well. When mysterious ancient enemies or allies emerge, there's always a nice shot of this event, with a special sound track. The same is done for crucial ingame story events. The 'breaking dam scene' in C&C:Generals on the other hand is an example where the camera work is lacking. If you have a tidal wave rushing towards a village, show it rushing towards a camera that's positioned in the village, not in a default RTS overview shot!
Now, is it hard to do? Technically it shouldn't be, since most camera systems are capable of shooting at just about all angles that would be impossible for a real-world camera. As the OP pointed out, it's much more an issue of how to use the technology the right way so that it actually adds something to the game. And that's the main challenge for any game. So I think the camera issue isn't exactly a new issue, but rather an extra thing on the list 'to do right'.
There may be some merit in using movie techniques to get that dramatic feel in video games, but as pointed out in a gamer's manifesto, cinematic angles take away from the immersion in a game. For RTS's and the like, this might not really matter that much, but for FPS's and most certainly RPG's, I agree with the manifesto... Come to think of it, it's a must-read for anyone interested in games, really. To quote the stuff about camera's though:
Quote:Original text from A Gamer's Manifesto
Let's ban all IAC's (Immersion-annihilating contrivances). These include:
(...)
"Cinematic" camera angles. No, thank you. Understand that we need to see what our character sees. As soon as you start panning the camera around Mario for no better reason than to see the pretty sunset on the distance, we lose control. And here's another tip: If you have a single level where the player's character is required to run toward the camera, send the fucker back for more programming because you're not done yet.
(...)
Chances of that happening...
The cameras in 3D games have actually gotten worse (Mario Sunshine's camera system wasn't half as smooth as Mario 64's) because in the game-making world camera and player controls are decided-on after the game's pretty artwork. When 3D games were new the only question was, "how can we make the controls as responsive and fluid as 2D?" Now it's, "how can we show off these really cool-looking trees? That's what the little sons of bitches care about!"
In short, the first 3D games were designed around their cameras, now they're designed around their graphics.
Food for thought :)
Edited: added the last line of the quote because it's so true
[Edited by - remigius on December 30, 2005 3:13:22 AM]
Quote:Original post by zedzeek of those games i only know re (never seen it in the flesh though), but i was under the impression that the camera was pretty static ( i heard u didnt have much control over anything much, a bit like dragons lair )
A static camera doesn't preclude cinematography.
Quote:i dont know System Shock 2 although mentioning lighting (which i agree is very important), its a totally differ mother to camera movement which is what your post was about.
Not different at all. It's another situation in which visual decisions are made which make the view more dramatic, at the expense of making it less "useful".
You should have been reading my journal a little while ago. Prince of Persia: The Sands of Time prompted a whole slew of posts from me on this topic. If you've got time you may want to try and dig them up.
The camera is all about presenting a slice of information about the game world to the player. This information is presented visually and consumed visually, and consists mainly of information about object state and about the spatial relationships between objects. The player will use this information to make gameplay decisions.
So, problem number one: Frequently the information communicated by the camera is not communicated via any other means. In Prince of Persia, I can hardly tell anything about my current status without looking at the screen - the music might tell me whether I'm in a fight or in a platforming section, and the sound of enemies shouting might tell me whether I've been spotted or whether I can still go for that stealth kill, but beyond that everything else is conveyed visually. If I do look at the screen, as I'm willing to accept that we are a primarily visual medium after all, then the only things I get regardless of the camera are my health and the number of sand tanks I've got left. For everything else - what I'm currently doing, where I'm currently going, etc - I rely on the camera.
If we want to free up the camera for more artistic purposes, we need to begin by offloading the information that the player is relying on it to present to other means. Currently the only open avenues seem to be sound (which is underused anyway), force feedback (which isn't really a PC thing), and fixed screen elements like HUD. Step one is more innovative use of these.
Take the Dahaka chase sections in Prince of Persia: Warrior Within. One of the pieces of information given to the player - not that they really use it directly, but it adds to the tension - is how close they are to being caught by the Dahaka. If this were to be conveyed using the camera, it would have to show both the Dahaka and the Prince in the same shot - and given that the player also needs to see things that are in front of the Prince in order to negotiate them, the only viable shot would be something like a bird's-eye view (and given the wall-running nature of the game, even that would not be very good). Instead, what the developers have done is to use a camera invariant concept - as the Dahaka gets closer, the screen begins to grey out. You get the information about how close the Dahaka is (and as an added bonus, the information is less precise, adding to the tension) without the camera showing you it.
If the information load on the camera is reduced, then we can start looking to present the information we are conveying in more artistic ways. Concepts like the rule of thirds, the idea of 'good framing,' these can all begin to be applied once we're sure that doing so won't be cutting out things the player needs to know.
Another thing that can help is use of multiple camera. Fahrenheit (ok, ok, Indigo Prophecy) has good examples of this. Pick up the demo, and watch the sequence where Lucas is trying to get away while the cop walking down to the bathroom; at the beginning of the sequence, Lucas is in the bathroom and the cop is outside, but you have a camera in each location. The information about how close the cop is getting is presented by the second camera, and the first camera tells you enough to let you get out successfully. If you've not checked it out, you should do.
Richard "Superpig" Fine - saving pigs from untimely fates - Microsoft DirectX MVP 2006/2007/2008/2009
"Shaders are not meant to do everything. Of course you can try to use it for everything, but it's like playing football using cabbage." - MickeyMouse
As my favorite games are console-style RPGs, where I play to experience the story in a film-like manner and don't really care if a cutscene is 10+ minutes long (or 30+, as in Xenogears... -_____-), I love the cinematic feeling you can get with great camera angles. However, as I struggle my way through Shadow of the Colossus, I've seen just how terrible camera work can get. With that game, they not only have a complete analog stick devoted to camera movement, they also have two buttons. And as cinematic as it can sometimes look, it is a CONSTANT struggle to get my character to actually move where he needs to (though the control scheme might cause at least some of this).
I have to a gree with a couple of the above posters that in any game where the focus is on movement and reaction experiencing the world is best done through a camera designed to facilitate such movement. If you want to see neat effects in such games, position those effects in front of the camera, don't position the camera to see those effects.
However in RPGs, adventure games, and other types of games where timing button presses and jumping across platforms isn't as important as experiencing the game world, by all means, go wild with the camera. Prince of Persia's camera COULD be cinematic without getting (too much) in the way of the game because the controls were responsive enough NOT to require precise movement or jumping (the combat got a bit annoying, but the camera mightn't have been the cause of this). The difficulty in the game came from figuring out what you needed to do in a room, not from actually doing it. If you were facing in the general direction of a pillar and pressed the jump button, you'd go there.
"For sweetest things turn sour'st by their deeds;Lilies that fester smell far worse than weeds."- William Shakespere, Sonnet 94
