Summertime 3D-fun The conclusion

After laying around on ice for a long while after the modeling was wrapped up a number of months ago I finally got back to it and gave it some well-needed UV:s and textures! The normal map took a couple of rebakes but I found a compromise between either too hard edges or having gradients in my normal map that took care of some annoying artifacts (which probably could have been avoided by paying closer attention to topology and especially triangles way back when I modeled the base shapes and focus on quadflow rather than squeezing down the polycount but live, learn, do better next time).

Getting the generators and curvature map to work my way in Substance Painter was also something of a struggle but the more educational for it.

The end phase was a nice opportunity to try out Marmoset Toolbag to get some nice renders of the model as well as do some photocompositing in Photoshop.

Below is the link to the Sketchfab model:




Summertime 3D-fun: Part 1 – The Model

For a while now I have been working on Tim Bergholtz’s tutorial for an AKM 47 in 3DsMax ( to learn more about hardsurface modeling techniques and the modeling-part itself is slowly being wrapped-up (going into 7,5 hours of UV-unwrapping-material wiee!).





So, because my memory is short and goldfish-shaped, what follows are some in-process pictures and notes taken on some of the more complex and new (for me) aspects of creating this model for future projects. The approach was to divide the model into it’s real-life parts and work on one part at a time, from the stock to the barrel.

The magazine:


The magazine was created by making the top shape and pulling out the edges, add supportive edges with the connect tool and then using the bend modifier. For it to work properly though I had to re-aling the pivot point to the opposite side from the bend. Then using slice plane to create a middle edge and applying the symmetry-modifier to save work. For the chamfered edges: set upp support edges where the bevel starts and ends, make sure there’s an edge in the middle, extrude that edge and chamfer it. If the bend tool didn’t quite cut it, the FFD-modifiers make it easy to fix.


Same technique used to create the details and bevels on the piece showed below by the way plus some smoothing group magic and proboolean subtraction with a box.


This little detail was pretty cool. After creating the basic shape (plane, set up edges, chamfer the corners and mirror it),


go into border mode and select “create shape”


which allows you to  enable a render in the viewport, creating the desired shape.


And the rest is just chamfering the middle edge, extruding and pulling out new faces from the upper poly-edges by holding shift.



Another fun part: After creating the basic shape of this part of the barrel (bit of inset, extrusion, scaling and pulling out enough geometry) you can bevel the desired faces inwards to create that shape. After adjusting the edges…


(pulled out the middle edge sideways a bit to round out the shape. If you’re wondering why the end of the barrel is rounded out there in the lower picture, it’s to give the impression that the pieces are molded together)


center the pivot in the middle and delete everything except the affected faces.


Then use Array with angle snap to create enough copies of the detailed face to close the loop (had to redo this part once because I used a cylinder with 18 faces instead of 24 which doesn’t loop correctly). Weld the vertices together (select all, weld with 0,01 threshold) and you get this, ready to go.



Edgeflow of the piece with smoothing groups before applying chamfer and turbosmooth


Finished highpoly result.


Working with hardsurface on this project is a bit like being Oprah:


Speaking of which, working in highpoly for the first time I quickly learned the importance of support edges and keeping your base mesh edgeflow clean to avoid distortions after applying the chamfer and turbosmooth modifiers.

The front piece was another fun detail43

Created from a basic cylinder, with the slice plane then applied at an angle to create that cutoff shape. Then, using the back face of that cylinder, the inner cylinder, which is then used to cut out the right shape inside the outer cylinder with the proBoolean (subtract)-operation set to “remove invisible only”).



And a lot of fun little floater decals to be baked down into the normal later.






On a final note, toward the end I had 21 versions of the file. Might seem excessive but if you’re a bit of a clutz like me and ever accidentally collapse a stack with turbosmooth on an object without noticing before going over your undo-step limit (Which I was so extremely happy to find out is adjustable. It doesn’t take the infinity symbol as valid input though) you will be VERY happy that you can just merge in that part into your project from an earlier save instead of redoing something like the aiming box above which is comprised of multiple parts and made plenty use of the cut tool. True story.

I will probably think of something I forgot but so far it’s been a really great learning experience. Until later with (hopefully) newfound UV-mapping wisdom!


BGP: Second Post

This week I have been mainly going over assets, UV-mapping, lightmapping and re-assigning material ID:s to make the environment texturing easier and tried out a couple of texture setups to see what would work for the visual style we had in mind. It’s been a bit tricky, since it’s our first time working with 3D in Unity so every step is a bit trial-and-error. Nothing I’d like to show so far unfortunately but here’s a piece of promotional art for Kei to make up for it.


BGP: First post

Having started off Big Game Project it’s inevitable to do an info dump or two on it. Last week I designed and started modeling the three interchangeable weapons for the main character. We knew we wanted one light melee, one heavy aoe and one weapon for the ranged attack so the design kicked off from there. We also knew the game was going for a pun- and reference-heavy humor so designing weapons like The Banhammer (sledgehammer) or The Good Game (Bow) played into that.



I threw out a couple of colour variations for each chosen weapon based off the assembled moodboard:




We knew we wanted the weapons to look cohesive but still be distinguishable by both colour and shape so in the end this was the chosen lineup:



A screenshot from one of the quick weapon blockouts:


GDC: Overwatch animation talk summary

Long time no see!

This post is for once not related to any classes but rather to my recent trip to another continent during which I attended a mixture of various talks, took some photos and scribbled down some notes, all which will be featured here in a pseudo-orderly fashion eventually.

One of my favorite talks was given at the animation bootcamp by David Gibson who spoke about approaching character animation in Overwatch, using the hero Mei as an example:


The first part of the talk was about nailing down the movement for a character based on their personality type and how that is not always directly translated from the initial concept and how he would approach gathering references.


Especially that third point, that references don’t have to be exact. It could be anything, from direct looks to movement style, posture, personality traits, anything that would capture some aspect of the character that was being created. Inspiration for Mei (who firmly sits in the “young, awkward and adorkable”-category) for example came from, amongst others Zoey Deschanel, Rapunzel from Tangled and Giselle from Enchanted.


Also taking input from everyone! It’s highly unlikely that the entire team share the exact same set of references so there’s always new input to be found from speaking to others and seeing what gets them excited about the character.

And then to the practical stuff. Mei’s initial concept art did not reflect her archetype. Gibson showed concept art for other heroes like Tracer and Widowmaker who are firmly planted in an A-pose, feet apart, head up, exuding confidence. Mei was not supposed to be like that. She was younger, less confident and shy (feet together, slightly off-balance, body angled a little away from camera and weapon), which segways into the next part:


(Would love to see the bone structure under that rig. Really need to learn to create better rig controls for the next animation project).


In Overwatch, Gibson said, the idle of the characters needs to convey their personality on top of their movement speed and function in-game. In Tracer’s case that means wide stance, facing the camera head on, leaned back, peppy and ready to go.


He showed some other stills from the game which shows how they break their rigs (McCree’s arm, D.Va’s leg) to achieve more snappy timing and enhance the movement arcs with smear frames in a 2D-style.


It all looked absolutely gorgeous and definitely got my interest up for learning more about rigging to be able to achieve try out these kinds of animations for myself.

Unfortunately I can’t insert videos here but Gibson mentioned other animation principles they were making heavy use to achieve that snappy timing like always overshooting the extreme pose and then holding the key poses a little longer than usual as well as having slightly different timing on arms (total symmetry looks stiff). Tracer’s reloading animation on youtube is a good example of this.

But also: Mariel Cartwright who animated for Skullgirls gave a presentation in 2014 which covers pretty much the same principles (overshooting, holding keys).


Again, can’t upload the video but we got a look at the first pass of Mei’s run cycle. Gibson’s explanation of his process for creating a run/walk cycle was that if you get the full extension keys and passing pose to really match the character, the rest of the animation is pretty much just adjustments.



3D-III Pre-production

It’s been a while!

Since the last update a new, animation-focused course has started, 3D-III. Our first assignment was to (in groups of three) create a pre-production package for a character of our choice. I work with Kim Teroni Borg (modeling main mesh and head, finalizing animations) and Nayomi Arvell (modeling of secondary parts, character background, texturing). My responsibilities for this project are concept art, rigging and skinning and blocking out the main character animations). We brainstormed and agreed that a small, exaggerated, military, napoleon-like character would be a fun and interesting challenge (when doodling I tend to draw women so an elderly man seemed like a good push out of the comfort zone). To give his macho, militaristic appereance a twist, Nayomi came up with a backstory for him about how he lost his leg and got a not quite synchronized, new one. Quick first doodle:


After agreeing on the base concept I explored basic body shapes and facial shapes:


Then after we’d agreed on a body-shape (potato on legs), I made a “character creator” (psd.file) that let you mix and match eyes, noses, facial hair, hats, jackets, shoes and robotic legs. Some mixtures below:




And then the colour palette:



Only sliightly obsessively. Spot the differences!


In the end this became our (ex-) admiral Edward “Ladyleg” Hemsworth. (The funny thing is, if you look back, the final design ended up pretty close to the original doodle throughout the iterations. It ended up very much being an ‘archetype-with-a-twist’-sort of design). In short, his character background is that he used to be an admiral in the navy so devout to his duty that he wasn’t about to let little trifles like the fall of a government and civilized society stop him from doing it. So he still wears the uniform and sticks to his values whereas his crew have taken on a more pirate-y ‘practical’  approach.


For the modeling turnaround, some adjustments were made to make him more suitable for animations. For example covering up the attachment of the leg, moving down the boot below the knee, moving up the belt and adjusting the sleeves:



The final part was planning out the rigging and the edgeflow before going into modeling the base mesh:


Next time, the base mesh, rigging, skinning and some animation poses!

Outside of the course I’ve been slowly working my way through Loomis’s “Drawing the head and hands” which I’ve found really helpful to get a better idea about facial proportions, both male and female.


Concept Art – Week 5

So during these past weeks in the Concept Art course we’ve been talking about character design, shot design inbetween the basics of art like perspective and colour theory. The good thing about that is that even if you’ve heard it before, you can always find something new to take away.

This week it’s been rendering and while the final assignment is to create and render our own character designs for a given script (which is supposed to be our own take on a Alice in Wonderland-themed car commercial), we got an in-between exercise to colour a finished lineart.



We tried two approaches, either painting with grey values and then overlaying colours on top or painting with colours directly. Usually, I find that painting with greys underneath kills the colour quite a bit but that can be remedied by fiddling a bit with photoshop layer and colour adjustments to saturation and contrast and just painting in the stronger colours.



I’ve finally learned to get rid of the lineart! And defining different materials! Greatly helped by the magical acrylic brush which feels more reminiscent of traditional oil painting. Still got boggled down in the details a bit but for once the whole thing is actually done (painting a background was not part of the assignment).

Have a good one!

3D-I Character turnaround and other things.

For our next 3D Assignment (character creation, yay!) we first have to create a pre-production package (which means concept art! Drawing! 2D!). The character had to be reasonably humanoid, not too detailed and stylized. So I came up with this cool chick:


(Not entirely sure about the colour scheme yet). She’s Mojo and she’s supposed to be the chemistry whiz in a 3D platforming game where you play a kid detective, “Codename: Kids Next Door” style. (It’s a work in progress…) Hopefully the concept works for the assignment. It was so much fun to linework again and design for a cartoony style. When I haven’t been doing stuff for school I’ve been trying to work on colours and values so it was nice to step back and put everything into gesture, silhouette, simplifying shapes, overlap and all that lovely jazz we spent much time on during Life Drawing.

I forgot about the action pose-shot we had to do so I quickly threw together the picture below. Since I already had what I needed for the model from the turnaround I gave the action pose a little more “schwung”.


I’m still behind on the second assignment for 3D2 but I’ll crunch it out during this weekend. I chose to step back and redo the top part and arches because it was turning into a clusterduck. Everything is much cleaner now, the half-sewn in/half overlap is gone and now all the separate parts overlap cleanly (as I had planned in the beginning but somehow lost along the way. Don’t model when sleep-deprived kids!). The silhouette also looks closer to the reference material. What’s still lacking are the details on the top to cover up and with around 1000 500 triangles left to spare, that should work out fine.


Below is the final version of the model. Luckily there were few construction errors to fix and it clocked in at a nice 2973 triangles out of the 3000 triangle limit.


On a final note, what’s happening outside of schoolwork are mainly two things. For one, working on getting my painting skills up to a decent level. This is what I’m currently working on and I think as far as as skintones and staging go, this one is the best I’ve managed to make yet. Hopefully there will be time to finish it soon.


The other thing is that I and a friend signed up for King’s release of their Defold Editor to work on a 2D title for mobile! It’s going to be awesome.

That’s all for now!

Theme Park: Pre-Beta – It’s time to penguin up!

Hi again!

At the time of the last post we were in the middle of a hectic pre-alpha scramble with an overhauled concept and a lot to to do with little time to do it. By now I can tell you that we passed our alpha-presentation and have been working diligently  toward our Beta-deadline next week.

With all  basic mechanics in place, we have been working on implementing (amongst other things) new levels, hazards, powerups, a dash-mechanic and a new GUI.

As you might or might not have guessed from the title, I’m going to talk a little about the power-ups in the game. Project “Save the Penguins” has now gone from a puzzle-type of game to a fast-paced  4-mother scramble for 1 penguin baby! You player, the One True Mother of Penguins (Queen of the Antarctic, First of Her Name), have to get hold of the penguin baby and hold on to it. Sounds simple right? And it would be, except there are three other penguin mothers who want the exact same thing and only one of you gets to walk away with the prize. With chaotic and hectic gameplay, we wanted to enhance that mentality with power-ups ranging from the purely benificial to the more risky If-I-Can’t-Win-You’re-Going-Down-With-Me-So-Help-Me!-type of last-minute game-changers.

In the spirit of Mario Kart, a power-up comes in a mysterious box, occasionally dropped by parachute onto the playing field. The power-ups are not explained to the players beforehand. I tried to design them in such a way that the power-up’s icons hint at their function as well as being reasonably easy to tell apart. We tied most of their uses to the core mechanic of the game – bumping into other players so even if a player does not know what the ice-cube-like power-up does, they will find out as soon as they bump into another player!

Psst! Hey. Hey Buddy. I’ve heard you’re looking for some power-ups? I can hook you up, know what I’m saying?

So what’s inside the box?

opening box

We are (currently) planning for seven different types of powerups and will try to implement as many of them before the Beta-deadline as possible. As of this moment we have two.


The powerups are:


1. BlowfishBomb! This one is risky, after being picked up it will start gradually inflating until it explodes! Doesn’t sound very benifical though does it? The player can however pass along the fish to another player by bumping into them.

If you’re a blowfish in this game…You’re gonna have a baad time.

2. FireStarter!

This power-up turns your penguin into a temporary god of fire, leaving a fiery trail behind you (yes there is a bad bowel-related joke there, shush). Other penguins crossing the trail will be set on fire and panic!


Hey you! You’re doing great! With the baby secured and only a couple of seconds left on your meter, there is nothing that can stop y…Why is your suddenly babyless penguin on the other side of the arena? The swapper, that’s why! This trickster power-up randomly switches the positions of all players for better or worse.

4. SnowShoes!

As our fluffy penguins’ coordination would make Bambi look like a figure-skating prodigy the player will need some help to nail those sharp turns. Enter, snow shoes! This fashion-forward footwear will give you much better control of your penguins movement….for a while.

5. BabyCall!

Similar to a dog whistle, this little trinket tells penguin baby that playtime’s over and it will come rushing to your player!

6. Freezer!

This handy power-up gives your penguin temporary mastery of snow and ice, turning every player they touch into a helpless popsicle. Sparkly dress and power ballad sold separately.

7. SuperMoma!

Temporarily granting you superpowers, this power-up increases your penguin’s speed and strength, knocking back the players SuperMoma touches. Comes with a fashionable red cape.

There’s a lot to do left before Beta and besides these powerups there is some work left on the basic mechanics, controls, penguin-customizations, levels, feedback…well mostly everything. But everybody is working hard to make sure we can present the best possible version of our game at this year’s GGC. Do you feel like partaking in some penguin mayhem? Come and see us there!

Until next time!


Theme Park: Alpha Week – Seal the deal

Hi again!

Since the last update our game has gone through a re-haul to a battle arena type of game where penguin mothers will try to snatch the last penguin baby and hold on to it for as long as they can, while avoiding various traps and obstacles! To match the new direction, our input was changed to trackballs.

Tomorrow is alpha and we are feeling the heat but considering the major change that was made last Friday it’s better than it could be.

What I’ve mainly been working on lately are the seals that will circle the arena and drag off unsuspecting penguins. I’ve been doing concepts, movement animations in four directions and an idle animation. Fairly simple, low-frame animations to match those of the penguins but it still took up quite a lot of time.

I started by sketching out concept shapes:


Together with Lead Art it was decided that rounder, softer shapes would suit the overall style of the game best. So based on that I refined the concept closest to that idea, rounded out the face more to make it cuter and adjusted it to the intended perspective. This was the result:


After that it was time to test out colors. I used the color palette in the Style Guide to create five alternatives that would both work with the colors and textures of the art assets while still looking “seal-like”.


With input from the group this selection was narrowed down to two alternatives and tested against the background and other assets (NOTE: This phase happened pre-concept-overhaul, so the ice blocks still follow a hexagon-pattern):


We finally decided on the grey seal since the light blue one blended too much against the water. So going forward I created a turnaround for the seal to base the movement-animations on:


Time to animate! As mentioned before, after looking at the penguin animations and the time available I decided to limit each animation to 6 frames each. After experimenting a bit I decided that since there would not be that many enemies that required movement animations I could afford to draw each frame by hand instead of just putting the body, head and limbs on separate layers, a method that would save time but did not yield the results I wanted.

This was my workflow for the movement animations:

Rough sketch with the primary motions:


Iterated sketch with details and secondary motions:


Cleaned up lineart:




and done!


Look at how happy it is!

Then the same process for the side- and back-animations (the timing is off in these, don’t worry ’bout it):


That last seal is totally going for the Olympics gold medal.

Finally, an idle animation. Seals aren’t all cuddly and cutesy after all…


That’s all for today folks! See you next time!