How far can graphics go???

[IronMongeR]

Imagine the cost and manpower in creating movie/real life like quaity graphics! Development teams would have to be so much bigger.....surely it would effect the cost of the game? I think that recently game prices have slowly been getting better and better, with more budget priced releases etc. Maybe it's just because I have more money than when I was a 10 year old scraping coins together to buy a Mega Drive game...?

[kisrael]

Personally, I think Halo and GTA:VC are some of the best examples of getting the balance just right.

 

Not sure about Halo, players can still tend to 'slide' too much.

I was thinking more in terms of...like, realism of the overall story. I mean, come-on, a battle like that being settled by ground forces and small squad combat?

 

Good points about the other games tho

[Fretwobbler]

Imagine the cost and manpower in creating movie/real life like quaity graphics!  Development teams would have to be so much bigger.....surely it would effect the cost of the game?

 

You are correct. Although the cost of the game depends on the size of the audience. Thats why you can see Titanic for $5 at the Cinema. If Titanics audience was the same as, say, Civ 3: about 3 million, it could cost you $1000 to see the movie at the Cinema.

 

It doesnt just affect retail games, online gaming will be affected.

 

This is one reason why consilidation is becoming more prevalant, smaller studios dropping like flies and Electronic Arts and Gamespy are laughing all the way to the future because they have budget and user base so big they could go to war and win.

 

IMO, 20 years from now game releases will be 95% of commercial stuff coming from huuuuuuge publishers and $billion budgets. 4 or 5 'big' releases a year (all released just before the major annual awards ceremony), a few bits of chaff and the rest will be freebie indipendant web games written in Flash MX 2024 GTI Pro-Am Platinum Edition.

[Mr.FoodMonster]

If so few games came out each year, nobody would game anymore. Plain and simple. I hate sports games, and they tend to be popluar. I like FPS, others dont. I hate tactical squad type games, others dont. There wouldnt be enough to make buying a system worth it.

Besides, there would be more then 5 sports games alone. Then you have FPS, adventure, puzzle ect.

[Fretwobbler]

Besides' date=' there would be more then 5 sports games alone. Then you have FPS, adventure, puzzle ect.[/quote']

 

There is an economic issue ahead for developers though Mr.FoodMonster.

 

They can no longer afford to produce that many games because the developments costs have been steadily rising, but net return has remained static. Sure you will get a 'Half-Life' goldmine released every so often but this isnt the norm.

[+Random Terrain]

Also remember that when the game engines of the future become so advanced that things can look as real as real life, the game design tools will also become advanced. It will take work to create a game, but it may not be as time consuming and expensive as people might think. Besides that, there will always be new classic-style games, ugly anime cartoony crap, and other types of games for people who do not like games that look and "feel" so real that you will crap your pants and swallow your own tongue.

[Fretwobbler]

Also remember that when the game engines of the future become so advanced that things can look as real as real life, the game design tools will also become advanced. It will take work to create a game, but it may not be as time consuming and expensive as people might think.  

 

I disagree here, how many man hours did it take to make Titanic, and how many man hours did it take to make Singing In The Rain. Industrial Light & Magic are a growing company, getting more expensive every year, but delivering better special affects as a result. You cant get away from the every-oncreasing cost cycle.

 

and other types of games for people who do not like games that look and "feel" so real that you will crap your pants and swallow your own tongue.

 

These are the ones I cant wait for! Hopefully Doom 3 will take us 1 step nearer

[+Random Terrain]

I disagree here, how many man hours did it take to make Titanic, and how many man hours did it take to make Singing In The Rain. Industrial Light & Magic are a growing company, getting more expensive every year, but delivering better special affects as a result. You cant get away from the every-oncreasing cost cycle.

That's because the technology isn't advanced enough. Do you know how much work they still have to do to get something to look real? Today's movie special effects will look like crap compared to what the generic game engines of the future will produce.

 

I don't know why you are comparing today's lame tools with the tools of tomorrow. I have had this on my web site for a fairly long time now:

 

Soon Our World

We have A.I. game design agent programs that understand speech. If you want to make a game, all you have to do is describe it to the program "conversationally" and it will do all of the work. It will ask you questions and you will just supervise like you're George Lucas, giving thumbs up or thumbs down on various aspects of the game. Anyone can be their own game supervisor. It's like being in charge of a group of highly talented programmers, artists, and musicians.

That's far into the future, but until then, the advanced (easy to use) tools combined with the ultra-realistic game engines of the near future will make the tools we have now look like Tinkertoys.

[Artlover]

Do you know how much work they still have to do to get something to look real?

 

More then they actually bother to do?

 

I've yet to see any special effects that looked really real. I've seen plenty of good effects. Van Helsing, The Day After Tomorrow, so on. But you can still tell it's cg, models, forced perspective and so on. Mostly, it's in the movement. CG objects just don't move right.

 

Then again, I'm the type of guy who when watching movies looks for the bloopers and stuff. Color mismtches from overlayed cells, clocks & watches that never change time, background scenery changes, looking for the seam between live action & painted overlays. So I'm probably a bad choice of person to ask an opinion of.

[Fretwobbler]

I don't know why you are comparing today's lame tools with the tools of tomorrow. I have had this on my web site for a fairly long time now:

 

Soon Our World

We have A.I. game design agent programs that understand speech. If you want to make a game, all you have to do is describe it to the program "conversationally" and it will do all of the work. It will ask you questions and you will just supervise like you're George Lucas, giving thumbs up or thumbs down on various aspects of the game. Anyone can be their own game supervisor. It's like being in charge of a group of highly talented programmers, artists, and musicians.

That's far into the future, but until then, the advanced (easy to use) tools combined with the ultra-realistic game engines of the near future will make the tools we have now look like Tinkertoys.

 

Fair enough, but there are two things that need developing here, the engine, and the game. You talk alot of how easy the development of the end result will be because the engine is so powerful, but what about the development time of the engine?

 

I agree in some respects - once they have created, for example, 1 perfect tree, they can recreate a tree with no work at all....

 

once they have created 100 unique perfect Birch trees, they can create a perfect Birch tree forest with very little work.

 

and once they have created 100 unique perfect trees of 100 differant species of tree, they can do the what the hell they want

 

But how much ground work is there to be done, thats the crux IMO.

[+Random Terrain]

Do you know how much work they still have to do to get something to look real?

More then they actually bother to do?

 

I've yet to see any special effects that looked really real. I've seen plenty of good effects. Van Helsing, The Day After Tomorrow, so on. But you can still tell it's cg, models, forced perspective and so on. Mostly, it's in the movement. CG objects just don't move right.

 

Then again, I'm the type of guy who when watching movies looks for the bloopers and stuff. Color mismtches from overlayed cells, clocks & watches that never change time, background scenery changes, looking for the seam between live action & painted overlays. So I'm probably a bad choice of person to ask an opinion of.

Yeah, me too. My family hates watching movies with me. Every once in a while you will see bits that look real though. It's like crap, crap, good, crap, crap, crap, good, almost perfect, crap, crap, what in the hell were they thinking?, crap, crap, crap, Ok, good, crap, crap, crap, nice, crap , crap, crap. . .

[mos6507]

Also remember that when the game engines of the future become so advanced that things can look as real as real life, the game design tools will also become advanced. It will take work to create a game, but it may not be as time consuming and expensive as people might think. Besides that, there will always be new classic-style games, ugly anime cartoony crap, and other types of games for people who do not like games that look and "feel" so real that you will crap your pants and swallow your own tongue.

 

Movies can be realistic because you are doing a real-time analog scan of the real world down to a 2D plane over a limited field of view. Real world actors. Real life locations (or sets you can build with real materials). No interactivity.

 

The only way to get this into an interactive game (that isn't just FMV shovelware) is to be able to 3D scan objects w/textures a lot better and cheaper than is currently possible.

 

Right now it's pretty time consuming to take a real object and make a photoreal facsimile of it in 3D. Imagine having to 3D scan every object in a house. Okay, what about an entire city block. Or a city.

 

I don't see any way to make an environment that can hold up to close scrutiny without putting some limits on what you can interact with/get close to/etc...

 

What if you want objects that have true volumetrics. Like being able to cut an orange open at any arbitrary point and see the inside. 3D is not ready for that. You practically have to model every individual atom to give the player 100% control over how he wants to interact with the world. And I'm assuming any alterations you make to the world should persist. That takes a crapload of memory to manage the state of every object. Is it scuffed? Is it crumped? Is it burned?

 

What we have as the state of the art now is the Thunderbirds phase of graphics, basically advanced puppeteering. It's not VR and just because you wear a motion controlled helmet doesn't make it true VR either.

[+Random Terrain]

The only way to get this into an interactive game (that isn't just FMV shovelware) is to be able to 3D scan objects w/textures a lot better and cheaper than is currently possible.

In the future, you know that at least one advanced tool will come along that has thousands or hundreds of thousands of pre-scanned images included and you could adjust them to suit your needs. It would be possible to make a game without doing any scanning yourself. The software will be so advanced that you could also design a non-scanned object and have it look real and act real. The software will be able to reproduce any kind of gravity, clothing will move realistically, various types of skin (human or animal or invented alien or monster) will act according to believable rules. Breasts will jiggle, fat bellies will bounce, and skin will stretch.

 

 

Right now it's pretty time consuming to take a real object and make a photoreal facsimile of it in 3D.  Imagine having to 3D scan every object in a house.  Okay, what about an entire city block.  Or a city.

Remember that I'm talking about the future. If the tools become as advanced as I think they will, scanning may not be necessary (except in certain cases). Besides the software's ability to allow you to quickly create realistic object, it may come with many prescanned images (as I said above).

 

 

 

I don't see any way to make an environment that can hold up to close scrutiny without putting some limits on what you can interact with/get close to/etc...

You don't see any way because you live right now. You're only looking at what is available today. Look at how far software and computers have advanced in the last 20 years. As long as we don't blow ourselves up, we will have some amazing tools that will take a lot of the hard work out of the process. It won't hinder our creativity, it will enhance it.

 

 

What if you want objects that have true volumetrics.  Like being able to cut an orange open at any arbitrary point and see the inside.  3D is not ready for that.  You practically have to model every individual atom to give the player 100% control over how he wants to interact with the world.  And I'm assuming any alterations you make to the world should persist.  That takes a crapload of memory to manage the state of every object.  Is it scuffed?  Is it crumped?  Is it burned?

Again, I was talking about the future, so there will be enough memory, the computers will be powerful enough, and the software will be advanced enough. A new type of harmless scanning may come about that will quickly and cheaply allow the scanning of both the outside and inside of an object at the same time. The software would be preset with the knowledge of how an orange, a watermelon, and similar objects will react when sliced, or dropped from a rooftop or whatever you can think of.

 

 

What we have as the state of the art now is the Thunderbirds phase of graphics, basically advanced puppeteering.  It's not VR and just because you wear a motion controlled helmet doesn't make it true VR either.

Right, it sucks. As I said, if we don't blow ourselves up, we will have amazing things. It would be nice if I could live long enough to see and use many the cool things that will come along.

[ozyr]

I remember when it was all about gameplay....

 

You said it....

[+Random Terrain]

I remember when it was all about gameplay....

You said it....

Gameplay is the most important thing' date=' but the addition of great graphics can make a good game even better. You can have great gameplay and great graphics at the same time, but too many game designers seem to focus on everything but gameplay:

 

http://www.randomterrain.com/gamedesign/gameplay.html

I'm constantly amazed at how many games are developed—even at some of the biggest and most established game companies—by teams that have no one responsible for the actual game. There are artists, programmers, musicians, and a producer to coordinate them all, but no one actually concentrating on the business of making sure that the interactive experience is as rewarding as it can possibly be. No one thinking about how it will actually feel to play the game. Instead, the actual "gameplay" will be added at the last minute, when all the graphics are ready. Almost as an afterthought.

 

It's like these developers are trying to invent chess and have created a superb, glossy-looking board and a whole new set of exciting pieces and then sit back and say, "Look! Look at his new board game we've made! Look at these shiny pieces and this state-of-the art board! What a great game this is!" But they haven't thought about how the game is played. They haven't thought about what pieces can move in what directions. They haven't thought about how these pieces then interact with each other. They haven't developed a set of rules. In short, they haven't thought about the actual game itself.

--Neil West

[mos6507]

The only way to get this into an interactive game (that isn't just FMV shovelware) is to be able to 3D scan objects w/textures a lot better and cheaper than is currently possible.

In the future, you know that at least one advanced tool will come along that has thousands or hundreds of thousands of pre-scanned images included and you could adjust them to suit your needs. It would be possible to make a game without doing any scanning yourself. The software will be so advanced that you could also design a non-scanned object and have it look real and act real. The software will be able to reproduce any kind of gravity, clothing will move realistically, various types of skin (human or animal or invented alien or monster) will act according to believable rules. Breasts will jiggle, fat bellies will bounce, and skin will stretch.

 

If you want to look at what's going to happen in the near future for games all you have to do is look at the rendered 3D industry (i.e. CGI in movies).

 

CGI is always a few steps ahead of games because they don't have to render in real-time.

 

Most of the newest features in 3D rendering has to do with modeling aspects of nature algorithmically for things that would be otherwise impossible to manually animate via keyframing (think water, smoke, fire, soft-body dynamics, hair, etc...).

 

It will take a long time before all of these features start showing up in videogames.

 

You know, it does take a programmer to write the rule-set that gives a rendering engine the ability to create prototypical objects algorithmically instead of having to 3D scan stuff.

 

You can express a lot in nature mathematically. For instance, for a long time it's been possible to render forests and trees algorithmically using fractal and other math. Or various algorithmic textures like woodgrain and marble and reptile scales.

 

However, it's a lot harder to express compex objects as a mathematical formula. I mean, you can't just feed a DNA strand into a computer and tell it to simulate the creation of a human being by modeling every cell division and chemical interaction. Everything involves a certain amount of fudging.

 

I don't think we'll get even close to that level of graphics until we have quantum computers and holographic memory with net power that is as much of a leap as a 3GHz PIV is to an RCA Studio II.

 

It's purely in the realm of science fiction. Current CPUs are quickly reaching physical limits in advancement and at the same time there is a de-emphasis on speed with the general public who primarily use their computers as communications and media playing. For the majority of us, the current generation of CPUs are as fast as we will ever need and mobility is becoming more important.

 

So I don't see a mad rush on the part of CPU manufacturers to double or quadruple clockspeed due to all these factors for almost anything BUT game consoles.

 

I have to say I'm pretty skeptical about Sony's Cell processor endeavor. If it doesn't generate the kind of leap in graphics that Sony is hyping then I think it will bankrupt the company.

 

And Microsoft I don't think are being as ambitious as that with XBox2.

 

So who is going to get us from point-A to point-B?

 

I think what's going to happen 10-20 years down the road is we'll all wind up running out of oil anyway and we'll be happy to run notebook-style power-friendly chips in our desktops.

[ultravi0let]

i think all of the time and money spent on photorealism should be used for gameplay instead

[kisrael]

i think all of the time and money spent on photorealism should be used for gameplay instead

 

Yeah, but I think in general, you can throw more money at something and get better graphics, but I'm not sure if that works for gameplay...that requires good ideas for the most part, though cash helps in the tweaking stages.

[Artlover]

What if you want objects that have true volumetrics.  Like being able to cut an orange open at any arbitrary point and see the inside.  3D is not ready for that.

 

It probalbly is ready for that if you only worry about that orange and nothing else. But in a sick way, I'd kinda like to see that.

 

Virtual Orange 2004.

 

Be up to the wee hours of the morning.. "Ok, so lets see what happens if I have the orange angled at 34.6 degrees pointed at heading 162 and slice straight down the middle. Look at that one vein, it's off! What kinda crap is this!"

 

(I think Naturaly Artificial is going to expand outside the vaporware of fake food, and include a whole line of virtual food simulation software.)

[Raijin Z]

Meh, whatever is prerendered will eventually be rendered in realtime. Much of the mars landscape in Doom 3 reminds me of old prerendered backgrounds from games like Myst.

[sigma]

I can't be the only one to have seen eXistenZ...