Gears of War: Grand animation in a little package??????

[dcotto21]

i been playing the game gears of war and see that the detail on the characters are grand. so grand that i think that my cpu would just crash if i only created one of the drone characters…im wondering what the polygon count would be on just 1 of those video game characters. i use zbrush and when i sculp just a face i cant have it look nice and smooth unless its about 100,000 polygons… i have a 2.4ghz and at that rate my cpu starts to slow down a bit… i export my sculpture onto maya and the wireframwork is to the point where i cant even see the spaces inbetween the polygons… im wondering…how do they get those geogerous characters into a small dvd9 8gb disc with so much mass of levels and content were they shoot each other huge levels, etc…i seen xbox 360 games where u could download a whole level with beautiful detail which only about 13 mb in size!!!. im wondering if its just an illusion trick done with textures, were the detail is actually done in the textures with the character being at about 5000 polygons a chracter or if it has something to do with the hardware of teh 360…i see some of these videos of artists workign with 3d characters and the characters look great and then i see that the polygons arent as much as i thought they would be cuz they show the framework… i was wondering what would be the polycount of just one character in a game detailed like this and what would be a method they use to get such detail on the faces like that character below which looks like it could be a million polygons and looks almost the same way in the game…do they actually boost it up the polycount and then eliminate some of the polycounts? how would u be able to still have the same look of tha character but then just lower the polygon count without altering the appearence? is that possible? how extactly do some of these top of the line video game companies actually produce something that looks like it would take terabytes to make (which probably does) and then have it on a small 8gb disc? if anyone knows a link give me a link to were they talk about stuff like this more in depth or something similar please let me know or if u have some knowledge then please share thanks

Attachments

[Luminous_Anubis]

uuh well you’ve come to the right place. The process is called normal mapping. Begin your search.

[dcotto21]

thank you! with ur information i was able to find detailed information on wikipedia on what normal mapping is and found a pretty good tutorial on normal mapping with 3ds max here>http://zbrushcentral.com/zbc/showthread.php?t=32119&highlight=normal+mapping+tutorial
the video i seen shows extactly something similar to what i’ve seen while playing video games… usually on gears of war online multiplayer matches and on halo 2’s campaign and online m.p. as well. when i see the game start from a new level the detailing of the skin slowly begins to build up right before your eyes. what im guessing is that that process that i see is what they might call “real time rendering” correct me if im wrong …so what im thinking is that they use a normal polygon with low resolution and convert them into a nurb or a nurm (such as in the example of that tutorial with the elephant.wmv) so as to smooth out the edges and then apply the texture with normal mapping. and thus creating a simple small in size… i guess since the xbox360 is not a pc it does not have direct 9.x real time renderer but instead uses one of its own. im thinking that each game uses its own real time render and that the unreal engine 3 has a patent real time renderer which creates a grand scale of detail…is unreal engine more than a great real time renderer game engine or it is something more than just real time render/game engine? what im actually wondering is what does unreal engine actually does in order to create great games. but thanks for ur answer i was not expecting an educationed answer and i found one which answers my question fully. normal mapping would explain it… normal mapping uses the red green and blue dots of a selected polygons to shade the selected area with the image…but now i have one question…how would u be able to do this process is zbrush? when u normal map do u have to take a high resolution 3d model, sculpt it and mold it. then copy it as a texture perhaps or .jpeg .bmp. then optimize the 3d model into a lower resolution and apply the jpeg,bmp or texture? (is that how the process of creating normal mapping 3d model occurs?) i usually see these purple, pink and blue images of what looks like the skin of a character which i think they call UV mapping i suppose…normal mapping is almost or somehow is the same thing i suppose where the skin is ripped off and then properly adjusted onto a lower resolution model. i dont think u could do certain functions like real time rendering on zbrush but would u be able to convert the resolution and make it a smooth nurb…usually i have a model that is so pixolated (squarey on the face for example) that i cannot get great details on the nose and mouth cuz they been pulled out so much without having to divide it to the point of nearly 500,000 polygons which drives my cpu to become slow. i wanna make my high resolution 3d characters look as great as a million polygon model but yet not have such an impact on my processor speed in order to apply more models into a 3d world will normal mapping help me with this and could it also be applied in normal movie animation? i know its alot of stuff im talking here but if each question could be answered it would ease my mind a bit from this immense curosity im getting from all this specutacular effects im seeing now and days in these games and images. thanks

p.s. normal mapping must be the norm now for every game i suppose…games no longer just create polygons and paint them…they actually use normal mapping for everything correct? please if anybody has more tutorials on how to use this technology let me know thank u

[womball]

The new Zelda game doesn’t have normal mapping. A normal map is just additional detail. The main structure still have to be in the base mesh. If you used correctly they create the appearance of tons of detail. Otherwise it just doesn’t look right.

[dcotto21]

yes thats true i guess certain parameters need to apply in order to perform the correct appearence one wishes to obtain but in the case of gears of war they either used a real good amount of polys in their mesh. its hard to find any “blockiness” or improper details on the characters (except in the waterfalls, they are blocky) in my opinion from what i’ve seen. i suppose great effects could be done without normal mapping but in that “REEL” video…it doesnt look as though the artist MIGHT NOT (im only guessing though) didnt use normal mapping but rather if its an example of good painting of the texture rather than normal mapping a high resolution model WITH good texturing. if u seen the video though it still looks like the characters uses low polygons (an example of the character is the fat monster in the bottom of the picture) the edges are nicely curved rather than straight…im wondering how did they get lines to curve nicely with the appearence of a high polygon count in this character…was it subdivided to nurms or a nurb spherecial geometry or some sort…cuz like i said before my characters gotta be about 100,000 in order to get that nice smooth shapes if not it would be all blocky …i will show u an example of an image i created where the lips of the mouth are blocky as soon as i get to my house computer ill will upload my high framewire image…but for now look at this monster…the polygons are spreadout nicely they are not compressed full of lines like mine is this an example of normal mapping …coloring texturing or what i know there could be many possible methods to get this and i would like to know more just like i learned about normal mapping

[dcotto21]

could the wireframe on that image actually just be a projected image? lol