• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Question for Developers: Is Normal Mapping the big thing for next-gen game dev?

Ark-AMN

Banned
Coming off my other thread, I wanted to get a broader opinion on if Normal Mapping is going to be the big "must do" in graphics for next-generation development. I ask this because now that we are pretty much in the next-gen already, and I'm seeking employment, should I go full steam ahead and research/learn Normal Mapping for modeling jobs in the industry?

Its sad that in my classes, we never got lectured on the subject even though it should have been anticipated considering when we graduated, it would become important. Then again, we still haven't installed Z-brush here at school yet. Morons.
 
element said:
normal mapping will/already has become one of the most used techniques.
Damn, they totally didn't talk about it at all. Bastards, rip me off, glad I checked on this, because I probably would have ended up with a portfolio with no Normal Mapping. Now I just have to research and learn myself I guess, hopefully I can get the techniques down by June.
 
it is already here, but for the most part it isn't any more involved than creating two models and then running a normal mapping program between the two to extrapolate the differences.
 
Ark-AMN said:
Damn, they totally didn't talk about it at all. Bastards, rip me off, glad I checked on this, because I probably would have ended up with a portfolio with no Normal Mapping. Now I just have to research and learn myself I guess, hopefully I can get the techniques down by June.

If you are talking about college, from my experience - universities lag behind the real world about 5-7 years usually. I'm sure that fluctuates, but it never seems like they are up on the latest and greated (at least in my experience).
 
In all honesty, I don't think any graphical element is going to be the next big thing for next-gen games. I think physics and AI (hopefully) will be the next big things. Unless of course you only meant this question to be applied to graphics alone.
 
Normal mapping is just a bump mapping technique. The technique has been around for a while. You can take bump map, which is essentially a height map, and then calculate the normals for each pixel. There you have a normal map.
 
RiZ III said:
Normal mapping is just a bump mapping technique. The technique has been around for a while. You can take bump map, which is essentially a height map, and then calculate the normals for each pixel. There you have a normal map.
not entirely. while the implementation of it is as such, actually creating the normal map is a little more difficult and definitely can be done well or poorly.
 
Normal mapping will be a common standard.
Stuff like parallax mapping and some special lightning techniques/effect will be the new buzz
 
Wyzdom said:
Normal mapping will be a common standard.
Stuff like parallax mapping and some special lightning techniques/effect will be the new buzz

I want fully realized ray-tracing like they do in the Pixar movies.
 
Question from a guy who really isnt a major tech-head/doesnt care to be:

Why does bump/normal/parallax mapping always result in the fucking awful plasticity/shiny effect?

Thanks in advance.
 
borghe said:
it is already here, but for the most part it isn't any more involved than creating two models and then running a normal mapping program between the two to extrapolate the differences.

Technically, yea. It has a lot of caveats though and particulars that you need to get a grasp on. Nothing horrendously difficult but the first few times it freaks out on you and you have no idea why, its frustrating.

There's also talent in known when and where to use it instead of just slapping it on everything at max intensity like some people have a tendency to do.

Parallax.

A parallex map is really just a normal map with a height map to aid in the occlusion. Parallax maps are good for environment pieces, tiles and floor boards and the like. The workflow from an artist standpoint is practically identical.
 
Normal mapping is no big deal. If you learned how to High res model than you should be fine. The actual process of normal mapping varies from software to software and engine to engine so it's no biggie. Just take a few of your low res models and add more detail.
 
borghe said:
not entirely. while the implementation of it is as such, actually creating the normal map is a little more difficult and definitely can be done well or poorly.


Well yea, but thats the gist of it. Point is that the technique is already used quite a bit because its pretty basic.
 
HomerSimpson-Man said:
Ah yes, ray-casting the elusive unicorn of realtime visuals, the really next big thing in realtime graphics.

I changed my text from casting to tracing. I can't remember which is the one I want. Sorry.

Someone fill me in.

It's casting isn't it. I was right the first time I think. Dammit!
 
Solo said:
Question from a guy who really isnt a major tech-head/doesnt care to be:

Why does bump/normal/parallax mapping always result in the fucking awful plasticity/shiny effect?

Thanks in advance.

Because of specular maps. This stuff is still in it's somewhat infancy in games and much more basic than what is used in movies, our lighting models are still no where near as complex.
 
HomerSimpson-Man said:
Because of specular maps. This stuff is still in it's somewhat infancy in games and much more basic than what is used in movies, our lighting models are still no where near as complex.

No doubt. Our lighting models are simplistic and huge generalizations/approximations of the real thing. It's really pathetic how bad our lighting is in games still (well, it's no ones fault though - we don't have the CPU power still). There are huge strides left to be made in the world of light in games.
 
Thanks.

I'll be some happy when the tech gets a bit more refined. The trend in games the past year or two of being "shinefests" is vomit-inducing. The effect looks nice on say tile floors, but when the player models are also like that, it becomes overused.
 
Solo said:
Question from a guy who really isnt a major tech-head/doesnt care to be:

Why does bump/normal/parallax mapping always result in the fucking awful plasticity/shiny effect?

Thanks in advance.
- Bad/non-existant art direction on the lighting pipeline of the game. i.e. Graphics programmers changing specular powers to show the technique more.

- Bad art direction in general (e.g. PDZ).

- Artists use the specular power more to show off detail coming from extra lighting results of normal maps.

- Unorthodox specular lighting equation (seen this in a few games).

- Artists/Art Director WANT to up the specular power/lighting intensity to show off normal maps to justify the time put into them.

- Not enough control for artists to change material attributes or a slow turn-around time to see results (i.e. bad pipeline). I'd be surprised if this would happen anymore.


If a game development team has a dedicated lighting expert, then most of these problems can be alleviated.
 
Dr_Cogent said:
I changed my text from casting to tracing. I can't remember which is the one I want. Sorry.

Someone fill me in.

It's casting isn't it. I was right the first time I think. Dammit!

Wait, you're right it's ray-tracing that's the extremely hard one, ray-casting is a simplified variation.
 
Zeenbor said:
- Bad/non-existant art direction on the lighting pipeline of the game. i.e. Graphics programmers changing specular powers to show the technique more.

- Bad art direction in general (e.g. PDZ).

- Artists use the specular power more to show off detail coming from extra lighting results of normal maps.

- Unorthodox specular lighting equation (seen this in a few games).

- Artists/Art Director WANT to up the specular power/lighting intensity to show off normal maps to justify the time put into them.

- Not enough control for artists to change material attributes or a slow turn-around time to see results (i.e. bad pipeline). I'd be surprised if this would happen anymore.


If a game development team has a dedicated lighting expert, then most of these problems can be alleviated.
+ maps being a lower res.
 
Well, it seems pretty obvious now that this is something I need to become proficient in. I suppose there are tutorials out there on the web. Hopefully I have the right software for it (I've heard that Z-brush is good for this, and MAX 8 has some things that can hep). I guess there are probably also plug-ins and other things that can be used in MAYA and such.
 
Is a common misconception that normals cause the shiny plastic looks. Its a result of shaders and spec maps.
Well maya 7 comes with a normal mapper, its pretty easy to use and you're able to view the results in real time.
 
Ark-AMN said:
Well, it seems pretty obvious now that this is something I need to become proficient in. I suppose there are tutorials out there on the web. Hopefully I have the right software for it (I've heard that Z-brush is good for this, and MAX 8 has some things that can hep). I guess there are probably also plug-ins and other things that can be used in MAYA and such.

Since you want to get into the industry, it might be beneficial to ask some questions here. I'm sure you'll get some answers from a few of the 3,000+ game developers on that forum.
 
Zeenbor said:
Since you want to get into the industry, it might be beneficial to ask some questions here. I'm sure you'll get some answers from a few of the 3,000+ game developers on that forum.
Wow, I never heard of that place before, thanks :)

Ah, thanks for the link Doc
 
Doc Holliday said:
Is a common misconception that normals cause the shiny plastic looks. Its a result of shaders and spec maps.
Well maya 7 comes with a normal mapper, its pretty easy to use and you're able to view the results in real time.

So true, some people even on this boards far too focused on the wrong aspects and trying to downplay normal mapping when it's one of the ways realtime graphics should be heading anyway.
 
HomerSimpson-Man said:
Wait, you're right it's ray-tracing that's the extremely hard one, ray-casting is a simplified variation.

Along with ray tracing and casting theres global illumination, radiosity, photon mapping, sub-surface scattering and other complex algorithms. As I understand it there really isnt one Holy Grail (ray tracing is just a part of it), but combinations of different algorithms to calculate lighting that are far too expensive to do in real-time right now. Although I think we'll see some decent approximations over the course of the console cycle. Hell we've already had Cell tech demos doing sub-surface scattering (Alfred Molina) and ray casting at 30+fps. Diminishing returns my hiney. :)
 
Tons of people use Zbrush to build their normal model

I'm sure you'll get some answers from a few of the 3,000+ game developers on that forum.
hardly anyone from the private forums reads the public forums at TCE.
 
hukasmokincaterpillar said:
Along with ray tracing and casting theres global illumination, radiosity, photon mapping, sub-surface scattering and other complex algorithms. As I understand it there really isnt one Holy Grail (ray tracing is just a part of it), but combinations of different algorithms to calculate lighting that are far too expensive to do in real-time right now. Although I think we'll see some decent approximations over the course of the console cycle. Hell we've already had Cell tech demos doing sub-surface scattering (Alfred Molina) and ray casting at 30+fps. Diminishing returns my hiney. :)

There is some a nice little approximated solution called Pre-computed Radiance Transfer/Spherical Harmonics that can simulate ray-traced global illumination in real-time. Sub-surface scattering (translucency), colored interreflection (i.e color bleeding), ambient occlusion, self-shadowing, etc. There is even a method to do this for deformable meshes called Local Deformable PRT.
 
element said:
Tons of people use Zbrush to build their normal model

hardly anyone from the private forums reads the public forums at TCE.

Well, even though most us browse the private forums, I do seem to remember some posts being answered on the public programmers forum a few times.
 
HomerSimpson-Man said:
Ah yes, ray-casting the elusive unicorn of realtime visuals, the really next big thing in realtime graphics.

I think that's a bit far off. The next big thing should be volumetric smoke and water
 
Apenheul said:
Well raytracing is basicly just recursive raycasting.

Actually it isn't.

Ray casting sources from the viewer. Ray tracing sources from the light source to the viewer. Ray casting is an approximation used for performance reasons. It's not really the same thing at all.
 
pj325is said:
I think that's a bit far off. The next big thing should be volumetric smoke and water
And how about finally making animations and direct control work together without any strafing, jittering or anatomically incorrect looking shit. :) I swear, play the Fight Night 3 demo, the shit looks real, until you start playing...

Really, the ray tracing is probably so far off that it's a few generations at the least until that will rear it's pretty head.
 
The biggest thing with the new generation, in my opinion, is definitely the following:
-motion blur on all elments
-normal mapping
-High Definition
-ultra detailed models/textures


If i had to pick one, it would be motion blur.
 
For me....

Animations, if the animation is great i can ignore pretty much anything..and i dont mean motion capture. This is a talent thing mostly but still. I think more devs should be focusing in this more than anything else right now.

Normal maps, once the artists get used to this feature ie GOW, MGS4 things are going to look amazing.

Lighting, raytracing is overated..it gives you sharp shadows and its cpy intensive. Shit pixard didnt use raytracing till very recently. Faking shadows is fine with me. What needs to improve is the materials with translucent properties work in light examples: Skin, glass, clothing, liquids etc etc. But again this is all icing.. improve gamplay and animations first!
 
Top Bottom