Category Archives: Materials Used

Blender Foundations, One Year Later

Remember how this all started? I followed the Blender Foundations book from start to finish, actually doing everything step by step, to get myself familiarized with Blender. I was done mid-May last year, just a little over one year ago.

Back then I exchanged a few e-mails with the author, Roland Hess, both when I needed help and to point out a few issues for the errata, and a couple of days ago Roland sent me a follow-up e-mail. He wanted to know, one year later and after having completed a few projects of my own, if I felt that the book had indeed achieved his goal of “laying a great baseline for people to develop from”. The answer, while straightforward, is long enough to warrant a blog post :)

Simply put, it is a fact that this book got me through the learning curve as painlessly as possible. “Yeah, but how do you know if you never tried just learning from all the tutorials out there?” Well, I didn’t… not until Cycles came up. And I’m telling you: having to learn this new engine by grabbing bits and pieces from different tutorials out there (after filtering out the bad information) is a very inefficient way to do it. Which is probably the reason why I haven’t given Cycles a proper go yet, and that says a lot about what could have happened if I had tried to learn “the whole Blender” that way. That’s why I still recommend Blender Foundations to people who are looking for a structured approach to learning the app.

The education resources out there, namely the video tutorial sites like BlenderCookie and BlenderGuru, have been growing in an awesome way, but to me the true value in them comes when you already know your way around and are just looking for solutions to problems you encounter, or want to learn the advanced stuff. That’s where I think tutorials are the most efficient – when you have the foundations set and are just looking to build your knowledge, which is pretty much all I’ve been doing since finishing the book. Also by giving you a look at the complete pipeline within a project it gives you a pretty good idea of how to segment areas of knowledge you may want to dig deeper into, and how this or that tutorial may fit into that program.

I’ll also take the opportunity to assess how well the book has stood the test of time. Quite an accelerated time too, since in a year Blender has grown like crazy! Well here’s what: the omission of Cycles is the only noticeable thing. Big features like camera tracking and dynamic paint would probably only deserve a quick mention in a beginner-level book anyway, and the addition of B-Mesh goes mostly behind the scenes – surely the new tools that it brought, like vertex slide, the improved knife and of course the ability to use n-gons are things very much worth learning, but if you pick up the book today to learn basic modeling you will be able to follow the instructions in just the same way as you could last year without B-Mesh. Same goes for the rest of the book. Aside from the very slight differences in the UI here and there, you’ll be able to follow everything and while you’ll maybe miss a few tricks that have been added meanwhile, you’ll learn the basics just fine. And the book was published two years ago..! I said it then and I’m saying it again now: that’s a great future-proofing job.

Ah by the way, here’s a funny twist: the Blender 2.6 glitch on the cover is actually more deceiving now that we are indeed on 2.6x. Before, it was harmless. “Hey, there is no Blender 2.6 yet!” But now people may pick it up thinking it covers Cycles… They should be researching these things though! :)

If I had that Fax Seal of Approval I’ve always talked about, this would still get it!



Learning Texturing – Finishing

Heh, I was going to say that this post could be a let-down after what I wrote about the importance of finishing, and then I picked that title.

Well here’s what: I’m no longer going to follow along with the book step by step to texture the buggy character. I am going to finish the book (in fact I’m almost done, will probably finish tonight), and I will try out some techniques it describes that I’m not yet familiar with. But considering all the things I have and want to do I believe it’ll be much more efficient to refer back to the book and use the techniques it teaches directly in my own projects.

If you remember the post I linked above, one of the things I noticed in that Corolla project was that I wasn’t fond of texturing. I want to find out if this really is so or if I just didn’t try hard enough. If I find that I am indeed just not cut for the job, I’ll just accept it and focus on my stronger skills. If it goes the other way, well that will obviously be a plus!

So what this means is that my next project will be much less ambitious in terms of modeling (although it may include some sculpting, which is also somewhat new for me) and will focus mostly on textures. Stay tuned ;)


This post is part of a series on the book Digital Texturing and Painting.
You can find the base post of the series here.

Previous post on the series: The Goggles (Procedural Textures)

The Massive Mammoth Masterclass on Rigging

As promised here is a review of CMIVFX’s Massive Mammoth Masterclass on Rigging, by Nathan Vegdahl (I’m linking to a presentation video because there is no product page on CMIVFX for the bundle. You’ll find links to Volumes 1 and 2 on the video description).

Well… it is awesome. No really, it is. I’m beginning to wonder if there’s anything short of awesome in Blender education out there. Either there isn’t or all my homework picking the products really pays off ;)

That's the awesome end-result (thanks to Sebastian König for providing the image!)

First things first: video and audio quality are very good. Wasn’t expecting otherwise but hey, I thought you’d like to know :P In terms of the information contained… well the product pages are a good summary of what is included. Just to reinforce though, it is 6 hours full of information.

What I really like about the tutorial is the structure. Instead of jumping right in to rig the model, Nathan first shows you the techniques you’ll be using for rigging different parts of the character, one at the time, with a very thorough explanation and no distractions. This makes the first part of the video a great reference for your future rigging problems, and in fact it is extremely informative. Then he brings it all together and starts rigging the model. Another advantage of this is that it makes the rigging workflow much more clear, since he doesn’t have to break it to explain the intricacies of rigging a leg or what have you.

The fact that the weighting process is detached to a separate module is also a nice idea. It makes it a good reference, since you may have different methods for weighting and just want to use those and then figure out how he sets up the controls. Or you may be tired of auto-weighting or your manual weighting not getting good results, and you can just see how Nathan approaches the weighting process. This is actually my second favorite part of the tutorial (my favorite being the first part where he tackles one rigging problem at a time). Maybe it’s because of the (bad) experience I had while correcting auto-weights on the kiddo character in Blender Foundations, but although Nathan’s method involves a lot of work you really see the results in the much better deformations. The included add-on to go with this method is also very useful. In fact I’m curious to check other rigging resources almost just to see how different the approaches are, because this does look top-notch.

Rigging a quadruped brings some interesting nuances into play. More often than not, the tutorials on rigging you will find out there will be on a humanoid subject, and that alone makes this tutorial more valuable. Rigging a quadruped isn’t just a matter of putting a humanoid armature on all fours and giving it a Spline IK trunk or a tail – there are differences in the way that these characters should move that will influence decisions in the rigging process, and thus the rig itself. So while a tutorial on a biped can just rely on the fact that we possess a basic anatomical knowledge of ourselves and that we can always pose in front of a mirror to figure out how the rig should work, Nathan takes the opportunity of having to rig a mammoth to put a reference image of an elephant skeleton in front of you and making you think to understand how it works. I found this very valuable – not every character that will come my way will be a simple humanoid, and if it weren’t for this I would possibly just adapt my humanoid rigging knowledge to it instead of trying to figure out how it should work.

It’s actually hard to think of negative things in this product. The only thing I can think of I’m not even sure about: while weighting the mesh, Nathan picks two cross-sections of the model to show you how he weights the vertices in them, but then pauses the video to work on the rest of the cross-sections; when he comes back there is quite a lot of work done and many cross-sections set up, so I thought it would have been good to hear why he picked those particular cross-sections, but then again the process is so long and repetitive that it would easily add a couple of hours of video just doing the same work over and over. Plus, to be fair, Nathan does give hints on which cross-sections are important to define, like ones that will be under the influence of several bones. So yeah, it really is hard to point out anything not so great about this, honestly.

Finally, Nathan’s style is very friendly and laid-back. It’s as if you had a friend come over and teach you this stuff. While he’s explaining things clearly he also has fun with funny things from time to time – he’s not joking around, but sometimes weird stuff happens and instead of going silent he just plays into it. That really helps maintaining attention and not getting bored throughout the 6 hours of video, at least for me :)

“Would you recommend this product to your friends and/or family?” Absolutely! My mom would actually like at least some parts of it, like when Nathan shows how you could plot parabolas on the graph editor :P (she’s a math teacher). In all seriousness, yep, if you have just the basic understanding of what bones are and what they are supposed to do, this product (of course with effort on your part as well, as always) will show you clearly how to best rig a character and above all in a very well structured way.

I should get a Fax Approval stamp :P

Learning Texturing – Goggles

Digital Texturing and Painting

Procedural Textures

I’m back with another instalment of this series, this time using procedural textures to texture the goggles. Despite all the tweaking and tweaking and more tweaking, this was actually more fun than I thought ;) Ready for a truckload of images?

The book actually started with the strap, but since Blender doesn’t have a procedural texture type that more immediately emulates fabric, I left that to the end and went straight to the lens casing. This was supposed to be a dark rubbery material, so using the workhorse of Blender procedurals, Clouds, I went ahead and set the base for it (go ahead and open the images in a new tab if you want, a small part of them is cropped by the blog layout):

But we actually want it to be almost completely covered in sand and dust from the desert, so I used a Blend texture to work as a Stencil, and applied another Clouds texture, this time much more fine, to work as sand and dust. Result:

Then we tackled the leather mask, and what a nice surprise. The author (using an early version of Maya) was cleverly varying the colors on the mask (both on the skin and the cracks) by assigning them to other textures instead of solid colors. I thought this wasn’t possible in Blender and was trying to get the same result by using the method described above (a stencil texture). But then I noticed something I’d never noticed: we have compositing nodes, material nodes, and texture nodes! I immediately started playing with those and achieved a nice very weathered result for the leather:

Here’s the node setup:

Cool stuff huh? Next up were the lenses, which were inspired in bug eyes. I used a Brick Pattern node for these, but the limitations were obvious:

Here’s the node setup (aside from the pattern there’s a noise texture to dirty up the edges of the lenses; unfortunately I wasn’t able to make this work well):

So all that was left was the strap that I mentioned in the beginning of the post. My plan for this was to see if there was something in the Blender Material Repository that I could use, and then figure out how they made it. I actually tried manipulating the Brick Pattern to get something useful, but I failed miserably and reverted back to this plan. I ended up finding a Carbon Fiber material that looked more or less like what I wanted. Turns out it uses the Magic texture type, ha! After a few tweaks, our goggles were done:

After all the experimentation of this and the previous chapters, the next one (on the car body) starts this way:

The car body is an example of how I create 90% of my textures.

Gotta be useful ;) Stay tuned!


This post is part of a series on the book Digital Texturing and Painting.
You can find the base post of the series here.

Previous post on the series: The Face (3D Painting)
Next post on the series: Finishing

Learning Texturing – Face

Digital Texturing and Painting

3D Painting

Heh, the age of the book does show a bit. I guess back then NURBS were all the rage, because up until this chapter the author never mentioned UV unwrapping and I’ve been doing it myself (NURBS come with UVs that you don’t manipulate like you do with polygon meshes). So the first part was actually about how to prepare a polygon mesh for texturing. The second part was more fun: painting textures directly onto the 3D model.

This time we approached texturing the dude’s face by starting with the bump map. Since his skin is supposed to be scaly, it makes sense to draw the scales first anyway. So after firing up Blender’s Texture Painting mode, a lot of wrist flicking ensued:

Just like painting an easter egg. Not that I ever did.

By the way, as you can probably tell I only painted one side, and it was automatically mirrored to the other side. Setting up this mirror isn’t as immediate as in sculpt mode, for example, but this tutorial at BlenderCookie will tell you how to do it. Thanks to Adrian for pointing me to it!

Oh and by the way again, remember to save.  It really is a pain to come back the morning after and realize a lot of work from last night is gone. “But I saved my file!” Yeah, me too. Not the image file though. Saving your blend file (Ctrl-S) does NOT save your image file (Alt-S on the UV/Image Editor)! Funky, I know. Don’t say I didn’t warn you. ;)

Once that was done we took it to Gimp and started the translate-from-Photoshop dance. If we used the current image as a bump map there would be sharp gouges, but we want the scales to be softly bumped. So we applied a couple of filters to get that effect: Gaussian Blur and then Value Propagate (to mimic Photoshop’s Maximum filter). Then for the color map we used a grunge map the author provides and colored it green, masked another brown layer above it to give it a slight variation, and then colored the “cracks” more brown with the help of Channels (yet another thing learned). Taking it back to Blender and playing with the influence values, here is the final result:

Not too shabby, huh?

Once again I do prefer the author’s result, but if I didn’t then probably it would have been me working on those Matrix films, wouldn’t it?

Next up we will look at procedural textures, and something tells me I’ll have to cheat a bit more… but we’ll see!


This post is part of a series on the book Digital Texturing and Painting.
You can find the base post of the series here.

Next post on the series: The Goggles (Procedural Textures)
Previous post on the series: The Tires (Displacement Maps)

Vehicle Modeling Series, Done

Yep! In fact it has been done for a while – I said on the previous VMS post that I should be done around now, but after that I went on a spree and finished the thing in just a couple of days :) But… I was planning to do something special with it, since I thought it was too cool to just take a few still shots and post them up. So in the last couple of days I worked on a little something. And here it is (watch it in HD in Vimeo for best effect):

Since the model is not textured and only has simple materials it was a challenge to think of an as-simple scenario, but I think this does it. Quickly accomplished, but quite a lot of fun and learning to do!

So now that it’s over, and to avoid repeating myself, a final verdict of the thing as a product: 98%! I’ve already told you how much of a great catch this is and how I feel working on it has boosted my modeling skills and confidence, so no news there. I nitpickingly took 2% away because I was expecting a clean-up in the end (renaming objects, maybe linking some, and closing a couple of holes that we ended up forgetting about), but that’s kind of unfair because a) you can’t really blame Jonathan for letting those slip after 15 hours of extreme modeling and teaching, and b) fixing those things is easy-peasy and can be left as an exercise for the reader :)

And that’s it! Or is it? No, I still have at least a couple of plans for this baby. Stay tuned ;)


This post is part of a series on BlenderCookie’s Vehicle Modeling Series.
You can find the base post of the series here.

Previous post on the series: Part V (Detailing the Wings)

Learning Texturing – Tires

Digital Texturing and Painting

Displacement Maps

This one was a little trickier to follow. You see, this book was published in 2002, and apparently they painted displacement maps back then. I’m sure they still do, and in some cases it may very well be a good solution, but from the beginning of the chapter I was feeling weird about painting greyscale shapes on a rectangle that I’d later apply to a torus-ish shape. It just wasn’t intuitive to figure out what the result would be, not to me at least.

Then I also had a hard time, for the first time, to translate things from Illustrator to Inkscape. Mesh gradients, for example, are not implemented in Inkscape (or rather, they are not supported by the SVG standard, so they’re not in Inkscape because of that if I understood correctly). I more or less solved that by feathering edges, but then those got funky when I exported to PNG… and exporting to Gimp isn’t as straightforward as one would expect it to be… so after all these setbacks I realized “hey it’s 2011 – all the times I heard about displacement mapping were related to sculpting and baking the displacement from the hi-res mesh”.

So I gave up on painting and picked up my sculpting brush. I tried (not very hard, I confess) to create a similar effect to what the author envisioned, using symmetry to make my job easier, and then baked the resulting displacement into a map (after getting help from Ben Simonds’ recent tutorial on BlenderCookie). The process is not as intuitive as it could be, but that has recently been solved for Blender 2.58 (due any minute now!) with Bake from Multires. Applying the map got me this:

Sorry for the lousy lighting

So now I had to create a color map. I created a layer of solid brown below the copy of the displacement map, duplicated this displacement layer twice and set them to multiply to get me values of dark brown going to lighter brown, which I then just had to enhance by hand-painting on another layer on top of the others. None of this was described in the pages, which made me realize again how awesome the book is – I already just know these things I didn’t before I started.

Then I duplicated this file, made it black and white, sharpened it, and added scratches and bumps to create a bump map. Final result:

The UVs were pretty stretched and there's a nasty seam there, but I really have to move on now. Better luck next time.

Less exciting this time around, but I learned some useful new stuff and now it’s just a matter of practicing. Oh and a final note: in the middle of a thread in BlenderArtists Ben Dansie suggested what looks like a pretty useful resource on texturing. Check it out!


This post is part of a series on the book Digital Painting and Texturing.
You can find the base post of the series here.

Next post on the series: The Face (3D Painting)
Previous post on the series: The Hoses (Tileable Textures)

Vehicle Modeling Series, Part V

BlenderCookie’s Vehicle Modeling Series

Part V, Detailing the Wings

Getting really close to the end now! This part had moooore modeling experience and quite impressively still new things to teach.

That’s right, I thought by now we’d just be using all we’d learned and racking up experience but the Series had at least one new thing to teach me: Lattices! I’d been spotting these on and off on videos on the web but never really caught an introduction to them and their capabilities. Now it’s clear to me that no tutorial on modeling would be complete without explaining this feature. We also revisited constraints to reinforce their utility.

As for the model itself, whereas in the blocking stages I made sure my mesh was coherent with Jonathan’s (but not being paranoid about it), now that we’re detailing I’m giving myself more room to do my own thing. Before it was important that I could follow along, now we’re just doing things and “never going back to them”, so I might as well shape them to my own liking. Overall the design is the same, but some details are more “me” (like the lower top and the wing insertions in the back). Here’s how it’s looking:

That tail fin's lack of details is really standing out now.

Only the tail and the cockpit remain, at this rhythm I may be able to have this finished by the end of next week. Stay tuned!


This post is part of a series on BlenderCookie’s Vehicle Modeling Series.
You can find the base post of the series here.

Next post on the series: Done (Detailing the Tail and Cockpit, bonus Video)
Previous post on the series: Part IV (Detailing the Mid-Section)

Learning Texturing – Hoses

Digital Texturing and Painting

A Small Tileable Texture

Wow, I am learning lots. This is exceeding my expectations quite a lot! As a byproduct I’m also learning lots of Gimp, and even other things: this chapter instructed me to use Illustrator, so I used Inkscape for the first time ;)

So this chapter was about creating tileable textures, which we applied to texturing an engine braided steel hose. Ready for awesomeness?

First step was to define the repeatable tile. Based on a few reference images, an “over-two-under-two” pattern was established, and so we put it together in Inkscape:

Fairly intuitive, Inkscape. I couldn't find direct translations to some of the instructions for Illustrator, but I solved those issues in other simple ways.

Repeat that pattern a few times and you see it taking shape:

Remember the days when desktop wallpapers had tiles like these? Fear not, our purpose is another.

All we have to do is isolate the tileable square and we’re good to go:

Here we go.

From here on we take this to Gimp. These braided hoses are made with very thin steel cables, so with the help of Gimp’s gradient fills we created two layers of weaves, one vertical and one horizontal.

Like this.

Using layer masks (like I said, I’m learning a lot of Gimp!) we applied these weaves to the respective directions on the base braid we made in Inkscape. With the airbrush tool we added dirt and shadow to the edges of the braids, and things are beginning to look like what we want:

Cool stuff huh?

I did desaturate the color map a fair bit after. Before applying the texture to the hoses though, to give them an extra kick we put together a bump map. Without it it just doesn’t look right. It wasn’t just duplicating the color map and making it black and white though. The vertical braids have a darker grayscale value, so they would appear to have different depth than the horizontal ones. So basically we just had to make the tones equal. We also duplicated the layer of dirt and shadow to accentuate the creasing.

Here's the final bump map

Time to apply it to the hose mesh! Simple UV unwrapping, some specularity, color and bump maps applied… and there we go:


I’m really liking this, for some reason I was a bit pessimistic but these results are just too satisfying. So stay tuned, because there will definitely be more!


This post is part of a series on the book Digital Painting and Texturing.
You can find the base post of the series here.

Next post on the series: The Tires (Displacement Maps)
Previous post on the series: The Bandana (Scanning in Textures)

Learning Texturing – Bandana

Digital Texturing and Painting

A Simple Tile and Scanned Texture

This chapter was about using a scanned-in image to put together a texture for the bug(gy)’s bandana.

The author used a paint rag of his, which worked really well. I don’t paint though, and I looked around for something grungy I could use but the closest I could find were my ragged jeans. So I decided to use the author’s scan. Good thing Gimp is able to import Photoshop files!

Obviously the original is in (much) higher resolution

Then I had to fill in the empty spaces. I used the clone tool with a faded brush, which gave a pretty good result in my opinion:


I then had to make the texture seamless. I knew Gimp has a Make Seamless tool but I wasn’t too happy with the result, so I went ahead and did it manually like the author suggested (using the offset + wrap around tool (exactly the same in Gimp) to see the seams, then clone over them again).

Next up, making it look like it’s been speeding in the desert. Sun-bleached, old, dusty… So we desaturated the image, and added a few layers painted over with a custom sprinkly brush of brown and green colors.

Then all that’s left is applying it on the model! After tweaking the unwraps a bit (the model is a bit funky), ta-daa:

Ha! again

I’m pretty satisfied with the result, and it makes me look further even more to continuing this project. Cool stuff!


This post is part of a series on the book Digital Painting and Texturing.
You can find the base post of the series here.

Next post on the series: The Hoses (Tileable Textures)