More Next-Gen In-Game Goodies From EA

Show some ingame. Cutscenes are so useless.

What's to say that isn't ingame? They could use the game engine to do cutscenes, you know.

Plus, cutscenes usually are of a higher quality, thus showing how far they can stretch graphics.

im not sure what you mean... if you are sayign its pre-rendered then i have to disagree... it is obviously in engine... there are a few noticable inications of this, the left forearmcloser to the boddy you can see some flat edges, so you can tell its a lower poly model than a pre rendered would be. If you are saying that its an in engine cutscene, i believe that is true, but i dotn see why that is a problem. i realize some one who is big into the actual football games woudl want to see a shot of the players on the line and see some of the interface, but these are early teasers, they want the most exciting images possible.
_______________

He looks like he's been startled just before taking his buddy up the ass.

I would classify in-game as just that; when you're actually playing it. Current-gen (and even some last-gen) cut-scenes or replays useing an engine, rather than FMV, can look better than "in-game", as it's not running "on the fly" (e.g the replays in GT on PS looked way better than in-game)

I agree, it's obviously an engine running it; and the in-game graphics will probably look as good. But the irony with most sports games, is that the camera is so far zoomed out, in order to give a wide enough view of the playing field/pitch/court, all that detail is redundant. Maybe in-game they'll concentrate on getting the crowds and sun-lit arenas looking perfect?

EDIT: Argh! I've been Kelkoo'd! SPOnG, I expect my cheque in the post ;)

i dont think the fact that is is not "on the fly" is the reason the after race cutscenes look better. Often times there are special effects added to help add realism to the cut scenes like focus effects(in GT3-4) Just because the system has rendered something once already doesnt help it render it the second tiem any faster, it still has to do all the physics calculations, basically the only info being stored to create the replay is the input rom the controller, anything more (actual frames or animation keyframes) woudl probably cause to much of a load. I believe alot of these effects are not used during the actual game play because it might hinder the play itself, the constant change in focus during a race in GT would probably cause you more trouble than its worth.

Now i know some games do use more detaie4ld models for (Actual) cut scenes, I.E. story progressions, but i dotn believe that is the case in something like madden. Its not liek its shenmue where there are prescipted animations and you can decide when the more comlx models should and shouldnt be used, these are just close up shots fo something that just happened... liek matrix style replays of big hits and such.



to true... no matter how good a football game looks, its not gonna look that impressive from 100 feet over the players, the more dynamic lighting and some fur style effects on the grass would help to add some realism to even the distance shots, that and some added tiem spent on idle animation of th eplayers ont he line, makign them more distinct when all lined up in a row (tapping there foot before the snap, turning their heads side to side checkign the line... keeping their eyes on the reciever as he moves about before he lines up).
__________________

Ok I know what the image is, its using the game engine to show a closeup. If the action is scripted and predictable of course they can use more resources to make it look better.

Nobody actually looks at this stuff every play, or the game would take 2 hours. So while it looks fantastic, what does the actual gameplay shots look like. They will obviously be worse than this.

I reall ythink the image is indicative of what the game will look like... the visual fidelity will be the same... the onyl difference will be that during the game you will be farther away so it will be a less exciting image and there will of course be little colored circles on your current cotroled player and images of corespoinding buttons for recievers and it will seriously detract from the realism, plus a camera placed perpendicularly to the line of sckirmage will never look like the camera angles used on tv, it just doesnt make sense for control sake. But i really dont think they are varyign detail at a distance, especially considering these games will be running in high def, the weird pop in on distant opjects in halo2 is bad enough, the sudden swing in when a reciever catches a ball doesnt allow any time for a change of resources and it would also tie up ram to have multiple version of every model. We are talking about a system that is theoretically 3 times as fast as a current top of the line mac with a s**t ton of fast ram (when you consider there is no OS) and a 2 generations from now graphics card, its really not that rediculous to think that they could render some decent face detail on 24 (22?) players at once. has anyone played halflife? there are a pretty good numebr of charcters running around at once in there (and very complex environments and complex physics)and top of the line pcs arent nearly as fast as the 360 will be when it is released.
_________

This image just proves to me that in the foreseeable future we will never have games that can replicate reality.

I want a picture where the ball is in a move, with the normal camera 30 meters away from the football players, and if the game has new ways to lead the team.

Realistic graphics is more helpfull in games like Shootings, RPG, action games.
Even in thes games It is hard to tell if next generation improoved graphic can add new elements for the gameplay comparing with this actual generation consoles.

Some times too realistic games look artificial and like a mock of the reality, thats why it is important a good artistic direction.

Since I haven't actually developed 3D games with replays, I'm not sure I'm in a strong position to argue.

However...

Surely the game must record the geometory of the various car models, in relation to the track (which is of course static), throughout the course of the race. As the game is obviously raced in real-time, and is decided by the user in real-time, the engine must refresh the rendered images in an instant. Therefore, the level of complexity in creating shadows & lighting etc must be limited to keep the number crunching down. Since the geometic positioning of each car model is known prior to the replay starting, surely the processer can process what's about to be diplayed before the replay actually starts?

i havent developed any games either but i have done a pretty good amount of 3d work and read and study it a good deal



the difference being that the replays are viewed form a different angle so it still si being rendered in real time, each frame has to be complelty drawn over, in GT for example you can instantly switch camera angles, the only thing you mention that i think might help at all is knwoing where the shadow of a car may fall ont he track or where the shadow of a tree mgith fall on the car, however in order for this to be worth while it woudl have to basically create a lightmap for each individual frame and make the texture on the car animated to match its position in relation to its suroudning objects, which woudl require the sytem to save the shadow info in a movie file of sorts. Its been a whiel since i have doen any real time work but when i dabled in Quake3 level designing i know it took an awfull long time for the game to produce the light maps... several minutes... and those where static. also the fact that the backgroudn is static doesnt really matter as nothign is techically static in the sense that it isnt being redrawn. an actual static backgroudn would be like those in resident evil (not4) weatehr or not somethign is moving doesnt matter its all relative, it seems as though its sitting still and you and your car are movign through it but that makes no diffe to the processor it still has to redraw it from a diferent angle for every frame as it is being scene froma different angle.

i think im over analyzing this, but i like to cover all the bases... if im wrong abotu any of this please feel free to correct me as i am certainyl not an expert.
_____________

LUPOS, you told me that you are an expert in software inventory, now i realize you are an expert in graphic processing too. It looks like you are jack all the trades.
What else you know about?

Yeah, that was my bad in using the word render. Obviously every 3D image displayed, that has been calculated by host CPU, is constantly refreshed.

My main point is that before a replay is shown, the gemoetric position of a car, relative to the track (and other car models) is already known. Whereas during the game, the geometric position of a car is decided "on the fly" by the user (and your AI controlling your competitors), which requires more CPU resource.

When it comes to the replay, the resource required for calculating geometric positioning is now freed; thus giving the CPU the resource to render more fancy effects.

Er, I think.

jack of many trades master of none :/

i went to college for computer graphics for 3 years, didnt have the scratch to finish however. I do web stuff, flash animation, 3d modeling, basic programing, if it involves a mouse and keyboard im all over it, which is probaboly why im not especially good at any of it... my mind wanders across to many differnt things.. i always thought game producer woudl be a good career. since it involves knowing a bit about all the different aspects... yet here i am... inventorying eggs :/

i have been told that starting out as a tester is a good way to get your foot in the door and prove your know how... unfortunatly i cant afford to live in the NYC area on a testers pay.
ces la vi!
________-