BlenderArt Magazine Issue 12 Texturing Special

Issue 12 | Sep 2007 | Blenderart Magazine

Blender learning made easy

- Cristian Mihaescu - ‘Eggland’

Gaurav Nawani [email protected]

Sandra [email protected]

Nam [email protected]

Phillip A. RyalsKevin C. BraunDerek Marsh

Abhihsek GuptaAdam FriesenAlex (blenditall)Andi SchumannClaas KuhnenEric PranauskKrzysztol ZwolinskiOlivier Saraja

‘BlenderArt Magazine’, ‘blenderart’ andBlenderArt logo are copyright of GauravNawani. ‘Izzy’ and ‘Izzy logo’ are copy-right Sandra Gilbert. All products andcompany names featured in the publica-tion are trademark or registered trademarks of their respective owners

Cristian Mihaescu - ‘Eggland’

www.blenderart.org Issue 12 | Sep 2007 - Texturing

mailto:[email protected]



http://www.blenderart.org

You know that you have finally become areal “blenderhead”, when not only do youlook at objects around you and breakthem down into their various vertices,edges and faces, but when you becomeawestruck by the infinite variety of tex-tures in the world around you. You will bewalking along, minding your own busi-ness when it will grab you, the way rustis peeling on a signpost, the way a flowerpetal looks with morning dew on it. Youwill notice how the light affects the tex-ture and next thing you know, you havegone from awestruck to breaking it downand figuring out how to recreate it inblender.

Welcome to the wonderful world of Tex-turing. In this issue we will be taking alook at how to create great textures toenhance your images and animations.Creating great textures is a skill like anyother and one that can be learnedthrough practice and experimentation.One of the first skills you will need tolearn is how to observe everyday tex-tures. What makes them unique, whatmakes them “work”. Then with thatknowledge and a good understanding ofBlender's texture tools, you are on yourway.

In this issue, not only are we coveringhow Blender's texture tools work, but howto effectively use them to create trulygreat textures.

We have gathered up some great tutorialsthat will show you how to remove UVSeams and create Normal Maps. Then youcan try testing your new skills by follow-ing along to create a great Skin Shaderand realistic plastic shaders. Which oddlyenough, is not as easy as one wouldthink and yet not as scary or difficult asthe last statement would imply. In'Keeping the Flame', we see how to set upand create a realistic flame. Then we gobehind the scenes with a look at how'Rooftops' was created, while in 'Tangramsof Light' we learn how lights affect tex-tures and how the whole can be used tocreate just the right mood.

So go grab a cup of coffee and curl up forsome great, informative reading, sure tochange the way you approach your nexttexturing adventure.

Happy [email protected]

Managing Editor


One of the first skillsyou will need to learn

is how to observe!


At some point in your blender exploration, it willhappen. The perfect material/texture for yourcurrent masterpiece is one that you have al-ready created in a previous project or one thathas been created by a generous communitymember that has offered to share his/her mate-rials with the community. So how do you goabout getting that perfect material/texture intoyour current project? Well it couldn't be easier.

Blender has two options for reusing materials,Linking and Appending (actually both these op-tions can be used for all of blender's assets.Blender assets can include actions, armatures,cameras, images, IPOs, lamps, materials, mesh-es, objects, scenes, text, textures, and world,etc.).

will place a independent copy of thematerial into your new project.

to a material will link to the original filethat contains the material, meaning that anychanges to the original file will also be saved to

the new file that you linked it to. This is a usefuloption when working on a large project that willgenerate a large number of files that may beworked on by one or more people. This will al-low any changes made during production to bepropagated throughout all needed files withoutthe added work of updating each file individual-ly.

Okay, on to actually seeing how to use theseoptions. Both options are started the same way:

1 From your new/current project file, either bygoing to File> Append or Link or using thehotkeys Shift + F1, open the file browser win-dow. (All blend files can be used for Append-ing or Linking.)

2 Browse to where you have saved the blendfile that contains your perfect material.

3 Click on the file name of the desired blend filewhich will open the library list ofAppendable/Linkable assets.

4 In this case we are looking for materials, soclick on Material.

5 That will open a list of all available materialsin that blend file.

6 At the bottom of the file browser there are twobuttons, one for Append and one for Link.Push whichever option you have chosen foryour project.

7 Right Mouse Click on the Material you want toLink or Append, to highlight it and then MiddleMouse click to confirm (load) your material toyour new project file.

8 Now your material will be available to as-sign to whatever object/model you wantedit for.

There, just as I promised, easy as can be. Withpractice, you won't even have to think about it,you will just click your way to reusing yourblender assets.

Bump MappingWhat is a Bump Map? Bump maps are texturesthat store an intensity, the relative height ofpixels from the viewpoint of the camera. Thepixels seem to be moved by the required dis-tance in the direction of the face normals. Youmay either use greyscale pictures or the inten-sity values of a RGB-Texture (including images)(definition taken from the blenderwiki).

Well that is a nice definition and all, but whatdoes it mean to the average artist? It means wecan fake details and geometry that would betime consuming to create or result in a modelthat due to high vertex/face counts would takeentirely too long to render (that of course is forthose of us that are impatient with long rendertimes.)

Let's hear it for faking! Now let's look at how touse Bump maps.

1 Add a new Material in the Material Buttons,assign color of choice

2 Click on the Texture button (F6)3 Click in a Blank texture slot4 Choose an texture type (there are many to

choose from, some procedural textures don'twork as well as others, for more informationon the different texture types and their usesyou can refer to the blenderwiki:

5 Apply the Texture in the Material buttons MapTo panel by clicking the Nor option. Thestrength of the effect is controlled with theNumButton Nor on the same panel.

Texturing tools & Techniques


A look at various tools andTexturing techniques toenhance your images.

http://wiki.blender.org/index.php/Manual/PartIV/Procedural_Textures

Multiple Bumps maps can be used together tocreate any number of detailing effects. UV maplayouts can be used in combination with proce-dural textures to create more complex and real-istic effects.

Render BakingRender baking creates 2d bitmap images of amesh object's rendered surface. These imagescan be re-mapped onto the object using theobject's UV coordinates. Baking is done for eachindividual mesh, and can only be done if thatmesh has been unwrapped. While it takes timeto set up and perform, it saves render time. Ifyou are rendering a long animation, the timespent baking can be much less than time spentrendering out each frame of a long animation.

Use Render Bake in intensive light/shadow solu-tions, such as AO or soft shadows from arealights. If you bake AO for the main objects, youwill not have to enable it for the full render, sav-ing render time.

Use Full Render or Textures to create an imagetexture; baked procedural textures can be usedas a starting point for further texture painting.Use Normals to make a low-resolution meshlook like a high-resolution mesh. To do that,unwrap a high-resolution, finely sculpted meshand bake its normals. Save that normal map,and Map To the UV of a similarly unwrappedlow-resolution mesh. The low-resolution meshwill look just like the high-resolution, but willhave much fewer faces/polygons.

Can significantly reduce render times Texture painting made easier Reduced polygon count Repeated renders are made faster, multiplying

the time savings

Object must be UV-unwrapped. If shadows are baked, lights and object cannot

move with respect to each other. Large textures (eg 4096x4096) can be memory

intensive, and be just as slow as the renderedsolution.

Human (labor) time must be spent unwrap-ping and baking and saving files and applyingthe textures to a channel.

Bakes all materials, textures, and light-ing except specularity and SSS.

Bakes ambient occlusion asspecified in the World panels (F8). Ignores all lightsin the scene.

Bakes camera-space normals to an RGBimage.

Bakes colors of materials and texturesonly, without shading.

If selected, clears the image to selected back-ground color (default is black) before bakingrender.

Baked result is extended this many pixelsbeyond the border of each UV "island," to softenseams in the texture

Windows Users do AO first: If you are runningBlender on a Windows operating system, youmay have to first bake Ambient Occlusion beforebaking any other option. If you do not bake AOfirst, you may get the error message "No Image

to Bake To" and will not be able to bake anythingfor that mesh.

1 In a 3D View window, select a mesh andenter UV/Face Select mode

2 Unwrap the mesh object3 In a UV/Image Editor window, either

create a new Image or open an existingImage. If your 3D view is in Textureddisplay mode, you should now see theimage mapped to your mesh. Ensurethat all faces are selected.

4 With your mouse cursor in 3D View,press Ctrl Alt B to popup the menu ofavailable baking choices. Alternatively,access the Bake panel in the Buttonswindow, Scene (F10) context, Rendersub-context.

5 Bake your desired type of image: FullRender, Ambient Occlusion, Normals, orshadeless Textures.

6 After computation, Blender replaces theimage with the Baked image.

7 Save the image in the UV/Image Editorwindow via Image->Save

(Render Baking information taken from theblenderwiki) Further tips on usage can befound at the blenderwiki

Texture Painting in BlenderartBlender features a built-in paint mode, calledTexture Paint, designed specifically to help youedit your UV Textures and Images quickly andeasily in either the UV/Image Editor window orthe 3D View window.

In the 3D window in Texture Paint mode, youpaint directly on the mesh. In the UV/Image Edi-tor window, you paint on a flat canvas that iswrapped around the mesh using UV coordinates.Any changes made in the UV/Image Editor win-



http://wiki.blender.org/index.php/Manual/Render_Bake

-dow shows up immediately in the 3D window,and vice-versa.

A full complement of brushes and colors can beselected from a floating Image Paint panel in theUV/Image Editor, or a Paint panel in the Buttonswindow, Editing (F9) context. Brush changesmade in either panel are immediately reflectedin the other panel.

When satisfied or at intermittent intervals,please save your image using the UV/Image Edi-tor window.

Once you have un-wrapped your model to a UVTexture (as explained in previous pages), youhave to:

Load an image into the UV/Image Editor(Image->Open->select file) , orCreate a new image (Image->New->specify size)and saved it to a file (Image->Save->specifyfile).

You cannot paint on a mesh in Texture Paintmode without first unwrapping your mesh, anddoing one of the above steps. After you havedone these two things, you can modify the im-age using the Texture Paint mode. Once you do:

In the 3D View window, select Texture Paintmode from the mode selector in the windowheader, and you can paint directly onto themesh.

In the UV/Image Editor window, enable Tex-ture Painting in the Image menu (shown tothe right).In the UV/Image Editor header, click the magicpencil (highlighted to the right).

At this time, you may choose to show the alpha(transparency) channel by clicking the button to

the right of the magic pencil. The dot icon to theright of that button allows you to paint the alphachannel by itself.

Once you enable Texture Painting, your mousebecomes a brush. To work with the UV layout(for example, to move coordinates) you mustdisable Texture Painting. To work with the meshin the 3D View (for example, to move it in 3Dspace), or select other objects, you must leaveTexture Paint mode.

When you enable Texture Painting, use the View->Paint Tool option in the UV/Image Editor win-dow or the Paint panel in the Buttons window tomodify paint settings.

All painting you perform in either window will beinstantly reflected in the other window (if the 3DView is in textured viewport mode). However,the modified texture will not be saved until youexplicitly do so by Image->Save in the UV/ImageEditor window.

If you want to paint directly on themesh in the 3D View window, change to UV/FaceSelect Mode first to show the edge outline of themesh, then switch to Texture paint mode, whichvisually overlays the previous mode, but keepsthe outline.

As soon as you enable Texture Painting or switchto Texture Paint mode, a Paint Panel becomesavailable in the Editing (F9) buttons. This panelhas all the same controls as those available inthe Paint Tool. Use this panel if you are onlyworking in 3D view in order to change brushes(colors, patterns, function). (Texture Paintinginformation taken from the blenderwiki)

Further tips on usage can be found at the blend-erwiki blenderwiki

This is a short overview of some of the avail-able blender texture tools that you may ormay not have been aware of. For more infor-mation on texturing tools and techniques (aswell as other areas of blender) refer to theblenderwiki documentation.

A wealth of information is available there foryou to browse through and use. A big thanksgoes out to our documentation team forkeeping our blenderwiki up to date and easyto use.



http://wiki.blender.org/index.php/Manual/Texture_Paint

http://wiki.blender.org/index.php/Main_Page

Whenever you want to create a digital materialselect the best objects with that surface.Images can help as well, but only with thephysical object in your hand is it possible foryou to study the interaction of light andsurface in detail. And 45 degrees above theX-axis. Select the vertex to extrude [RMB] snapthe cursor to the vertex [Shift-S] Constrainingmoves to an axis and dimensional input worksfor extrusion [E] rotation [R] move (grab) [G]and scale [S]. However with scale you willneed to use a scale factor rather than adimensional input. This is easily calculated bydividing the required finished size by theexisting size of the object.

The majority of shaders we create and the waythey are calculated are in general comparableto an egg shell. The light that hits the surface,gets reflected, it does not penetrate it andilluminate the inside. This limitation for exam-ple also formed the expression plastic render-ing.

The current implementation of Sub SurfaceScattering tries to simulate the penetration oflight into the volume of the model to createproper mass illumination, enabling Marble,wax, skin, and other materials that can besimulated in a more realistic way. Howevernot every plastic surface is the same. And assimple as it sounds, good plastic shaders canalso take quite a bit of work to get right.Specifically when the surface finish is layered,a lot of extra steps have to be included toproduce a realistic shader. The material setupshould be similar to the real world counterpart.

These type of surface finishes very often havea primer layer, two effect layers with irides-cent particles, and one clear lacquered finishlayer. The most common and visible exampleis in transportation design and they are alsocommonly found in both product design andjewelry design. See title image for reference,you can se more examples of Phillip Carrizzihere.

As you can see in the example, we havedifferent specular reflections, we have truemirror reflections, and different colors. Wehave to analyse this structure first to be ableto translate it into a shader. If we break those

visual elements down we would get thefollowing hierarchy:

Translucent, high polished, with a strongburning specular reflection / highlight and amirror effect with a slight Fresnel value.

Translucent, with iridescent particles in aspecific color, producing an individual spec-ular reflections pattern.

Translucent, with iridescent particles in aspecific color, producing an individual spec-ular reflection pattern.

Opaque, basic color for the main object.

In addition to the layering, we also have tolook at the scene and watch for indirect anddirect lighting differences. For example,when you watch cars with iridescent shadersdrive into a shadow, you will notice that thecar lacks changes. Because of a lack of directlight, the specular highlights will change, thediffuse pattern will change.

When you take a look at a car very closely,you will also notice that the specular reflec-tion is not just plain white. It has a nice setof rainbow colors. This, for example, is truewith many plastic materials when they arein direct light. But it is all dependent on theangle of light impact and from where yousee the object. Those colored specular reflec-tions are very often an overlooked character-istic. For example, you also find them inbrushed metal and coarsely brushed alumi-num parts.

Multi Layered Plastic Shaders


Multi LayeredPlastic Shaders

- by Claas Kuhnen

Image1: Work examples fromPhillip Carrizzi jewelry work

http://www.flickr.com/photos/optilevers/sets/72057594066404718/show/

So how do we translate this into a digitalshader? Lets use the material mixer. Techni-cally we do not need the same layer hierarchy.The primer layer, for example, could be com-bined together with the mirror reflection andburning highlight reflection. The reason why Iwould select the material mixer is simplybecause of the flexibility to create individualmaterials and the option to mix only diffuseand specular together.

In Blender this material hierarchy could looklike the following:

Diffuse color valueSpecular reflection - hard burningMirror reflection value - very low Fresnel setupSpecular light model : WardIso

No Diffuse outputSpecular reflection - focused spreadColor Specular - Color value set to 0.02Specular texture is scaled down: Size 0.01Specular light model : Blinn

No Diffuse outputSpecular reflection - wide spreadColor Specular - Color value set to 0.02Specular texture is scaled down to: Size 0.02Specular light model : Blinn

But how do we now create the pointy high-light texture? As a specular texture, we canuse the Voronoi with the Col1 tab selected.This texture can be used for the bump mapchannel as well as the color specular channel.

The bump map will produce the grainy look,the Color specular will give the specular reflec-tion some color. Because we do not use anydiffuse, the specular will still be spread out bythe texture size and normal value.

WardIso will produce very sharp, hard edgedreflections which are a good indicator of a wellpolished surface. Blinn, which has a little lessburn quality, can be used for the iridescentparticles of the following two inner layers. Thespecular reflections of those two layers areless strong compared to the top layer.

By setting the Spec value inside the MaterialButton from white to a different color we canalso tint the complete specular sparkles intoa desired color.

As you can see in the reference objects, thetwo layers have two different colors.

Inside the material node mixer, we combinethese three materials together.

The two specular reflection materials are beingadded together and their color result is beingadded together with the main shader material.

Add and Mix produces different results. In ourreal world objects they add together. With Mixyou can specify if one input dominates more.Add defines how much the second channel is

being added to the first one. Here we alsodeactivate the Diffuse output of the twoSpecular Reflection materials.

This setup should produce a rough represen-tation of the desired plastic object.

Now it is time that we setup all materialvalues correctly to our needs. Using thematerial mixer, we can always quickly turnon/off complete networks,a specific color output, or change the mixingproportions.

Looking at the reference image we can iden-tify that the first specular layer is very fine,is very close to the main burning reflection,and also shows a similar while less strongburn intensity. This tells us that the firstspecular layer has to be fine and not spreadtoo widely. It should start to blend visuallytogether with the main reflection and notspread too much over the main body.



Image2: Blender First Specular Layer Material

Image3: Blender First Specular Layer Texture

The reason for that is that the particles in thesecond layer seem to be denser together andthus react more to the incoming light. Thisalso means the light direction is important tokeep in mind.

In my example I set the Spec value to 1.17,Hard to 20, and Refr to 8.27. This produced thedesired shape and coverage. With the Norvalue we can shape the material in a way thatthe individual reflections will stand out orblend more together. I want them to blend alittle more together, so I use a value of Nor 0.1.

The next specular layer, the first layer appliedto the body has a much coarser pattern. It alsospreads much more over the entire body andhas a much less strong intensity. I set Spec to0.52, Refr to 3.47, and to get a more spreadHard to 10.

This produces a quite close representation.Depending on scale and view distance, wehave to adjust the texture values. However ifwe compare it to the real object, we alsonotice a significant difference. The spread inthe digital model is too even. To help this itwould make sense to actually paint the tex-tures in Photoshop instead of using a proce-dural texture. We would only need a coarseand a fine point map.

The transparent areas in between will allowthe surrounding area to remain transparentwhile keeping the point size the same. Some-thing difficult to create when you can only usescale to bring points closer or more far awayfrom each other.

Additionally, it will be very visually interest-ing to also add a soft iridescent color changeover the complete visible body, mainly at theobject edges. Make the color changes verysoft. However if we look at Phill's object, wecan see that there is no typical iridescenteffect. The edges are darker simply becauseof the lack of glowing metal pigments. Weplainly see the base color of the primedsurface.

Also keep in mind when you work from animage, that the reflection in your image isshowing the environment of that shot. Ofcourse, in that case it will be difficult toproduce the same looking reflection in yourwork.

Reflections always depend on what you putyour object into in terms of physical objectsaround it and the light setup. If nothing isthere to reflect, then your reflections willalso look logically empty, which can be thecase in studio setups.

In addition, because in nature there are no100% dark shadows and every illuminatedsurface illuminates the environment, we alsohave to take this indirect illumination intoaccount. A very simple way to make ourmodel look better without using time con-suming AO, is just to utilize a filler lamp. Nextto the speed advantage, this also has theability to simulate color bleeding, whichBlender's AO cannot produce at this point.



Image4: Material Mixing

We can easily simulate the color bleeding ofthe ground color to the object, by giving thelight the same color as the ground plane. Wejust need to place a lamp below the groundmesh and let it shine upwards.

But which lamp to use? Sun produces perfectlinear light rays, but we cannot specify anendpoint. Point and spot are emitting lightfrom one point. With the fall off value, wecould specify where the reflected light wouldstop. Only the lower part of the model wouldreceive bounced light. But we would need tocreate an array of lamps to produce an evenillumination.

Area lights produce a nice even illuminationand with distance, we can somewhat controlthe fall off a little bit. Let's use the last one.With a very low Dist value of 0.3, I place thelamp slightly below the ground. The more Imove it away, the more the light intensity islowered. Please keep in mind that in case youwant to increase the area of the lamp use theSize slider and do NOT scale the area lamp. Ifyou scale, you also scale up the distance ofthe fall off.

In case you change the Size value, you in-crease the surface from which light will beemitted and thus the overall illuminationintensity will be different. Scale it down andthe light rays will be compressed and illumi-nation intensity will be very strong. Use theDistance indicator as a reference for scaling.See image5.

First, I just set a distance and then scaled thelamp to the right fall off result. In the followingstep, I did use Size and Energy to get the

desired area illuminated with the desiredstrength. Scale and Blender dimensions arethe key here.

Inside the Blender scene are two cameras.One is nearly parallel to the ground. I use thatone to check how well the area light is illumi-nating the mesh from below. I use multiplecameras to quickly switch between differentviews for testing light and material settings. Iprefer having individual cameras, than havingto move the camera all the time.

In addition we could also simulate the colorbleeding a little bit here. All we need to do isto create a spot light with a very soft edge andlet it shine downwards onto the floor.We have to lower the light intensity and alsogive it a color equal / darker than the bodycolor of the plastic. We want to limit the Lampto only illuminate the ground plane.

For that we have to put the ground planemesh and the lamp onto the same layer andactivate the layer option for the lamp underthe lamp buttons.

The illumination should only be a very smalleffect. Depending on ground and body colorthe bleeding effect would be more or lessvisible.



Image5: Area Light.

Image6: Spot Light Layer.

None: Energy 0

Too Much: Energy 1.00

This should give you a good idea about whatto look out for. The supplied test scene forexample, would require some work in the lightset up.

Without a scene around it, all reflections lookdark. It even looks like there are no reflectionson the object.

All this will change with some more work andactually having the bracelet relating to anobject. Imagine a scene in which the braceletis on an actual arm. With the reflection of theskin and the environment, it will look verydifferent.

Initial material setup from Matt.http://mke3.net/weblog/car-paint-redux/

the distraction inner reflection, but thenice window like reflection on the upper andright part.



Just Right: Energy 0.50

Image 6: Spot Light.

Image 7: Finale Rendering with slide Fresnel reflec-tion value .

Image 7: Finale Rendering with slide Fresnel reflec-

Germany

MFA 3D Studio Jewelry/MetalBowling Green State University, USAFocus in Functional Metal Art and 3DDigital Art

Dipl. Des. (Fh) Color – Advanced ColorConcepts HAWK University of AppliedScience and Art, Germany Focus on Func-tional Graphic and Product Design

After grad school I taught for one year atthe University Wisconsin-Stout where Iintroduced Blender for industrial designand interior design to students. Throughthat exposure to the students I focusedmore on researching the usability ofBlender for this field.

It has increased my knowledge and un-derstanding to see how NURBS and SDScan be combined in a professional workflow for CAD and Rapid-Prototyping usingBlender Blender proved itself to be actu-ally not only quite useful but rather beinga real treasure and workhorse for thedesign students.

www.ckbrd.de |- [email protected] 309 9795

http://mke3.net/weblog/car-paint-redux/

www.ckbrd.de


Normal maps have become very popularthese last few years, shown as a large im-provement in the Game Industry, but also inmore traditional CG Industry and in the movieIndustry, even if less obviously. Nowadays,solutions to sculpt models to an incrediblelevel of detailing exist (ZBrush, Mudbox alsoour beloved Blender with its now famousSculpt mode) but building normal maps toenhance the surfaces of our models is stillsomehow lacking. Blender now has the capa-bility to bake Camera Space Normal Mapsdirectly from a higher level of resolution andapply it to a lower level of resolution, thanksto UVmapping and the Multi-resolution mesh-es. But the very last step, the one which willbring Blender up to the standard is still miss-ing): and that is Tangent Space Normal Maps.

If Blender now supports these maps, it stilldoesn't know how to bake them

These are different tools for different jobs.None should be considered better than theother, and sometimes, they should becombined/mixed in order to get the bestresults possible. They all rely on the normalsof a given object, so this is the first thing wewill try to address in this article) :

A normal is a vector perpendicular to thesurface, at the given point. In polymodeling,as complex surfaces are made of small planarfaces, you can assume that there is only onetrue normal byface but as manynormals as thereare faces. This iseasily viewed inBlender. For exam-ple, select yourmodel, and thengo into Edit mode[Tab]. In the Editing buttons [F9], look for theMesh Tools 1 panel. It contain several buttonsof interest, especially Draw Normals (displaysnormals as blue lines) and NSize (changes thedisplayed length of the blue line for bettervisualization). A normal vector has a lengthequal to one unit.

A Bump map is basically a greyscale imagethat will tell, for each pixel of the renderedsurface, if the local normal length should be

scaled. If not, the pixel will be interpreted aslaying on the 'ground' level of the face(middle grey pixels of the Bump map); if yes,it will be interpreted as laying above theground (dark pixels of the Bump map) orunder the ground (light pixels of the Bumpmap); a 50% grey will mean no scaling of thenormal. A Bump map basically simulatesembossment of the surface, effecting theshading by virtually displacing the renderedpixels along the normals.To get a Bump effect, you need to associate

a greyscale texture (it could be an image ora procedural texture) to a texture channel,and activate the Nor option in the Map Topanel of the Material buttons. Use the Norslider, in the same panel, in order to modu-late the intensity of the Bump effect.

The Nor button has three states:: the texture channel doesn't affect

the rendered output : the texture channel does

affect the rendered output : the texture channel does

affect the rendered output but thenormals are inverted (bumps becomecrevices)

Tangent Space Normal Maps for Blender


Tangent SpaceNormal Maps for Blender

- by Olivier Saraja Image1: On the right, you can see the greyscale im-age used for the Bump, and on the left, the ren-

dered result.

Very much like the Bump map, the Displace-ment map will also use a greyscale image inorder to actually displace the polygon alongits normal according to the value stored in thetexture. It is really efficient with a finely subdi-vided or subsurfed mesh. You can check theDisplace modifier in order to learn more aboutthis, as it is beyond the scope of the article :Bump maps and Normal maps actually fakecomplex surfaces on simple geometries, whereDisplacement maps need a very dense geom-etry to be effective at all.

In the case of the Normal maps, the normalsvectors are all normalized (e.g., they have thesame length equal to one unit) and a colorimage is used to tell, for each pixel of therendered image, what is the simulated orien-tation of the local normal at the pixel. TheNormal map basically simulates the local ori-entation of the surface by virtually rotating thepixel according to the normal. The normaleffect will be quite visible even with non (orvery slight) specular materials. To get a nor-mal effect, you need to associate a colortexture (an Image type one, because normalmaps need special tools to be produced) to atexture channel, activate the Nor option in theMap To panel, as previously, but also toactivate the Normal Map option in the MapImage panel of the Texture buttons [F6].

Since version 2.43, Blender has been able tobake Camera Space normals into the texture

of an object. This is only a step toward aproper production of Normal Maps withBlender, because Camera Space suffers fromsome limitations: the normals baked areonly used with the object with the samelocation/rotation/size versus the samelocation/rotation/aperture of the camera. Ifeither the camera or the normal mappedobject are moved, deformed or animated inany way, then the normals will look wrong.

Of course if more Normal map types exist inBlender that would be quite useful or handy :for example with an Object Space normalmap animations of onlly non-deforming ob-jects could be carried on, while with a Tan-gent Space normal map you can alsoanimate deforming objects (using arma-tures, shape keys, softbodies, lattices, what-ever) while retaining a decent normalvisualization. Unfortunately Blender is notable to bake these kind of normals at themoment, so you will have to rely on externaltools to build efficient Tangent Space normalmaps.

Baking the Camera Space normal map intothe texture of an object is pretty easy. Selectthe object you want to bake then UV unwrapit in UV/Image editor add a new image(menu Image >> Add...) and then go in to theScene menu [F10]. Here in the Bake tabchoose the Normals option and hit the BAKEbutton. Don't forget to save the Normal mapgenerated in any file format of your choice(Targa is the default, PNG tends to be veryconvenient).



Image2: The Map To tab, showing the Nor option andits slider.

Image3: The Map Image panel with the Normal Mapbutton enabled.

Of course baking the normals of a mesh inorder to use it with the same mesh is almostpointless. The interest is to use this bakingfeature with multi-resolution meshes. For ex-ample, you start with a low poly mesh thatyou unwrap, as specified before. In the Editingmenu [F9], find the Multires panel and clickthe Add Multires button. Add Level and startsculpting your mesh, adding details alongwith levels. When you are satisfied with yousculpting, you then can check that the maxi-mum Render level is set and proceed with thebaking, just as explained before.

In the previous image, the Normal map hasbeen computed in the Camera Space for thecentral cube, but applied to a low-poly in-stance of the same cube, slightly on its right.Of course, at first glance, the fact that thenormals are wrong (they are not exactly in thesame Camera Space) is not easy to spot in thisstill shot, but within an animation, wrongshadows will quickly betray the wrong Normalmap.

Testing this kind of Normal map within ananimation should be a good illustration of howmuch Camera Space could be useless whenused on objects that actually move or arerelocated in a different location than the onethe Normal map has been built from. Unfortu-nately, as stressed before, the Camera SpaceNormal map is the only one that Blenderactually knows how to build. Object SpaceNormal maps would be more suited to anyanimated rigid model. Finally, only a trueTangent Space Normal map would be useful atall with any animated deforming object. The

last part of this tutorial is about how toproduce such a Tangent Space Normal map,with external tools, and how to use it withinBlender.

There are many free tools available on theweb for building Tangent Space Normalmaps, but only a few are available for manyOS at the same time. One of these, actuallystands out as having a quite good supportfor Blender, is the Drag[en]gine Normal MapGenerator and is available for Windows,Linux and Os X. Source codes are also availa-ble under GPL v2.0. Go here.

Go to the Downloads section and select thepackage you want to use. There are depend-encies to solve for Unices and Os X, butnothing too fancy (libfox1.4, libfox1.4-dev,libpng3, libtiff4, libxmu-dev are a few exam-ples) and perhaps link the libtiff.so.3 tolibtiff.so.4 (sudo ln -s libtiff.so.4 libtiff.so.3).

Once DNormGen has been extracted, payattention to the dnormgen executable andthe dragengine_dim_export.py file in thescripts/ directory. There is also a lot of goodinformation on the usage of the tools in theREADME file.

The Blender script that is shipped with thetool will take two models : the low-poly oneshould be named *.low (in the OB: field) andthe high-poly one should be named *.hi (*being the original name of the object). Thisis absolutely mandatory for the script to



Example of Normalmap baked withinBlender from a highlevel multi-res of a

sculpted cube

Image4: The Baking Tab.

Example of render result, from left to right: normallow-poly cube, high-poly and sculpted multi

-res cube, and normal low-poly cube with Normalmap issued from the sculpted high-poly cube

http://epsylon.rptd.ch/denormgen.php

work, and could be done in the Link andMaterials panel of the Editing menu [F9] but-tons. It will then export the two models in the.dim format, which is specific to the normalgenerator tool. Then this tool will be able tobuild a Normal map from this .dim file.

There are quite a few steps to do withinBlender, before using the script, like renamingthe objects, setting Subsurf on the high-polyone, and selecting the two objects to export.The exact procedure is as follow :

1 Select your low-poly object and add to itsname, in the OB: field, the extension .low.Unwrap it, give it a material (if it doesn'talready have one) and a texture (Image >New... in the UV/Image editor) that you needto save (Image > Save or Image > Save As...).

2 Now select your high poly object and add toits name, in the OB: field, the extension .hi.

3 Add a Subsurf modifier to the high-polymodel (a Level 0, as specified in the docu-mentation, is no longer possible in Blenderwhose minimum is 1; if you don't want toalter the shape, use Simple Subdiv. optioninstead of Catmull-Clark).

4 Select the two objects : the low-poly one andthe high-poly one.

5 Change one of the views into a Text Editor.Use File > Open... in order to load thedragengine_dim_export.py script. Hit [Alt]+[P]with the mouse cursor pointer in the TextEditor to run the script, choose a name forthe exported .dim file (or accept the defaultone).

That's it for the part to be done within Blender.You should now have one .dim file that youcan feed the DNormGen tool.

When you run the DNormGen tool, you will gettwo distinct windows :

1 this windowwill show the 3D models (low-poly,high-poly) and will let you configure thekind of map you need (Displacementmap, Normal map, Tangent Space orObject Space, etc.)

2 this window will showthe resulting maps.

At first, you will need to load the previouslycreated .dim file in the DE Normal Map Gener-ator (File >> Open...). The original low-polyshould appear on the screen. Using the LeftMouse button, you can click and drag the 3Dobject to make it rotate around its center. Youcan also zoom in or out by holding [Shift]down and using, the very same way, the LeftMouse button. Now you can check and ex-plore the View menu. You can decide to showthe Low-poly model or the High-poly model. Ifthe maps have already been computed, youcan also show them in this window.

The next step, for the generation of a TangentSpace Normal map lies in the Texture-Mapsmenu: you will have first to define the TextureMaps Size, by setting the Height and Width.These parameters defaults to 256.

Now, you can pro-ceed with the gener-ation of the Normalmap. In the Texture-Maps menu,select GenerateNormal/Displacement Maps and pay afew seconds atten-tion to the pop-upoptions window.

At first you need to specify if the High-polyshould be smoothed before proceeding(check box Smooth Hi-Res Mesh Normals)then, you can select the kind of NormalSpace map you need (radio button Tangent-Space or Object-Space).

After accepting the options the texture mapis generated and the Texture Preview win-dow automatically updated. You will notethat it shows the UV coordinates set on thelow-poly object and that the newly com-puted Normal map fits the proper space.From the Texture-Maps menu Save NormalMap in the location of your choice. It shoulddefault to a normal.tga name.



The borders of my Normal map are slightlymessed up? In View > Set Hi-Res Mesh Options,increase very slightly the Vertex Offset option.Alternatively, using a greater map could help,too. The bumpiness given by the Normal mapis not strong enough? In Texture-Maps menu,select Apply Detail Normal Map and experi-ment with the Strength value. Applying it withthe default Strength of 50 could lead to dra-matic results, so use carefully. I read some-where that the Green layer of my Normal mapshould be inverted. Is that true? It used to betrue, yes, but Blender's normal mapping nowis fully compatible since v2.43 (previously thescale of the Z vector differed). So you don'thave to invert the Green channel in a softwarelike The Gimp anymore in order to use DNorm-Gen normal maps within Blender if you areusing the latest version.

You now have a correct Tangent Space Normalmap to be used in Blender. Select the Low-poly model you unwrapped a few minutesago. In UV Face Select mode, in the UV/Imageeditor, discard any temporary map you mighthave and load the normal.tga built by DNorm-Gen using the menu Image > Open... In theMaterial buttons, menu Shading [F5], checkthe Shaders tab and activate the NMap TSbutton.

In the Map Input tab, the texture should usethe UV coordinates, and in the Map To tab, itshould be set to Nor channel (invert it if youfeel the shading is wrong). Finally, in theTexture buttons, in the Image Map panel,check that the Normal Map button is activated

and that the map selected in the Image panelis the normal.tga built by DNormGen. Save [F2]your project, that's it, you are done with aTangent Space Normal mapped low-poly mod-el!



Great news for character animatorsespecially those who use blender formodeling and animating charactersfor computer games.

Blenderhead Niels Krook has justreleased a script titled AutoRig. Thescript automatically rigs up the char-acter and all you have to do is modelan character change few propertiesand fire up auto rig from the scriptsmenu and your character is ready foranimation.

The AutoRig script can be down-loaded from:http://niels.degooier.net/blender_scripts/

AutoRig also come with installationinstructions as well as complete doc-umentation on how to use it.

Niels website has more scripts foranimators go check it out.

http://niels.degooier.net/autoRig.blend

http://niels.degooier.net/autoRig.blend

'KeepTheFlame' is one of my works inspired bythe book of Pete Drapera, "Deconstructing theElements with 3ds max 6”. Even though thisbook is aimed at users of 3ds max, I recom-mend this book everyone.

Before we start to make materials we have totake a minute and wonder about characteriza-tion of our object. In this tutorial we make aflame from a lighter, lets think about it?:

Flame is fading out with the environment It has two different colors, half blue and

yellow Each half has a different property..

Now it seems that we can't make this flamejust from a simple material. So to do thismaterial, we will use Node materials. In thistutorial we use a simple object Sphere scaledin Z. The Flame in the 'KeepTheFlame' was alsomade in this way, but it takes a little time andeffort to set up everything to have this effectfelt right.

First we make a simple material and Turn onZTransp , Alpha 0,0 , Spec 0,0. Now for theNode, in Node editor we erase everythingleaving only Output (I am not going to explainabout the properties of every Node we havein the editor, I will only show you how toconnect them and what this connection re-sults in). Now add Output, Geometry, Normal,Color Ramp, and RGB Curves nodes. Mix andconnect them like on the image below.

We use a setup like this almost all throughthis tutorial as it gives great adaptability. Inthis case the object is covered with twocolors which are setup in Mix Node, and MixNode is setup by ColorRamp and NormalNode. So we can setup fading one color toanother on the object Normals. Later wechange Normal Node to Texture Node.

And this is the result of this setup. Gentlefading of the white color in the black andbecause we also put this setup in the alphachannel the object disappears and fades outwith the background.

Now we have an object that is gently fading,so now we have to split him into two differ-ent parts. First we make Blend texture likethis(Image on Next page.).

Creating Realistic Flame In Blender


Creating RealisticFlame in Blender

- by Krzysztof Zwolinski

Copy all of the Nodes and change Normal,ColorRamp on Texture, Mapping and set upadequate coordinates of the texture,

and the results.

Our flame is now split by the texture in twocolors, white and the black. In the alphachannel, black is transparent. So now we putall of the Nodes made in first step to fadeoutwhat is in the white slot of Mix Node.

Create another Blend texture, which will re-place the Black slot in the Mix Node and add

another Mix Node and mix them together likethis.

If we render it now we can see that the bluepart of the flame is not fading out, that'sbecause the Mix from Normal fade in is notattached into the blue slot. So now copy themand set them up like in the image.

Also copy Normal and ColorRamp and putthem into the black slot, this give us theability to change the outer line of the bluepart. If you setup everything correctly youshould be have some thing like this.

Now we can take care of the upper part ofthe flame. Make another Blend texture likeon this image.

And connect this texture to the first MixNode we created to the white slot



Now we have complete setting for the flame.

Now its up to us and what we want toachieve. We could add a Material Node withEmitter and cloud texture. With emitter wecould use Radiosity and the alpha channel inthe clouds to make the flame more real.

You can see that making a good Node materialis hard work and you need to study them in

real life. Most of time the best results aremade by trials and error.

Information in this tutorial are only the basicsof the Node Editor and making of the flamematerials, but I hope that the basics and studyof the image of the final settings of Nodesused in the work 'Keep The Flame' will let youto make more realistic flames and many moreNode materials in future.



Poland

I have work with 3D softwares for almost10 years. I started out on AutoCAD thenmoved to 3dsmax. Throughout this time Iworked on many programs but foundBlender to be the best of them.

I currently study journalism with special-ization on "CyberMedia and ElectronicalCommunication". In my free time I workas 3D graphic freelancer.

[email protected]


Since the development of Yafray has sloweddown a little, I was frantically searching forsome other external renderer which could beused with Blender. So for the past few monthsI have been trying my hands on renderers likePovRay, SunFlow, Indigo etc. Then just amonth ago I bumped into Kerkythea whichdoes have a bit support for blender too. Butwhy Kerkythea?

It has a good development pace It’s the most feature rich renderer for Blend-

er. Has six different global illumination methods A very easy to use G.U.I. (this is the best

part).

A superb physically accurate material editor. Layered material support Download-able materials Has a python exporter from blender Has a Blender build with Kerkythea inte-

grated just like Yafray.Supports windows and Linux platform

That’s really great. But still it has got someglitches which can be easily sorted. Here arefew to name

Its integration with blender is not on parunlike Yafray.

Its separate build and exporter are still inW.I.P. So if you want a renderer that is tightlyintegrated just like Yafray, than Kerkythea isnot for you.Currently Kerkythea is good for stills only.

But even after all these cons, the result I gotfrom my initial renders are so good that I justcouldn’t ignore it. So here is my tutorial sothat everyone can try Kerkythea. This tutorialis going to be about a basic global illuminatedrender. So hopefully every one should be ableto follow it.

Requirement for this tutorial

Kerkythea 2007.Blender to kerkythea build.dlls for the blender build.That’s it.

I have chosen a very simple scene which Ihave re-created from the tutorial for Sunflowrenderer (another very good renderer) which

appeared few months ago in Blender maga-zine. First fire up the Blender to Kerkytheabuild. Create the simple scene as shown inthe figure (fig 1). Assign basic diffuse colourto all the spheres. Place a light in the sceneand select lamp button. In the Kerkytheashadows and photons panel, click on enableand shadow button. Also go to the worldbutton select black colour as the world col-our. I will be explaining later on why blackworld is chosen.

Now press F10 and go to render panel. Hereyou will see a drop down menu just belowthe big render button. Click it and selectkerkythea from it (fig 2). After clicking it youwill see two more tabs will appear in therender panels. Select kerkythea from it.Move your mouse towards the 3d windowsand click on the display type (fig 3). Selectuser preference. Click on the file path but-ton. Go to the YFexport panel and assign adirectory of your choice. Now come back tothe render panel. In the kerkythea tab de-select Exec. Kerkythea button. Hit the renderbutton.

Blender to Kerkythea



- by Abhishek Gupta

What the hell just happened?

You just exported a file in a format which willbe read and understood by Kerkythea. Why Ihave gone for such a painful procedure? be-cause all the settings like material editing etc.will be done in Kerkythea itself.

Here comes the fun part. Fire up Kerkythea2007. (This is the latest at the time of writingthis.). I will be really brief about what you areseeing right now. The Kerkythea is divided intoabout four panels. On the left side is the treefor all the objects in the scene. In the middleis the 3d view. On the right there is a materiallibrary selector and below it a quick previewpanel with a small green render button.

This is going to come in handy. On the top isthe usual file open, save etc. menu. The rest Iwill describe as we move along.

Click on File>>Open or just press [Ctrl+O], navi-gate to the location where you have savedyour export from blender and hit OK.

Welcome to Kerkythea! After loading the sceneyou should notice two major changes. On theleft pane you should see all the objects fromthe scene. In the middle you should see awireframe view of the scene.

Move on to the right side and click the renderbutton (yes the green one). Congrats on your1st render. Now for the magic. In the left panedouble click on the name of one of thespheres. A star should appear in the leftindicating its selected (See fig 4). Now rightclick on it and select edit material from thepop up menu. Now we are in the materialeditor button.

Material editor has lots of tricks up its sleeveso I will leave that for you to explore. For nowjust reduce the specular by clicking and drag-ging on the dial towards left side under colourtab. Press apply and close editor. Do this withall the objects.

Remember I asked you to make the skycolour black? I did it on purpose. To under-stand that, here is a little insight on G.I.(global illumination). GI is a lighting tech-nique which takes into account not onlydirect light from light source present (like thelamp in our scene) it also takes the skycolour and light that bounces off the objects.Since I didn’t wanted the objects to receivelight from the sky I made it black. You willunderstand it more as we proceed.

In the top panel select render and thensetup. In the ray tracer tab click on the AAmethod dropdown menu and select produc-tion AA. In the soft shadows menu selectmedium. Now click the global illuminationtab. Kerkythea has lots of choices for a globalillumination method. In the method panelselect Photon Mapping + Final Gathering(SW). Set the no. of photons to many. In thefinal gathering menu select very many underrays tab. Set the accuracy to very good. Setthe gathering depth to 2 and press OK.



Now click on the render button. Sit back andrelax as this is going to take lots of timedepending upon your system speed.

After the render is complete you can view theresult and save it by pressing window in thetop panel and selecting rendered image. Nowlook at the subtle bleeding on the floor and onthe edges of all the spheres. Look at theshadows which are not that dark thanks toglobal illumination (see figure 5). Isn’t it beau-tiful? This is the best result I have ever seen inan external renderer for blender.

Kerkythea is a very advance renderer andsupport a lot more than you just saw. In thefuture I am looking forward to the followingchanges:

1.A very good integration with blenderso that a lot more parameters can betweaked from within blender.

2.Hopefully with the new upcoming UIchanges in the blender, a lot more

tweaking will be available for externalrenderers like kerkythea.

3.The GPU accelerated global illumina-tion is made available.

Newer Versions of Blender2Kerkythea can befound here.Kerkythea can be found here.

Till then happy blending and rendering.



India

I am currently pursuing C.A. (charteredaccountant) which does take my most ofthe time. But thanks to Blender it has givenme the power to pursue my dreams.

[email protected]

http://cobalt3d.free.fr/images_3dblender/kerkythea/documentation/Blender2Kerkythea_en.htm

http://www.kerkythea.net/joomla/


Triangles, artifacts and seams... all perfectlyacceptable in the real world, but commonlydespised in computer graphics.

Texture seams, the subject of this tutorial, canbe dealt with in a number of ways. Often, theyare strategically placed in some discrete loca-tion: clothing borders, hairlines, under a belt.Textures or materials can be blended to hidethe seams. The method discussed here in-volves the complete removal of seams usingBlender’s Multi-UV layer feature in conjunctionwith Render Baking.

Before we begin, be aware that one problemwith this technique is the possibility of texture

stretching. With the difference between twodifferent UV Mapping layouts, pixel details canbecome distorted. Two ways to combat this isto use high resolution maps and careful UVunwrapping (unlike the ‘quick & dirty’ exam-ples featured here!).

Start by selecting and marking the edges youwant to use as UV seams. Notice the brightorange edges I’ve marked as seams, on themodified Monkey mesh shown in Figure 1,along with the result of the UV unwrap to it’sright.

Figure 1a shows the Mesh panel with thedefault UV set, named “UVTex”. At this point,you may want to give it a more meaningfulname (I’ve used “UV1” in this example).

Create a new image (or, open an existingfile) in the UV/Image Editor, and set the 3Dview to Textured to see the map on yourmodel. After saving the UV layout, I used itas a guide for painting a quick texture in animage editor. The map and textured mesh(with very visible seams) can be seen inFigure 2.

Now we create the second map where theseams will be fixed. Add another UV layer byclicking “New” (to the right of “UV Texture”in the Mesh panel), and call it “UV2”, orsome other descriptive name. Notice in Fig-ure 3a that icons next to the UV layer namesindicate which UV set is active in the editor(cube), and for rendering (camera).

Texture Seam Removal


Texture Seam RemovalUsing Multi-UV & Render Bake

- by Eric Pranausk

To create a new UV layout for this layer, clearthe original seams and mark different edgesas new seams. These seams can cross the oldones, but should not share any edges. Mynew seams and the results of a new UVunwrap can be seen in Figure 3.

Before the second map can be baked, ourmesh must have a material applied, with thefirst map in the color channel. Also, be surethat the correct UV set is used for this map.Figure 3b shows the Map Input tab of ourmaterial, with the texture set to use UV coor-dinates, and the “UV1” layer.

With the mesh selected and in UV Edit mode,the second map can now be baked. Once the

baking process is started, the blank image inthe UV Editor window (Figure 4, right side) willsoon be filled with the pixels of the originalmap but in a different layout. Figure 4a showsthe menu path and hotkeys for the RenderBaking functions. For seam removal, use the“Texture Only” option.

Blender 2.44 seems to have a RenderBaking bug, so 2.43 or a later version may beused for this step.

Once the Render Bake has finished, you willhave a new map as seen in the UV/ImageEditor in Figure 5, and the seams are stillvisible on the mesh in the 3D View. But Figure5a shows that they are no longer on the edgesof our UV layout. Being located more centralto the individual UV islands, they can now beremoved with your image editor of choice.

Figure 6 shows the results of some quickmodifications to the Render-Baked map on

the SubSurfed model. The location of seamshas not been completely disguised in thisexample, but the hard edges have been‘smoothed’ to show that those parts of themap can now be edited without obviousdivisions in the map.



Figure 6a focuses on the softened details ofthe first map’s seams within the secondmap’s layout. Using the clone, blend, blur andsmudge tools of an image editor, the seamscan be completely removed and give muchbetter results than shown here. And remem-ber, to avoid introducing any new seams intothe second map, do not allow your editing totouch the edges of the UV islands. You canuse selection and masking tools to keep yourchanges within predefined areas.

Figure 6b shows that we are now using theUV2 coordinates in rendering, which is neededfor our seamless map (a modified version thesecond map) to render correctly. Be sure tosave the fixed texture, and load it into thecolor channel of the material. Now you’reready to render!

Figure 7 shows the SubSurfed mesh with thefirst map applied. Notice the unsightly seams.

Figure 8 shows the SubSurfed mesh with thefinal map applied. The original seams havebeen subdued, and could have been com-pletely removed . . . it just takes a little moretime than I was able to invest into this tutorial!

This tutorial was made with intermedi-ate to advanced users in mind . . . hence thelack of detailed descriptions or hotkey refer-ences. For beginners and any who requireadditional information or explanation ofterms, please utilize the search functions andother resources at the websites listed below.

http://www.blender.org/tutorials-helphttp://wiki.blender.org/index.php/Main_Pagehttp://blenderartists.org/forumhttp://www.blendernation.comhttp://forums.cgsociety.org/forumdisplay.php?f=91http://www.google.com/search?hl=en&q=blender+3d



USA

I'm a graphic designer for amanufacturing company in California, thehusband of a beautiful woman namedCarolina, and user of a program calledBlender.

[email protected]

http://www.blender.org/tutorials-help

http://wiki.blender.org/index.php/Main_Page

http://blenderartists.org/forum

http://www.blendernation.com

http://forums.cgsociety.org/forumdisplay.php?f=91

http://www.google.com/search?hl=en&q=blender+3d


Blender 2.45 Released!

Blender 2.45 has been released. This is abug release version and a large number ofbugs and fixes have been taken care of tocreate a stable release for the next roundof improvements. Go pick up your copy inthe normal places.

October 1st will see Project Peach officiallystarted, although a great deal of work hasalready been done. Concept art is beingfinalized as well as work on the script as aresult of a pre-production workshop at-tended by our team. There has been achange in the team members as Lyubomirhad to drop out for personal reasons and

was replaced by William Reynish (one ofthe team alternates.)

Arno Kroner, trainer at Disney FeatureAnimation, teaching at Woodbury Univer-sity and organizer of Siggraph ’s FJORGagreed to come to Amsterdam to organizeour first week. He’s planning to do a fullweek workshop about all aspects of visualstorytelling, storyboarding, layout, sceneplanning and acting. He’s also helping usnow as script consultant.

Main Blender feature work is of course thehair rendering and editing. Both topicsshould be tackled by mid November lat-est…

We will need tree/plant generating soft-ware, and probably landscape editingstuff. Here Python might be a great tool.The Blender Conference is set to kick off inAmsterdam October 12-14 at the newlyopened multi-media center De Zwijger,located in the Amsterdam cultural dock-lands, on the river IJ.

During the Blender conference we’ll revealthe movie concept and basic charactersketches.

Pre-orders for the Peach Project DVD arebeing taken through the e-shop. Moneyfrom the pre-sales will help fund theproject as well as individual donationswhich can be made through the ProjectPeach site. http://peach.blender.org/index.php/about

Donations of 30 Euro or more: you can getyour name mentioned in the movie cred-its, like for the DVD pre-sale credits.Please tell us which name.

Donations of 250 Euro or more: you canget your name mentioned as Main Spon-sor in the movie credits.

While you are at the e-shop be sure tocheck out the available books for sale.Not only will it increase your blenderknowledge, but it helps support theBlender Foundation. The e-shop currentlyhas three excellent titles in stock:

Essential Blender

Introducing Character Animation

French Blender Guide

Project Peach


Great things are onthe blender horizon!

http://www.blender3d.org/e-shop/product_info.php?products_id=95


http://peach.blender.org/index.php/about




This render started out with a simple idea thatI had: to create a scene of an old dormer.Soon, I was photographing textures, trying outlighting, and modeling more objects for therest of the scene. Unlike most art, this piecewasn’t planned; it just sort of expanded fromthe dormer out. In this article I’ll explain thecomponents of the final image, focusing ontexturing.

The most common question I was often askedabout the render is, “How did you do thetiles?” The modeling needed the mostthought, but here I won’t go into detail on

that. You can find my tutorial on modelingthose tiles at: blenderartists.

Reading that tutorial will give you some infor-mation on the modeling. I recommend thatyou read it before continuing, so you'll have abetter understanding of the whole process.Once the array modifier on the tiles had beenapplied, I assigned a simple material, called"rooftile" to all the tiles. Having a red-orangecolour, this would be the main colour of theroof.

I then created two variations of the material:one darker and one lighter. I assigned theseto random tiles (having separated the tilesinto individual objects). This gave a morerandom, old-looking effect (as if the tiles weremade of different types of clay). The next stepwas to use the ‘randomize loc rot size’ scriptto slightly jumble the tiles. Then I joined themback into one object.

This re-joining is an important step (althoughit’s not ideal for adjustments of the tiles’positions later on) because it allowed me tohave one moss texture spread out across allthe tiles. I used an empty to determine theposition of the texture.

Painting the moss texture itself turned out tobe a simple process in Photoshop. I addednoise to a plain green image, erased sec-tions, used the clone tool to make adjust-ments, added a bit more noise – you get theidea.

The texture was then applied to the tiles,affecting Col and Nor. The use of the textureas a bump map was essential, because itprevented the moss/dirt from looking com-pletely flat. However, you’ll notice it doesn’tlook particularly raised either; creating thattype of thick moss would require a differenttechnique.

making Of “Roof Tops”


Making of “Roof Tops”

- by Alex

http://blenderartists.org/forum/showthread.php?t=100650&page=2

Joining the tiles created one object with threematerials. I knew that I would have to apply themoss texture to each material to have moss on allthe tiles, but I chose not to. Most of the tiles are ofthe main colour (material: "rooftile"), with only afew lighter or darker ones in the mix, so it createdan effect of varied cleanliness across the tiles.

Texturing the walls didn’t prove to be asdifficult as I had expected. I simply went on asort of texture hunt around town, and endedup photographing a local running track (youmay notice it’s not in great shape) for the walltexture. After a bit of tweaking in Photoshop,I applied the texture to the dormer wall usingan empty (the same method used for themoss). On the dormer, the texture tiled un-pleasantly around the object, but that didn’tmatter since a window and planks of woodare covering most of the rest of the surface.

The textures for the other walls were takenfrom the same place, but there’s nothingusual about how they were applied. The finalstep for the walls, to add a touch of realism,was to add multires to the meshes and sculptsome chips out of the edges of the walls(noticeable on the dormer, on the edge clos-est to the camera). Simply beveling didn’t lookas good.

One of the challenges of texturing is to pre-vent repeating images, especially on separateobjects that all have the same texture. Toavoid this problem on the wooden planks, Icreated one material for all of them with anold wood texture (also required hunting).Then I joined some of the planks into oneobject, causing the texture to be stretched

across thoseplanks. The planksthat I didn't joinwith any otherskept the originaltexture. As long asthere are only afew objects thatremain on theirown (and thuseach have an iden-tical texture place-ment), this createsa look of variation– and only one im-age texture isused.

The material for the shutters and windowframe consists of (from top to bottom) a woodtexture (the same one used for the planks), acloud texture, and the same wood textureagain.

The first wood texture affects Nor only. The cloud texture acts as a stencil, and

also affects Nor. The third texture, wood again, affects

Col and Nor.

This whole setup gives the effect that roughold wood was painted, and then the paint was

chipped away over years. Note in the belowimage that even where the wood isn’t show-ing, you can see the grain, as if the wood isunderneath. This happens because of thefirst wood texture being used as a bumpmap. Also, the Col slider of the first texturecan be adjusted to mimic paint translucency.Try a similar setup yourself, it’s fairly easyand very useful.

That’s all of the most important texturinginformation about this render, and I hopeyou found it helpful. I find that the best wayto come up with original textures is to justget out there and take pictures of them. Thatway you’re guaranteed that: a) The image is100% yours, and b) it’s unique.

BlenderArtists Forum thread here:.

(aka blenditall) blendergallery

making Of “Roof Tops”


http://blenderartists.org/forum/showthread.php?t=100650

http://www.blendergallery.blogspot.com/

A lot of people asked me for a little tutorialabout my skin shader in this picture, andalthough it's quite simple it might not be easyto understand for someone who isn't into thenodes and SSS features of blender yet.

I began, of course with modeling the head andunwrapping it to create the necessary texturesfor the shader(btw my unwrap is really crappyso don't do this at home kids ^^).

I then created new image of size 2048x1024which is kind of small for that kind of texture,but it was enough for me now. I used blend-ers rather new "Bake Render" function tocreate a base layer so I can orientate myselfwith it in gimp. This was my starting point forall the other layers.

On one layer I painted the basic colors withoutshading plus some extras, like freckles,makeup and such things. Note that this layeronly contains colors without any shadingbecause all the shading will be done by therenderer. Basically I keep the whole image avery light skintone except the lips, the noseand other areas that generally appear morereddish because of thinner skin or higherblood vessel concentration.

Next up was the bump map. I started with a50% grey layer and painted black strokesand dots for anything that goes inward andwhite lines for everything outward. This ishow I created pores and wrinkles and thebasic structure of the lips. A few white dotshere and there make the skin look kindaunclean and more natural. Note that differ-ent skin regions have different sized poresand so on, just look in the mirror to find out:).

Last texture I did was the Specularmap,which basically defines what areas are shinyand not. Depending on what kind of faceyou're doing and what happened to the facethere are different areas to paint... normallyit's just the more oily areas like the nose thatare shiny but we also need the lips to beshiny so I painted them a bright white. I alsoadded a slight shine on the forehead and thecheeks. For this texture I didn't start with a50% gray because this would mean 50%shiny everywhere which would be way toomuch so I started with a very dark grey. Ialso kept the transitions between the areassmooth.

making of “Skin Shader”


Making of “Skin Shader”

- by Schuh

So now we have all our textures and areable to start on designing the material. Ibegan by creating a new Nodes materialand adding a simple diffuse matnode,which in normal blender way got thediffuse map set to "Col" and the bump-map set to "-Nor".

Then I activated Subsurface Scattering forthis matnode and set it to a very slighttypical skin setting for a very subtleeffect. With the "tex" slider at 0.5 it scat-ters the texture a little bit and for ussimulates the epidermal scatter.

Now on to the subdermal scatter. I addedanother node, without textures this time, setto a heavy scatter in a dark red tone, whichwas then screened over the epidermal node.You need to play around with the settings alot to get the right effect for your picture.Remember: sometimes less is more. The olderor the masculine the person is the less scat-tering is typically present.

Now onto theshine. So whydidn't I just acti-vate the specularsettings in theother matnodes?

Because blenderblurs the specu-lars togetherwith the diffusecolor. Sometimesthis effect mightbe interestingbut in my case I

wanted to have the speculars crisp. So Iadded another matnode set to specular only,which was again screened over the othertwo nodes. In the specular node I had 3textures: the specularmap set to "Spec", thebumpmap set to "-Nor", "Spec" and "Hard"(this way there's less specular in the dentsand more shine on the Zits :), and again thespecularmap set to "Hard" but with the DVar

slider set to 0.2. This way I got aroundcreating an extra gloss map, because inthis case the shiniest spots are also theglossiest. The Dvar slider is used to tunethe effect down a bit so the gloss is nottoo hard.

Hope this helped :)The rest is basicallyjust good lighting.

- .

making of “Skin Shader”


I decided to write this article after receiving aninvitation from Blenderart magazine to writean article about the lighting of this piece. Tostart with the basics, the modeling was donewith a bunch of extruded meshes with trianglefaces randomly deleted and the materialsused were a glossy material for the glass anda cracked material for the hall.

At first I thought, I’ll do the usual, try to placelamps to give it an atmospheric effect, butquickly found it was harder than it looks. Totry to achieve a good atmospheric effect withproper balance of light and dark is no easy

task. When Broken developed soft shadowsfor all lamps except hemi, the job got a littleeasier, but was still no easy feat. I wanteddecent contrast within the image as well.

At first, I just popped together some lights ofmany different colors with shadow lamps forshadows, but as you can see I scrapped thatidea and went with a largely single colortheme for the lights and very weak AmbientOcclusion. I knew bad lighting could reallybreak the image. I wanted the mood to bedark, but not overly gloomy, but also to havean impression of hope that says it’s not allbad, and there’s light ahead.

These were to provide an effect of looking atthe bright lights through the glossy surfaces.To give decent contrast, I cranked up thepower of the lamps so the gradient from lightto dark is visible easily enough without anycontrast adjustments in post processing. Inboth images, lights placements are usuallybehind several layers and can give the glossysurfaces a kind of glow, giving a better atmos-pheric effect. In the first image, I used nine ofthese lamps to provide evenness and thus notcause any unwanted darkening in the corners.I also set the lamps to sphere, the reason I dothis is because I can get an idea of just howfar any light from those lamps would go.Previous setup attempts included mixing thehigh powered lights through different areas ofthe corridor, but the contrast achieved waspoor in spots and I didn't wanted that.

Shadows can really help improve the atmos-pherics of the scene and provide a better

relationship between light and dark as wellas bring depth in the scene. Because of thehigh filter value of the gloss materials, someshadows came out colored as well, whichwas a wanted effect. Broken’s build helpeda lot, because it enabled me to use the softshadows, which look much better aestheti-cally than sharp shadows. To many sharpshadows everywhere was not an appealingoptions, as the soft effect treatment for thescene would be shattered. In both imagesthese lamps were usually between the brightlights behind several layers and the camera,they’re also usually placed off center, gener-ally at this step lots of test renders weredone to see if the position of shadows weregood. I also noticed they also made someareas almost black but figured that theyhelped the mood

The two types of lamp I just described canmake things look pretty, but that wasn’tenough. I knew I could’ve used a light just infront of the camera to help illuminate someof the closer glossy surfaces a little bit better.This light was mainly for this task, and wasplaced a little towards the corner for thereason that in the first image, the lightingwould’ve looked too symmetrical otherwise.The second image had the camera evenfarther from the bright lights in the back, sofor this reason I needed a way to illuminatesome of the closer surfaces. That light isn’tas powerful and used a different color toprevent contrast problems, but also of aharmonious color to match the colorscheme, these lamps also cast shadows.

making of “Tangrams of Light”


Making of“Tangrams of Light”

- by Adam Friesen

The problem I had with the scene is that itappeared brighter near the camera, and be-cause of that it would be little difficult to drawthe viewers eye down the hall, and trust meI’ve made images where the lighting lookedweak. The general goal for using negativelamps was to be darker near the camerarather than down the hall. These lamps wereplaced near the camera area and made thearea almost black giving a reason to lookdown the hall, and offers smooth increases inbrightness farther down, sometimes darken-ing a little but that’s okay.

The effect is more easily seen in the secondimage, where you can see that the frontsurface is almost in completedarkness and then you see thetransition between light and dark.Sometimes negative lights are im-portant in drawing the viewerseye. Here, you don’t want theviewer looking at the edges or thefront, but draw him towards thebrighter area. As you notice in thesecond image, the first surface inthe way is darkly illuminated togive an impression of it being infront of the brighter surfaces be-hind it and to draw the eye to-wards the brighter area. Noticingthe surface layer, I know I’ve donethis on purpose as part of provid-ing the mood.

So it took me hours of lightingsetup to get a good relation withcolor, light, and dark. Almost any

material can be enhanced with the right light-ing. Lighting to draw the eye, lighting to pro-vide contrast, or lighting to provide aseemingly mysterious glow in an area toattract the viewer. Lighting with shadows tohelp provide the mood, as a great deal ofmoodiness is set by the lights. In this scenebright lights can bring about a generic placemood, but here you see that the lights and thedarkness purposely left and brought in, sets asort of mysterious mood that intrigues theviewer, materials are enhanced like as inglossy transparent materials where a lightfrom behind can give them a surreal glow orSSS materials can use lights to give it a glowin an area that contributes to the scene ratherthan take away from it.

Lighting can help achieve what your visionfor the image is, which is maybe to scare,intrigue, or interest viewers, I also purposelyleft the AO to be very weak as I needed darkspace in the latter ideas of how to light it asI wanted this to be sort of dark but not overlygloomy and have areas of light where youreye doesn’t have to remain in darkness, orif you’re in that dark hall then you may findthere’s hope to be found in the red area. Buttake heed that in scenes like this, atmos-pheric lighting can be tricky and it may takea number of lighting setups and many testrenders to get it right.

This image was ripe for atmospheric lightingand for many abstract type pieces, findingand placing the right lights with the rightshadows could be the difference betweenmaking it good or being slammed into obliv-ion with comments about bad lighting.

Hope you Enjoyed. Happy blending!

making of “Tangrams of Light”


USA

Adam Friesen is a graduate of Maize HighSchool in Maize, Kansas. He is presentlyinvolved in 3D art production and is anactive member of the Blender community.

[email protected]


· Tutorials explaining new Blender features, 3dconcepts, techniques or articles based on current theme of the maga-zine.

· Reports on useful Blender events throughout the world.· Cartoons related to blender world.

· Images are prefered in PNG but good quality JPG can also do. Images should be seperate from the text document.· Make sure that screenshots are clear and readable and the renders should be at least 800px, but not more than

1600px at maximum.· Sequential naming of images like, image 001.png... etc.· Text should be in either ODT, DOC, TXT or HTML.· Archive them using 7zip or RAR or less preferably zip.

· Name: This can be your fullname or blenderartist avtar.· Photograph: As PNG and maximum width of 256Px. (Only if submitting the article for the first time )· About yourself: Max 25 w ords .· Website: (optional)

Note: All the approved submissions can be placed in the final issue or subsequent issue if deemed fit. All submissions willbe cropped/modified if necessary. For more details see the blenderart website.


http://blenderart.org


Cristian Mihaescu -


Cristian Mihaescu -


Al-Haitham Jassar -


George Max -


Krzysztof Zwolinski -


Len van der Westhuizen -


Marcos Pires -


Rafael Tenorio-elGordon -


Yain Rodrigo Vieyra -


Zoltan Miklosi -


Petri Rantanen - Self Portrait

Fairies DragonsFantastic Creatures/Landscape/Plantlife Any otherflight of fantassy found only in the Artists mind


blenderart.org does not takes any responsibility both expressed or implied for the material and its natureor accuracy of the information which is published in this PDF magazine. All the materials presented in thisPDF magazine have been produced with the expressed permission of their respective authors/owners.blenderart.org and the contributors disclaim all warranties, expressed or implied, including, but not limitedto implied warranties of merchantability or fitness for a particular purpose. All images and materialspresent in this document are printed/re-printed with expressed permission from the authors/owners.

This PDF magazine is archived and available from the blenderart.org website. The blenderart magazine ismade available under Creative Commons‘ Attribution-NoDerivs 2.5’ license.

BlenderArt Magazine Issue 12 Texturing Special

Documents