It has been two years since I started learning Blender in the DK30 challenge, so I figured I’d update on where I am these days with the hobby.

As stated, it has become a hobby. I’ve continued dabbling in Blender, all of it to further refine my own skills and making figures I can print and paint for my own amusement.

I have learned that I have a knack in replicating models from reference photos. As I’ve joined a few miniature wargaming communities, I’ve also come across people who print Warhammer figures. And that is where the bulk of my own fun has been as well.

Here’s a gallery of bigger things I’ve finished since the challenge. Sprinkled in with the Warhammer models are a few of my own, original models as well.

I’ve also quite recently made my first commission piece for a gaming company, and they contacted me for a new model to sculpt and paint. On top of that, I’ve started selling my models on Cults3D. So This DK30 has ultimately created a new hobby and a new passive revenue stream for me.

Week 4: Missed the target date

Here we are, a few days past the goal I was aiming for. As I suspected, gaming with friends got the better of me and the weekend was spent enjoying some racing around various cities in the States. I did open Blender up on Sunday and Monday to get some progress going, but we’re not quite at the finished state where I was hoping to be.

Even though I didn’t manage to get to quite to a finished, print-ready mini, I’m still marking the project complete for my part. I’ve gone from not knowing how to move around a view port to being comfortable with creating new meshes from scratch, sculpting folds for fabrics and at least starting to see the process behind creating something of my own, even if it still looks to my own eyes like crap. This is the first artistic endeavor I’ve embarked on since I created some liveries in Forza Motorsport 4, so in that regard I think I’m doing ok, learning Blender in just four weeks to the extent I have.

What now?

I will try to finish the concept sculpt I’ve started for this week 4 at some point. While doing this, I’ve also started collecting concept art for a crap ton of other projects, and I have more ideas in my brain than I have free time I’m willing to spend sculpting away at virtual things. I’m still planning on getting a resin printer at some point, so even if I don’t commit myself to the creation of new models, I intend to print the works of others, and possibly alter them slightly to better match what I’m looking, and looking forwards painting.

Blender is daunting when you first install the program. But as I’ve demonstrated here, it’s possible for a complete newbie to learn the basics within a timeframe of a month. If I stick at it (and I suspect I will), I’m thinking at this rate it would take me six months or so to become adequate as far as miniature sculpting goes.

Realistically I’m not that driven, but eventually I’ll get there.

Week 4: only days to the end

Hard surface modeling from a vague idea is hard. Regardless, there has been some progress!

Most parts are now present for the skeleton. Bare essentials, as it were. Next, I’m thinking of toying around with the array and curve tools, creating a variety of tubes and piping. Aim is to bulk up the model, fill a lot of the places, cover up some nooks and ugly spots with cabling of all sorts. Most Mechanicum models look like someone gave life to their discarded cable drawer, and I intend to embrace that look. Also under consideration and in the back of my mind is clothing: How little and how. I’m thinking a simulation drop on top of the model once all the appendices are in place, but I first need to learn how that is done.

All in all it is looking like I might be able to finish in time. Epic Store having their insane sales might put a dent in those plans, as I’ve already grabbed three games from there.

Whether I can make it at the end of the week or a week from now, I already consider the project at least a success. I feel comfortable with Blender and I’ve become used to a workflow that works for me. I started this challenge with no real experience of modeling anything, and zero experience with Blender. I’ve already updated Blender three times during these weeks, I’ve created a few models, one of which could possibly be printed and might even be fun to paint. From here it’s just sticking with it, practicing more and forcing myself to try and create more complex things.

Week 4

Two work days behind me, little progress has been made. Total time spent in Blender around three hours. I’ve been toying with the idea of a mechanical leg. It’s still clear that I struggle with the basic creative process and have trouble visualizing the basic shapes from which to start the building process.

This issue is somewhat lessened when doing the Warhammer 40K Adeptus Mechanicus style of cyborg: When literally anything goes, the solution sometimes is as simple as “throw more shit at it”. This leg started as a single piston connected to the cup, and a single beam connected to that. It looked really boring. But looking at some photos, I realized that essentially every Warhammer mech has an extra, backwards bending knee. So I added another beam.

Better, but still too bare. So I doubled the beam with a mirror modifier to be twice the size. Still too simplistic. So I added pistons and brackets for these pistons to connect to. Now were getting somewhere!

To this, I think once I add two types of wiring (one smooth, one corrugated), it will look good enough to pass muster.

At the corner of the “build plate” I’m growing a collection of parts. The mechanicum style is nice in the sense that I can literally build everything from a dozen different parts, just added in different scales on top of each other in interesting ways.

Process continues, one leg barely a model makes.

Week 3 comes to an end

Sculpting of the warlock comes to an end. The staff was whipped up in half an hour, and hands were molded around it after the fact, using an hour of sculpting with podcasts blaring in the headphones.

Sleeve folds are a bitch to sculpt, that’s all I gotta say. I had to go and find a large hoodie to wear so I could see how thick fabric folds and try to put that into the sculpt afterwards. It’s a tough challenge to balance between realistic and stylized, and to reach anything that looks halfway decent.

Buth with that now done, I’m turning my planning towards a new project. Something of my own design. I’m almost certain it will be largely mechanized, Warhammer 40k inspired - even so much so that it could be used as a proxy, perhaps. That is the dream, at least.

Let’s see what I can come up with.

Saturday progress #2

The skirt didn’t take as long to sculpt as the cowl, the form factor is relatively simple in comparison, and the details are larger. To top it off, I created chain link details present in the original model as well.

Spending two consecutive days sculpting more organic shapes did help with getting familiar with the tools. I notice that I barely even need to spend conscious effort when trying to sculpt something, I go straight for the tools and start shaping.

With the hem after I roughed a shaped from a box (totaling maybe 50 faces), I went from Detail level 0.5, to 3, to 5, to 10. I barely spent any time at 5, but I realized 10 was detailed enough for what I was trying to achieve.

The new tool for the day was the process of creating longer chains of objects without deforming any of them. If you just create an array from a link of chain and apply a curve to it, the links will bend and look janky. To get around this, I learned, one needs to create a proxy, parent the items you wish to chain on that proxy and then apply the curve to the proxy. So I create the chain links (two pieces, at 90 degree angles to each other), a plane, and the path. I apply an array to the plane, then a curve to the plane, I parent the chain link to the plane - object, and turn on instancing for faces. This creates a new copy of the chain links parented to the plane for each plane instanced, and these objects will be created on top of each plane in the array. As long as one doesn’t create too steep an angle, this works great.

After the chain link arrays have been created, you can create geometry from the instances, and you have the proper objects you can fine tune if you so wish.

While trying to keep this model printable as well, at this point I’m uncertain as to how well those shoulderpads would be to print. Aside from them, everything else should be printer friendly. I made sure the chain links are always embedded in the cloth at least partially, so they should print just fine with it. Separating the outer layers of the shoulderpads to be printed separately, laying flat they might be printable without too many supports, even.

All in all I think the cowl took about 4-6 hours, the skirt maybe two hours, and the chain links another hour on top of that. Included in the cowl was increasing the mesh density on all the armor pieces as well.

What is left in this model would be to create a bad-ass staff, and to model the sleeves for both hands. And the hands themselves. If I can get that much done tomorrow, I’m well on target for the goals I set for myself for this DK30. After some sleeves what is left is to start on Week four, creating an original mini that should be aimed for print readiness. Several ideas are swirling in my head already, it’s more of a matter of just picking an idea and sticking with it. Most likely it’s going to be a machine-humanoid hybrid, since I’d like to utilize all the skills I’ve practiced over the weeks.

Another option I’ve been toying with is a cyborg-corgi. So it really could go either way.

Saturday progress

The better part of Saturday is behind me, and I’ve created a cowl. The creasing and folding of fabric and how to make it look better is slowly “clicking”.

While the Draw tool is still for the bulk of the rough drafting on every stage, Crease and Pinch tools are indispensable when trying to create flowing cloth. Creasing normally for the inner folds, then pulling the peaks sharper with Ctrl-Crease, then smoothing, adding or removing to equalize the width of any fold… It’s a full day job to try and make look natural.

Another thing with the cowl that I noticed was that you really need to go to a high level of detail to reach the kind of control necessary to make the cowl look good. While I start with Dyntopo constant detail level at 1 for the absolute roughest drafts, next I jumped to 3, and barely 30 minutes went by and I pushed it to 5. At that level the creases could be created to a reasonable degree, I didn’t add or remove any after increasing the detail to 10. To finalize the edges of the hood I had to push the Detail Levels to 20 to get enough control over the edge. I quickly smoothed out all the other creases to this level of detail as well.

I feel like I have to stop myself here for the cowl though, I still intend to get at least the skirt hem done before the weekend is over. So next up, rough detailing the hem.

Week 3 weekend starts

How much has been done over the week?

Not a whole lot, is the answer. There’s some progress. I’m not too satisfied with the look, but I also am trying to keep to a schedule, and so the quality is obviously going down as a result.

I’ve yet to get to the meat of this week, doing more sculpting. I did the Squig at the beginning of this week as a free sculpt, but trying to replicate a model with no reference photos in the program through sculpting has not been accomplished. I do believe that attempting this will give me some revelations and help me take a few steps forward in terms of my skills.

The model here does have the same pose as the previous update, I’ve just hidden it in order to model the rigid armor parts in the correct positions (and utilize the mirror modifier to save a ton of time). Once the parts get bound to bones, they’ll go where they need to go.

Week 3 halfway done

Not a whole lot has happened, mostly life has gotten in the way. I picked out a model I have hanging on a shelf to see if I can get some of that detail and replication going.

It’s an old figurine of the Warlock Tier 6 armor in World of Warcraft. I think, it’s been a few years. This thing dropped from The Eye and Black Temple in Burning Crusade. It has both flowing features such as the skirt, the cowl, and sleeves, and a lot of hard surfaces, like the armor, belt and of course the epic shoulderpads.

My initial thought was to simply throw a rough estimation of a robe over the character and then start sculpting from there, but then I spent more than two seconds looking at the model and realized it’s really simpler to just do the skirt, the sleeves and the hood by themselves, separate. This would also split the modeling portion up nicely, and give me a chance to learn a bit more about the armature binding of parts. I did get the pose close enough that I can sculpt the skirt and sleeves on the pose of the generic human that I got off Thingiverse.

Dynamic looking fabric I imagine is an artform in and of itself, and that will be the biggest challenge for this. I’ve already spent a few weeks doing hard surface modeling, and that part just takes time, but I’m confident I can do that.

Other interesting challenges on this model, and techniques to try, would be to make the chains that hang on the robe, initially with a path tool, and see if that would be a smart way of doing them.

Overall this model would challenge my skills on several parts, especially the sculpting of fabric, but having a real life model to turn and twist might be a boon to figuring out some of the shapes and how these meshes go together.

Hopefully tomorrow I will have some more time to get started. I think getting some of the modeling out of the way would be a smart move at this point, to get me motivated by getting something done and showing on the model.

End of Day 1

An estimated 8 hours spent on sculpting mode, and trying out various tools.

I found it a lot easier to fall back on this kind of malnourished look for whatever reason. Compared to when I was trying to create a healthy looking Squig, this seemed to flow much easier. As a final bit of flourish I tried a stencil for the scratch above the eye. I’m not convinced that was any easier than it would’ve been without the stencil, but I’m sure there are some uses for the stencil I haven’t figured out yet.

The sculpting itself was a lot faster than before. The idea of sculpting first in rough detail and only moving to finer detail once you feel like you cannot refine the shapes any further is a good approach. I did fumble a bit: once I had moved from detail level 1.0 to 3.0, in order to sculpt the gums I turned symmetry off. I still had not detailed the legs at that point, and only after crafting details for one leg did I realize the other leg is completely untouched. Final level of detail was put on with 5.0, which I think will be plenty.

When sculpting this I was trying to keep in mind the possibility of printing this. This is why for example I didn’t carve the mouth at all, but rather capped it right at the teeth. I put in more teeth and shaped the mouth to block the cavity from showing and then sculpted it right against the teeth. The base was also sculpted to conform to the legs to keep them quite level. Possibly the trickiest part in printing this would be the horns, but even they don’t curve too much, and if printed with the base, possibly at a slight angle (face looking down at 20-30 degrees) I’m guessing they wouldn’t even require any supports.

Overall I’m satisfied with the results from a single night of sculpting. Following week I’ll see if the multiresolution approach suits me better. I think I need to get myself a basic human figure as a baseline.

Week 3, day 1

I decided not to wait for longer for the sculpting, and also to slightly shift the goals and aims for the week.

I decided to try and do a squig, just to get a feel for the sculpting tools. I thought a Squig should be comparatively simple for a sculpt, it’s essentially a football with turkey legs and British dental work.

As you can see from that picture, my initial sculpting shows that for an absolute beginner, even Squigs aren’t that easy. I have some peculiar mesh issues that I noticed way too late: Almost the entire belly section, plus random places here and there had some vertexes doubled up, showed up as really weird glitches. There were also a bunch of holes along the centerline of the mesh. Both of these were likely cause by my attempt to use the Mirror modifier on the mesh. Unnecessary, as sculpting tools are mirrored by default to begin with, and also necessary, because it turns out the mirror modifier doesn’t play nice with sculpting.

Lessons Learned: Sculpting isn’t too hard, and the creative process seems a lot easier for me in sculpt mode. After fumbling around with the tools for a good while, I settled on the Clay tool as my go-to. Learning the shortcuts for removal, adding and smoothing tools made the process speed up considerably. Put on a pile of clay, hit control to sharpen the edges a bit, run the smooth tool over places where you need the surface to calm the fuck down… Pretty soon I was thinking in terms of where this frenzied ping pong ball would have muscles and making his legs beefier. At one point I even dove under the squig and gave him pecs and abs, until I realized that’s just silly.

Another lesson learned with DynTopo was to use the distance to my advantage. Crafting the crude shape from a far, trying to make the silhouette as good as you can, then zooming in closer to get the next level of detail and doing a new pass. This seemed like a great method overall, though there were some issues where I zoomed out further to get a better view of a bigger feature, and when putting on clay I lost the details I had already carved in. So while DynTopo is a fine tool, I think I need to switch away from Relative detail to Constant, so the level of detail doesn’t change by zooming in, but by me punching in a bigger number.

I did some googling and I’m going to attempt another squig today. This time around I’m going for an even more freehanded attempt, starting with just the lovely smile, eyeballs and a spherical body. I’m going to save the file here, then make an attempt at sculpting this guy with two slightly different methods.

First, I’ll spend the afternoon creating a Squig as best I can with Dynamic Topology sculpting. I’m thinking I’ll set the detailing to manual as well to force myself to learn to sculpt the rough shape first, instead of laser focusing on making sweet looking gums hugging the teeth while there are triangular facets in his ass the size of said teeth. I should also try and focus on NOT making every part of the sculpt insanely detailed: Most models aren’t, and for a good reason too. If the model is too busy, it’s going to be more time consuming to paint, and if you do paint the details in you’ll have to utilize other painting techniques to make the eye focus on a particular portion of the sculpt. The huge maw and the face I think are good places to focus most of the details in, ignoring the underbelly almost completely, as it’ll be in the shadow, underneath the model, generally hidden from anyone viewing the model. So Pecs as an easter egg are definitely an option.

Second, there’s a modifier called Multiresolution I could apply to the object. With this, I could essentially do the same thing as I am doing with the Dynamic Topology of working in layers from broad strokes to smaller details. In this mode though, I shouldn’t be using Dynamic Topology, and the model should be roughly in the shape I want it to be when I start working already. I believe this is a worthy method to try out as well, and could really lend itself to fine tuning an organic mesh.

This second method I’m not trying out today. Today I’ll focus on creating with Dynamic Topology.

Week 2

We’ve almost caught up to today. This week the progress has been a lot slower, or there’s less to show for it. The initial enthusiasm obviously wanes (I’m pretty peaky with my interests, like Sean explained himself to be), but I also didn’t find a project to really get me interested as much. The Space Marine I consider complete and I will leave to the sidelines. The model has tons that could be improved, and the way it’s modeled doesn’t lend itself too well for printing, which was one of my main goals.

While remodeling the same Space Marine is definitely in the future - I am dedicated to getting a fully poseable and print-ready space marine model into my toolbox - I wanted to go to a different place the next week. I grabbed a bunch of pictures online of Orks - I like the Warhammer design of orks a lot- and started creating random bits and pieces one would usually see on an ork.

The pictures weren’t orthographic, there were focal length distortions on some - I had to resize and rotate the references, sometimes even in the middle to make the front and side references match better - and some I’ve essentially just eyeballed and created taking “inspiration” from the source. Freehanding from a reference image is really damn tricky, and I still feel like I don’t have the workflow to a place where I’m comfortable just doing. I need to consciously decide what tool to use and where, there are long pauses and more than once I’ve wanted to Undo more steps than Blender by default allows me to.

I will continue creating random orky bits for the remaining two days I have this week, and Start the third week on Monday.

Things learned and improved on

While I might not have made as much vertexes this week, the working was a lot faster overall. The bare minimum of modeling, editing vertexes and moving them around, that part is pretty automatic at this point. Taking the sword as an example: I don’t think it took me more than ten minutes to whip that up, and half of that time was spent on tiny little fidgeting things when my perfectionist brain wanted to make it better.

I’m getting more accustomed to all the possibilities of the transformation tools as well. While the first week I was essentially just moving one vertex at a time, I’m now barely thinking about it when I grab a bolt, duplicate it, switch magnetize on and replicate a bunch of them on the surface, then joining them as a single object for simpler management.

An entirely new tools I was testing this week were creating tubes with bezier curves and simple rigid body physics (the chain link is in the picture). Bezier curves were surprisingly simple to pull off, The ease of use gave me ideas for future things to try with that tool. I could not for the life of me figure out rigid body physics for the chain link. They just seemed unwilling to collide with anything I put in the scene. I must’ve watched a dozen tutorials and none of them did the trick. I will ask in discord if someone there has an idea of what I’m doing wrong.

Overall the absolute basics I think I have a decent grasp on at this point, but there are still some stuff that feels really challenging in what I’d still count as basics. Looking at this from a miniature modeling goal in mind, simple physics and cloth simulation are both something that would make a ton of difference if I could get the basics of those down to a level where I could just drape an Ork with chains, or a mage with robes and a cloak, click to animate and let the pieces fall naturally where they should.

Next week

And I think that’s all I have to say about that. Looking forward to week three I’m already daunted at the task I set myself, but I also think it’s good to force myself to learn new stuff now when I don’t have an established method of doing things yet. When I’m still unsure of everything I do, I do think it will be easier to learn new tools, and possibly figure out a more efficient way of doing things I’ve already done a few times, just without the sculpting tools.

Week 1

I actually started my Blender project on the 24th. While I missed the announcement for DK30 completely, that very same Friday I installed Blender and started working on my goals.

I watched three videos to get me started, and in the first week I’ve found two more that are really useful when getting started with Blender.

1. 3D Printing Professor setting up the workspace. This helps you get started for printing minis by setting the workspace in the correct scale, and also sets up some other “nice to do” things for your default scene.
2. Blender for beginners. The first - and only - follow-along tutorial that I did when starting out. Just to get you acquainted with the controls.
3. Low poly vehicle, beginner tutorial. A friend of mine used this one as their first video, and it also has all the essentials shown in a single video.
4. Low poly spaceship in 10 minutes. This video was just more of a motivational video, showing that really you can create almost anything with the simplest of tools and getting the workflow really down to a point where you don’t need to think about the controls. S to scale, E to extrude, over and over again, and 10 minutes later there’s a spaceship.
5. Low poly for cosplay printing. This video shows the workflow for creating a mesh one vertex at a time, which is great for more organic shapes that can’t be as easily reduced to cylinders and cubes.

And with those videos watched, I picked a project and got to work.

Week 1

Day one: After installing Blender and following the sword tutorial, my own project folder was created and I started working on a Space Marine. I found a good, orthographic reference photo of a Mark X Primaris Space Marine armor that had both front and side views. Set the picture for reference for both front and side, and got to work. On that Friday, I spent an estimate of 4-5 hours creating the chest piece of the armor, free-handing the entire thing, vertex by vertex.

Day two, I created the ab plates, and thigh armor. While the chest plate was made completely by hand, I started using primitives for base shapes more on day two. I also used the extrusion tool a whole lot more, instead of placing almost every single vertex by hand. Thighs were created from cylinders and then modified extensively to reach the shape. Total time spent in Blender on Saturday around 10 hours. New tools used this day were the circle cut (Ctrl+R) to split a row of faces and get more definition in faster.

Day three: Shin guards, kneepads and boots. Shoes were scratch-made, shins from cylinders and kneepads from spheres. New tools this day were using the knife to create cuts more freely where I wanted them. I also watched some videos and people telling how n-gons(shapes with 5+ vertices) are of the devil, so while I was cutting and modeling the feet I tried to keep everything made out of triangles and four-sided faces.

Day four: Shoulder plates and arms, groin and the random flaps of armor, and the rubber suit visible in the joints. New tools today were mostly focused on smoothing out stuff. The shoulderpads were created vertex at a time and then smoothed out with tools. Found out you can move a vertex along the edge it sits at by hitting the G key twice, which made the smoothing somewhat easier, and if you select all the vertices on the shoulderpad excluding the edges, you can smooth the center out from the right-click menu (Smooth Vertices).

Day five: Armature and rigging for inverse kinematics. Pretty much what it says on the tin. At first I tried to just autoresolve the whole thing, paint by weight the distortion values for all parts of the armor, but a helpful person on a Discord server told me the proper way to do stiff armor(select the part, select the bone, hit ‘P’ and link to bone). I rigged my marine ready for posing. I did also finish the helmet on Day 5, it was an ongoing process during the week. Aside from rigging and armature building which were both completely new things, I did use a boolean difference modifier for the helmet face holes, which aren’t even visible in the final picture, because of the pose.

Day six was heavy at work, and no progress was made.

Day seven was also a working day, but I slapped together the signature backpack of the Space Marines of the 40k.

I’ve been putting off 3D modeling as something of a casual hobby that I might enjoy. My prior background in arts is art class in high school and some liveries designed in Forza Motorsport 4 (That is the only artistic thing I’ve ever managed to trace into existence). 3D modeling I’ve only dabbled into in university with a mandatory basics of Autocad.

Since I also have gotten back into my old passion of painting miniatures, the price of 3D printers going down and quality getting to the point where you can’t tell a print from a cast, I feel it’s high time to learn how to do at least some basic modeling. I have been putting off learning modeling for a decade at this point, using the high initial step of learning the basic controls as an excuse.

But to quote Picard, “The line must be drawn HERE! NO FURTHER!”

By the end of the month I’m hoping to reach a level where I don’t just trace from reference photos and can actually bring a vision I have in my head into the 3D space without too much of a struggle.