This Forum is dying ... Congrats great Pixologic !

[justadeletedguy]

Normal - is the direction a polygon, edge or vertex is facing, the blue here lines here are the normals.

UV ... Unrwrapped Vertex (I think). A model is Unwrapped like the seams in clothing so an image can be projected onto the surface. The image projected for color is considered the Diffuse Texture. A Normal Map can also be used to show detailed geometry thats not in the actual Model. The normal map affects the lighting channel of a model, casting highlights and shadows according to the Image rather than the model Geometry. There are several other kinds of Image textures as well.

Baking is the process of putting the high-poly details and color onto the Textures of a Low-poly model. The Normal map is the best example here. A Zbrush model with millions of polygons has details like wrinkle skin, cracks, noise. The detail can then be projected onto a texture image (Normal map) so these details can be visible on the low-poly (lowest subdivision or retopologized model).

The whole idea here is to be able to use your Zbrush models in Games and Animation. Notice how long your Zbrush or Sculptris render takes! For a game you need 20+ Frames per Second. Your computer is figuring for Lighting, camera animaton, character animation, enemy characters animation, plus it communicating location across the internet and keeping score, health, time, etc .. So We project the details of a High Poly model onto a Low poly model that the game engine can handle. To get a better Idea, take a look at the Dominance War V challenge winners. Click on a few of them and scroll down. They show the Textures as well as the wire frame. And the limit for the challenge was 12,500 tris!! thats 6,000 quad polygons and about 6,000 vertices. The Pink and blue image is the Normal map. Some also have Specular and Opacity maps as well.

Sculptris exports the same format as Zbrush (.obj). Any software that can use Zbrush models can use Sculptris models. To see your paint though, you have to save your texture map! Click "Show advance tools" in Paint-mode, and save Texmap and save normals, then export your .obj. Blender, Wings3d, 3dsmax, and most others support the Microsoft Wavefront object (.obj) format. However, it's an old format, to view your textures and Normal map, you may have to assign them to your material manually in your other software. Sculptris Materials can be mimicked somewhat in Blender as well.

[SeanJM]

justadeletedguy: Normal Map, think of how the models looks so good. A little math okay? 1024x1024 = 1,048 576. What does this mean? That means that on a map that is 1024 x 1024 you can have 1,048 576 (if you map was packed perfectly of course) polygon details mapped to your low poly mesh. So, on most systems Sculptris can't hit these numbers. And 1, 048 576 is a bare minimum detail because you can pack pixels, a lot of models for games come from higher resolution meshes that are in 30 million+ mark.

This is a very simple reason that we want better performance.

[justadeletedguy]

a 4096x4096 DDS is a 21+ mb file ...
a 2048x2048 DDS is a 12 mb file ....
a 1024x1024 DDS is a 3 mb File ....


Little Johnny is the End User on a laptop with a 256 mb shared memory graphics card. If your model has Diffuse, Normal, and Spec maps, How fun is your game gonna be with like 2 models and a skybox ? ....

Most people don't buy hot rodded, alien-ware computers. They get them from Wall Mart, they use them right out of the box and don't know what a graphics card is ... if his card is forcing mip-maps, he may not even see 1024x1024.

Sure Playstation and Xbox will handle it, but with the new Pads and Smart-phones ... I think you'll see less demand for Console games and more demand for the portable stuff ... and there's size on disc, time to download to consider.

As an artist, I certainly understand your concern on wanting to save your hard work. But whats the point if no-one else is going to see it !!

[Serek]

@justadeletedguy

UV stands for just UV because XYZ was already taken (not a joke)

[MealeaYing]

@JustaDeletedGuy:
Thank you very much, I just learned a LOT and that was EXACTLY the sort of answer I was hoping for. I have looked these things up and got nothing but a bunch of (to me) gibberish and long math equations.

Thanks!!!

[michalis]

@ mealea.
if you hate retopology so much, then I may feel your pain but there isn't any point of talking about baking, texturing, exporting to other apps, or similar. Retopology and UV mapping is needed to have real texture maps. When you just paint in zbrush, this isn't a texture. This is called vertex painting and it can't be exported.
I like to help you on all of what you asked, but, mealea, you asked too much questions, what a barrage. LOL
I already, among other friends here, posted some tutorials like the quasi retopo (start from this one because its easy). Its a prelude on what we call retopo, UVs etc. If you don't have animation in mind, it maybe enough for you, I can't tell.
What Zbrush can do for exporting, is to "bake" poly paint to textures. Or displacements or other useful maps. So, this is called baking. On the contrary, sculptris can auto UV somehow but doesn't bakes anything. Here you just paint directly on textures in a 3d environment.

OK, about normals, as justanotherguy said. In blender, select the mesh, edit mode, select all [A]. recalculate under [T] palette (the left one), better [ctrl+N].
If you involved with blender, better start from this tutorial.
http://gryllus.net/Blender/3D.html
Learn the basics and everything will be much easier and faster after. Spend ~ two days on this, not more.
Then, ask again.
We're the good guys, remember.

[justadeletedguy]

Shut up Serek ... It sounded good .. lol

[michalis]

And mealea did it fine, as you can see here.
http://www.zbrushcentral.com/showthr...l=1#post918525
As zb retopo may not be great but I can thing otherwise. LOL

[SeanJM]

justadeletedguy: fortunately, this is a big world and you can do what you want yeah? Good news is, if Sculptris can support the polycounts I'm looking for, it can support the polycounts you're looking for. We both win.

[MentalFrog]

Im going to babble a bit here ok?


So.... Im totaly lost and I suppose I should just ask since Im starting to look at Blender for some things I have been thinking about....

What are normals and what are they for, why do they "flip" and why is the description in Wikipedia useless to me?
What are UV's, Uv Maps, and why do they seem to destroy perfectly nice texture work in ZBrush? Do they suck as much as they seem to suck?
What is baking and why does everyone want to be able to do it to everything?
Can anything other then Sculptris see the painting you do in Sculptris, and if so what and how?

I would love this explained as I suspect these are things that will help me get stuff I make in Sculptris and Zbrush into other programs with all the textures I make for them, right?

These things are really not that difficult. The math, physics and science behind it can be rather complicated but the 3d artist doesn't need to know the details just the basic idea behind these concepts and why they're used.

You need to realize that digital 3d is not really 3d. In real life everything is made up of physical molecules. 3D graphics only emulates real life based on mathematical formulas, many from actual scientific laws, formulas, and ideas.
3d models are actually hollow and can pass through itself and other models, unlike in real life where an object's molecules would collide with another. So things like collision detection are created in 3d. Same thing with particles, lighting & shadows, fluid, dynamics, etc.. Everything is emulated and calculated for proper render output.

As computers and technology improve 3d graphics becomes more realistic. This is because of two things; One being the more complex math formulas can be processed quicker on faster hardware, The second being that the formulas and mathematical laws that 'interpret' real life can be improved, simplified, and made more efficient. Faster computers and more efficient formulas mean more information can processed.


Normals:
Normals are simply telling the computer which direction a face is facing. This is mainly important for lighting. This is because 3d lighting uses the idea of Lambert's cosine law: http://en.wikipedia.org/wiki/Lambert's_cosine_law

The computer uses this formula (or similar ones) to help calculate how it should bounce the light sources off the model and back to the rendering camera.

For an experiment try creating a basic model like a sphere. Then select an area of faces and flip their normals. Render it out and you should get some strange shadows where the edges of the regular faces and the flipped normal faces are.

Sometimes you'll get strange texture issues and artifacts. By understanding that normals affect your lights and shadows it can help troubleshooting issues that are really a matter of fixing normals rather than the actual material or texture.


UVs:
UV is just a cordinate system. A 3d model is made up of 3 cordinates x, y, and z. A map uses a 2 dimensional cordinate system. Like was mentioned the UV is simpy because the 3d cordinates already use x and y.

It's like if you had a 2 dimensional map of the world on a piece of paper and you also have a physcial globe of the world. Both show the same locations using the same cordinates of latitude and longitude. However because one is 2 dimensional and one is 3d they show them differently. Here's where the UV system comes in. A UV is simply telling the computer where a 2 dimensional cordinate is on a 3d object. You can edit UV cordinates in both 3d space and in 2d space, depending on your software tools. You can change UVs in 2d space without changing its 3d space and vice-versa. For example if you applied the 2d map of the world to the globe but turned the map upside down so the north and south pole are switched. This would be the same as changing the UVs in 2d space but they still the same in 3d. When you rendered out the globe it would be upside down. You could get the same affect by turning the UVs upside down in 3d space but leaving the 2d UVs the same.

Mapping is just when you (or the computer) have mapped out the cordinates between the 2d map and the 3d cordinates on the object.

Where does mapping come in? Mostly in texturing. In real life the Mona Lisa is a canvas painted with various materials and colors. In 3d you could model the Mona Lisa using one face for each particle of the actual painting. You would then assign a different material or color for each face. This means you'd be recreating the Mona Lisa. To emulate this in 3d is much simpler. You take a 2d image and apply it to your 3d object. When rendered the computer will match the pixels of the 2d image and apply it to the 3d object based on the UV cordinates.

So when someone is talking about a texture map, they're talking about the color map (like the paper with the map of the world on it) and how it applies to the 3d object (like the globe). You can map just about anything, diffuse, specular, reflection, etc.. You can even model using mapping. This is where things like bump mapping and normal mapping come in. A bump is simply the computer roughly estimating a bump or crease in the model using 2 dimensional lighting tricks (it's similar to when an artist paints or draws something like a still life, they create the shadows and lighting to look 3d but if you look at the painting at a greater angle it starts to look like 2d like the canvas or paper it was done on). Bumps are very quickly processed by the computer but when the object's faces reach the greater angles it loses its effect. To see this try applying a bump map to a sphere. You'll notice the effect of the bump will get lost around the edges of the sphere and it will look perfectly round even though it should look 'bumpy'. This is where displacement maps come in as they actually change the 3d model at rendering time. They also increase the process time compared to bump maps. Normal maps is just a way to tell the computer to modify the angle of the face's normal during rendering. This is more efficient than displacement but much more realistic than bump mapping.

http://upload.wikimedia.org/wikipedi...ap_example.png


Baking:

"Baking is basically the process whereby you "bake" all the textures applied to the surface into an image that you can then apply as a texture. This is useful because if you have a whole lot of procedural textures and image maps applied to a single surface channel (like color), it takes longer to render. But baking them all into a single image which you can then use then solves the problem. "

For things like gaming this is really handy. You can actually render out a scene and bake your lights and materials to the scene. When the baked texture is applied to the scene it's much quicker to render because you don't have to recalculate everything.

You can even bake things for modelling. It's like taking a high polygon model of a face with wrinkles and baking it into normal and displacement maps. You can then take a much lower polygon model and apply the maps to it getting a very similar rendering as the higher polygon model without the high cost of calculating time. Just like the example above with the normal map. This is how you can get things like high detailed Zbrush models into reasonable models to use for real-time renderings and lower end hardware such as gaming consoles without losing the important details.

And I should point this out too:
The only thing that indicates to me that Sculptris is in any sort of trouble (like not being continued) is that when you are working on a dynamesh in Zbrush and hit GoZ it crashes Sculptris every time, that is it, one thing. And I bet that gets fixed at the same time we get fuzz and fluff and fur and all that other wonderful stuff that is about to happen.

This may be due to the fact that they handle polygons at different counts. Zbrush models need to be decimated to a reasonable count for most 3d software. Sculptris will crash with high poly counts, poly counts that Zbrush has no problem hitting.

Something to note here is that Zbrush and Sculptris do things totally differently from each other and as an example I have to point out one thing:
Sculptris uses my GPU, ZBrush does not.
Is that not a fundamental difference going WAY past tri's and quads's?
Merging the two, or "simply" expanding Sculptris and not breaking compatibility with some current or future plan for ZBrush could be very very very hard.

The only difference between quads and tris are the algorithms used in the software and hardware. Which is better? At the moment probably quads, this is mostly because quads are calculated into triangles anyways for the rendering process. However some algorithms and formulas don't handle triangles as well as quads. Formulas that are efficient to apply to tris can usually be applied to quads, but that isn't always the case. This can easily change in the future if tris become more efficient in the overall process. Also the difference depends on your pipeline process and what you want from your end result to be. Still renderings don't matter. Animation (games, movies, etc..) is where the bigger differences between the two come in.

One last thing, I asked a few programers about making a 32 bit program into a 64 bit program and they said "it depends" I asked "on what" and they said "well, what is it that you want converted and why?" I told them "ZBrush" the results were unanimous: they all laughed and said variations on this: "Some programs you can just hit a few switches in your compiler, something like ZBrush its possible but more likely to take a few years."

It's also not that simple to just compile. You could compile a program to run on 64 bit but that doesn't mean it will use the 64 bit to its potential. This is actually the case with most software currently, even 3d software. It's not written to take advantage of a lot the 64 bit has to offer. It will in time and that means faster and better 3d programs in the future.

One other last thing... I hate retopo.

It sucks but currently it's one of the steps required to make animation more efficient. The day will come when it becomes a lot more automated. I'm not sure it will (or should) ever be fully automated. I think it will always require some user input, but that should be very minimal. I also don't think it's that far off, just needs someone to come along and make it happen. The tools are definitely improving.


Hope that helps.

[RedSquare]

@ MentalFrog. Ye Gods, this thread is like reading a DAZ3D Hexagon forum post; doom, gloom and still more doom. But points taken and ignored with contempt. Onwards and upwards.

[MealeaYing]

@Mental Frog:
Thanks for another excellent description, that was great too!

@Red Square, I have never been on those forums, but I dont see this as doom and gloom, just a wee bit of worry about the future of something very cool namely Sculptris.
Personally I am betting that they are still making Sculptris better and that we will see it when its ready, and that is going to be VERY FUN!

I said before "I HATE retopo" and I have to revise that somewhat, I didn't know that turning off transparency would make it a thousand times easier, it does..... honestly I didn't even know I had left it on, very embarrassing, but now my only problem with it is what I do next...
I think in my case its more like "Onwards and outwards" or: "To Oblivion and beyond!" or at least Blender.....

[michalis]

Mealea, I feel a bit responsible for this.
All these are great. of course, to make apps communicate. To have great exports and probably some nice renders, animations etc.
These won't make you or me a better 3d sculptor though. We're loosing our time and (talking about me) it's not the first time. Technical details around. I wonder
http://www.ownfineart.com/Drawings/S...nd_Falling.jpg