WIP Art Blog

I’m working on some fancy pentagonal bulkhead based on a design I saw in Ogure Ito’s manga Air Gear. There’s a lot of cool mechanical designs in his works.

I’ve had to add a lot of detail to it. Lots of greeble and such. It feels like a deep dark shaft to some sort of deep dark secret that somebody spent so much money and energy building such an elaborate, inefficient locked door.

I generally don’t get those types who are into swords, or weapons in general. People who stand around getting hard to gun magazines. The people I hope I’m just imagining... But I have to admit, I sort of understand medieval weapon nuts when I see a nihonto. Japanese swords are utterly exquisite. Trust feudal Japan to turn every aspect of killing into an art form. If I’ma be a killer, I’ma be a stylish killer!

So I’ve started with the prettiest part first. This happy little fellow is called the tsuba, and serves as both the hand guard and the main source of artistry and individuality among Japanese swords.

Zbrush’s new Live Boolean features made it a breeze to put it all together into one solid piece. And with just a little planning ahead and maintaining polygroups, once it’s all together you can mask by groups and then polish out the edges all at once without affecting the larger shapes.

The 2013 Tomb Raider reboot managed to be really fucking good by acting as a deconstruction of old Lara’s character. Deconstruction is usually abound when a franchise, trope, or genre has been utterly run into the ground or grown stale, so It speaks to how the next two games in the Tomb Raider reboot series, Rise and Shadow, have squandered Tomb Raider 2013′s potential that I feel like we’re already due a deconstructionist reboot (soft or otherwise) of new Lara. Shadow of The Tomb Raider pays lip service to that idea, pretending (for like five minutes) that Lara being weirdly unmindful of the consequences of her obsession was a deliberate character flaw that the writers are now going to make into her central ark, before forgetting the whole idea and letting said deliberate character flaw default back into an unintended lack of care for characterisation and tonal control. They’ve really fumbled the advantages provided by what I felt like was the best videogame reboot ever. I feel like the people in charge are afraid to let her ever finish “becoming” the Tomb Raider because if they do they suddenly won’t know how to market her any more, so she’s trapped in this nightmarish reality that reflects all the flaws she’s doomed to never recognise much less overcome.

I’m reminded of what Sophie McDougall cautioned, when writing ‘strong female protagonists’ of neglecting to make them anything more than simply ‘...strong’. Through a combination of a lack of emotional range and constant confusion as to how we as the audience are supposed to feel about Lara’s motivations at any given second, Lara ends up feeling kind of... non-sentient. She has no agency in her own story anymore because her trademark, unwavering determination exists seemingly isolated from surrounding story elements. That’s why detractors of the series question new Lara’s mental health all the damn time, because of this weird disconnect between what Lara does and what’s happening around her. And it’s why Shadow of the Tomb Raider felt like it was onto something early on when the story calls her out for being selfish and obsessed. Of course, for 22 years Lara has been selfish and obsessed, but now she’s those things within a new series that ostensibly takes its own themes more seriously. But I feel like there are commitment issues. Like the devs are in an ill-advised balancing act between making Lara more than raw videogame id, and... not doing that.

Marvel’s Insomniac’s Sony’s Spider-Man Not The Movie The Game 2018 is out, y’all!

S’alright, innit?! Not quite up to the standards Sony exclusives have been setting this generation – it’s a bit heavy on exactly the kind of inconsequential sandbox side-questing I’d prayed we’d moved past in a post Witcher 3 world, and blame myself all I want, but me and the combat just do not like eachother – but it’s still easily the best Spider-Man game ever and then some if only for the story.

And I wanna talk about a specific aspect of its narrative - its characterisation of long time Spidey love interest Mary-Jane Watson and why it’s the best ever.

MJ is today thought of as integral to the Spider-Man mythos. Peter Parker’s one true love whom he is ultimately destined to be with (no matter how much your personal Spider-fantasy insists that Black Cat is the obviously correct decision). So if she is such an important character… then why does she suck?!

That is perhaps a product of her publication history.

Strange as it might sound to our modern ears, Mary-Jane was not conceived by Spider-Man’s writers to be anybody important. Way back in the original Amazing Spider-Man comics, MJ was a background accessory. A running gag who Aunt May would repeatedly try and set Peter up on a date with. Mary-Jane’s did even make an appearance until much later, she was just talked about.

It wasn’t until years later, in 1987, that writers very hurridly reintroduced MJ to the Spider-Man story, got her and Peter married, more-or-less as an attention-getting publicity stunt.

And it turns out weddings, however fictional, aren’t the sort of thing one should rush into.

Spider-Man’s married life as Peter Parker immediately became less interesting to read and write about and Marvel writers would spend the next twenty years trying to undo it without breaking continuity and without putting the blemish of a divorce on Peter Parker’s permanent record.

The solution they ultimately went with? Spider-Man made a bargain with Satan.

Oh, I’m not joking.

But by then an entire generation who would eventually go on to create the next batch of Spider-Man cartoons, comics, films, videogames, etc. had already grown up reading of a Spider-Man whose one true love had to be Mary-Jane Watson, so it appears she’s here to stay as a permanent fixture within Spidey lore.

What’s so wrong with that, you ask? Well, Mary-Jane is boring. Think of the Sam Raimi films, or the classic 90’s cartoon. Peter only definitively ‘get’s the girl’ right at the end, because the writers know that Mary-Jane’s role as a perpetual object of Peter’s desire doesn’t give them much to work with once she’s been obtained. And now knowing the editorial history of her character, it’s pretty obvious why – She’s not a character. She’s a background decoration of the world in which Peter Parker lives who was hastily turned into Spidey’s ultimate romance. And writers were in no hurry to have her grow into a more substantial character in her new role because they regretted getting them hitched almost immediately.

Insomniac’s new Spider-Man game too is of the opinion that there can surely be no Spider-Man without a Mary-Jane to long for. So how do they handle the character?

Well, I initially thought the timeline of this new take on Spider-Man was an odd choice. Normally you’d expect if they’re starting a new videogame series based on a pre-existing superhero they’d either start at the beginning (either with the origin story, or within the infancy of their superhero career) or at the far other end, the big sell being “This is veteran crime fighter Spidey’s toughest challenge yet!”

But Spider-Man The Game No Not That One, Or That One, Or That One starts somewhere in the middle. Peter’s been Spider-Man for eight years -hardly a veteran, but still with an established presence in New York, a rich rogue’s gallery, and with certain characters already having grown past their classic roles, eg. J. Jonah Jameson has left his newspaper, The Daily Bugle, and become the Marvel world’s answer to Alex Jones. Decrying Spider-Man as a menace on the airwaves 24/7 and branding any evidence to the contrary as fake news.

But this starting point was actually a really interesting choice, because it gave Insomniac the opportunity to give characters rich historical relationships. Case in point – Mary-Jane is introduced to us already long aware of Peter Parker’s superhero persona, and already having had a relationship with him. At first I felt robbed of a dramatic coming out of the Spider-closet scene, but that would have been too easy.

This imagining of MJ is an investigative journalist working for the Daily Bugle (itself now a more sympathetic outlet without J.J. Jameson).

Stop and think about that for a second. How do you make interesting the boring, long time love interest who exists for no other reason then to be a love interest? Give them a career! Why’d it take us so long to think of that?!

Yeah, yeah. I know. She already had a career. She was a supermodel. And how little of that did we actually hear about? How relevant was that to the events of the story? That wasn’t a role for her to play, it was just an extension of her desirability.

The firsts time we see new MJ, she’s sneaking around, sporting an impressive looking camera lens - imagery historically associated more with Peter.

What’s this? A shared interested? Just like that, these two already have more chemistry than in any previous incarnation.

This version of MJ is introduced to us crossing paths with Spider-Man investigating some villainy, but it’s because she’s pursuing a story for her paper independent of his actions. They’ve been broken up for six months.

She’s happy to see him sure, and you know they’re going to hook up again by the time the credits role, but when we meet her for the first time, she’s doing her own thing. Being a hero in her own way.

She has an involvement in the plot important enough for us to care about without being directly connected to Peter Parker. Which is good because if we’re to care about their romance, we need to care about them both as individuals first.

A journalist is a terrific choice for her, because she lives in New York City, which means her escapades are naturally likely to overlap with Spidey’s, and when they do she has a role to play in the adventuring. And that’s exactly what happens. The player is given direct control of MJ for recurring stealth segments (which in defiance of most obligatory stealth segments are not even that painful) wherein she sneaks around and uncovers secrets relevant to Spider-Man’s ongoing battle with the today’s villain, Mr. Negative. The two correspond to help each other, and as they become part of each other’s lives again, MJ puts a lot of weighty emphasis on wanting to be his “partner” as opposed to a “sidekick”.

The story Insomniac crafted is very satisfying (if a little bloated near the end, with its villain switcheroo). Rather than going the Batman Arkham route of giving us just a taste of every Batman character under the sun, they focused the narrative on a smaller cast but explored more fleshed out relationships, and the standout in my opinion is realising how little there was to Mary-Jane’s character considering the gravity often afforded to her presence in the Spider-Man universe. And so they basically had grounds to constructed an entirely new love interested for Spider-Man and call her MJ, and they did so by making her a character first, love interest second.

There’s been so much Spider-Man stuff. Too much. The media landscape is saturated with Spider-Man. Every now and then something will stick and be properly absorbed into the collective consideration for how we think about Spider-Man lore. Usually because they add something to it, while dreck like the Amazing Spider-Man films, or most of the videogames from the last decade are swiftly forgotten.

It’s very frustrating when a game comes out that’s mostly great but with a few very easily fixed problems, and you feel like “If I had been in that design meeting, I could have said two words that would have immeasurably improved this game.”

Case in point: Spider-Man. from what I hear, the combat either clicks for you, or it really doesn’t. I’m in the latter camp unfortunately, and I can explain exactly why.

Oh sure, there are little things, like how the Spider-sense graphical indicator of an incoming attack should clearly be bigger and more prominent, and how the indicator showing the path of incoming gunfire looks too much like a web line which makes the automatic reaction part of my brain think “But web is good, isn’t it?”.

But there’s one critical decision in this combat system’s design that is a) more potentially detrimental, and b) more interesting to talk about...

*ehem* In the Batman Arkham games (oh shut up!), notice how when an enemy attack is incoming - and the second you have to react is indicated by the Spider-sense looking graphic - as soon as you push the button to either counter or dodge, the enemy’s attack animation is canceled and the counter/dodge animation triggers immediately. This is a small but very impactful and deliberate design choice. It means that while your task is mostly to react, you are able to control the flow of the fight more the better your reaction time. It does mean that Batman tends to snap between animations unnaturally, but it’s an example of responsiveness and ease of use trumping realism.

Now, in Spider-Man 2018: The Sony-ing, when an enemy begins their attack animation, and you’re given a second to respond indicated by the [now very literal] Spider-sense graphic, your pressing the dodge button in time doesn’t fast-track the dodge animation. The dodge will come when it comes and you just have to trust that you pressed the button in time and then wait for the would-be attack to connect. That immediate feedback to let you know that the dodge was (or in this case, will be) successful is gone.

It’s quite possible that Insomniac wanted to take their combat animation to the next level and didn’t want to resort to that ugly animation snapping. A worthy goal. I totally get it. (The upcoming DMC5 is doing something similar). But they could easily have had the best of both worlds. How about this - how about when you hit the dodge button in time, it doesn’t impact the animations or the combat timing, but the little Spider-sense graphic over your head changes - (eg. flashes with a little, unintrusive sound) - in reaction to the button press, just to let you know “Well done, your command has been received” while the dodge animation is still waiting for the right moment to trigger. This would restore that all important immediate feedback to the player with the added bonus making the dodge feel more weighty and satisfying.

Oh boy, it feels good to be working on this rig again! It’s been too long.

I’m especially happy with how elbows and knees hold volume as they bend due to the corrective joints that fire more as the limbs bend to push out against the usual collapsing. The knees could be a little neater. Next I’ll give the same treatment to the wrists and ankles. A little trickier since they rotate in multiple axes.

The two big problem areas are still the hips and shoulders. Lifting the arms above the head shows at least that the twisting is following the right directions, but I need corrective joints to hold volume. I don’t even know where to start! It’s going to take some witchcraft. The hips I think I know how to approach. The way the crotch sticks out when the legs raise is an issue I’ve observed in several videogames. It’s usually not a problem because legs don’t often raise that high. I think I’ll just give the groin a dedicated bone and average that bone’s position based on the rotation values of both hips.

Alright, ignore a lot of what I said in my previous post about rigging the wrist. Turns out that video I used as reference for wrist movement doesn’t know how to use a compass. The video claimed that movement is distributed between two points in the wrist bones and that anterior movement = 50° in joint A and 25° in joint B and that posterior movement = -25° in joint A and -50° in joint B. 75° of movement in each joint all up. But going this while keeping 0° in joint A equivalent to 0° in joint B also while keeping the entire range of movement at a constant speed turned out to be mathematically impossible. Irked, I measured the movement in that video myself and got very different results. I got 35° in joint A = 40° in joint B, and -35° in joint A = -40° in joint B. A simple geometric relationship, as opposed to ugly (not to mention impossible) quadratic one proposed by the measurements purported in the video.

The value of multiplication from joint A’s rotation to joint B’s isn’t a rational number, but I don’t need that kind of precision anyway, so I just rounded to the second decimal place. Joint B rotation = Joint A rotation multiplied by 1.14. Accurate and much much simpler.

And and since the ratio is now a constant, I can keep the locations of the corrective joints in the same place, smack between joints A and B.

The next challenge will be the big one: full range of shoulder movement while holding volume.

With just the first skinning pass, this is what I get with the dreaded hands-over-head pose:

There’s a lot of hubbub about Destiny, and its developer Bungie lately. I’m not interested in cataloging their various deceptions and plain old dick moves since the launch of the original Destiny, that’s been covered EXTENSIVELY already. But the accompanying questions: “Are Bungie in trouble?” ,“Is Bungie going to be bought out by Activision?”, “Has Bungie lost its mojo?”, and just generally “My god people, what’s happened to Bungie?”

I believe can answer this. But it will require some... philosophy.

Specifically Theseus’ Ship. A very old, still very relevant philosophical thought experiment. You’re probably familiar, and there are multiple retellings varying in unimportant details that don’t change the question posed by the story. This is the most succinct version: A ship leaves from a port in the town of Theseus. Its voyage will be long, spanning many years. As it sails, some of the ship’s components wear or get damaged and need to be replaced, be it a torn sail, a broken mast, or a rotted plank. As the ship visits various ports and islands, some of its crew decide to stay behind, and are replaced with new members. By the time the ship finishes its voyage and returns to Theseus, not a single component of its original structure nor a single member of its original crew remain. The entire ship has gradually been replaced, piece by piece.

The question posed by the story is this: Is the ship that returned to port the same ship as the one which left all those years ago?

There’s no definitive, correct answer. Maybe you feel like the ship should be considered technically a different vessel as soon as one of its pieces is swapped out, or maybe you think it retains its original identity so long at least one piece remains unchanged. Or maybe you feel like it’s a question of degrees, and that identities can be altered, stretched some way before being considered a properly disparate entity. Or maybe you feel like the ship’s identity should be considered changed when the amount of it composed of new pieces exceeds half.

The point is to make you consider what it is we mean when we speak of identity. When we say something is one thing and not another. It’s the same existential quandary we engage with when we talk about ourselves being different people than we were some years ago because so many of the cells composing our bodies have died and been replaced with newer ones.

It could be that you believe the ship is still the same ship that left Theseus as long as the pieces, new or old, are assembled in the same way. A question of how much of identity is a matter of that which can be physically measured and how much of it is perhaps historical, or even imaginary.

So what the hell is going on over at Bungie?

Bungie are one of the best and most storied game developers around. They achieved fame and glory with the Halo franchise, one of the most beloved and influential videogame series of our time. During Halo’s development back in the year 2000, Bungie was purchased by Microsoft, and in 2001 the first Halo game became a sensation that almost single-handedly put Microsoft’s new videogame console the Xbox on the map.

10 years and 5 highly successful Halo games later, the people working at Bungie, yearning for creative freedom, had secured their company’s independence in exchange for leaving the sole rights to the Halo IP with Microsoft.

The world was once again Bungie’s oyster, and after wandering around exploring different ideas for their big new IP, they began work in 2010 on what would eventually be Destiny... and the source of a whole lot of headache for Bungie.

Bungie had entered into a 10 year development contract with publisher Activision, and some of the bigger names to leave or be forced out felt that this relationship was becoming poisonous to the company spirit that had served Bungie so well. This deal was for a whopping $500 million, the largest development contract in videogame history, so valuable did it seem to secure Bungie’s next big series. Pete Parsons, Bungie’s COO told press they expected people would “...put the Destiny universe on the same shelf they put Lord of the Rings, Harry Potter, or Star Wars,” And this game which had started out with the intention on the part of some of Bungie’s heaviest hitters of getting as far away from Halo as possible in search of a fresh new experience, ended up being, by their own admission, a lot closer to Halo in genre and design then they had wanted.

When it launched in 2014 Destiny was rife with problems, some slight, some pretty glaring and certainly surprising of a game with Destiny’s pedigree. And while through iteration and expansion Bungie eventually managed to steer Destiny in a direction at least decent, and while Destiny 2 released this year and was far better received (at first...), It’s pretty clear that those dreams of creating a cultural touchstone to stand with the likes of Star Wars or Lord of the Rings, or even Bungie’s own Halo remain pretty illusive.

I’m still somewhat fascinated by the way the fans Destiny 2 earned for the franchise all now seem to be turning on the game and reporting that once you reach the end game content, Destiny 2 turns disappointingly back into Destiny 1, ie. a little bit shit.

And this is all to say nothing of the way Destiny (perhaps at the behest of Activision, we can’t be sure) has been treating its players with cloak & dagger design choices secretly aiming to sabotage any attempt of progressing through the game without turning to its in-game storefront and spending real world dollars. And whatever amendments they’ve attempted to make as they hop from one scandal to the next amounts to a cynically charged Peter Molyneux-esk cycle of 1) Enact predatory design decision/business practice. 2) If caught, apologise, fix problem(optional), vow to do better in future. 3) Repeat step 1. Immediately.

And news that Bungie is committed to making sure Destiny’s design going forward will revolve heavily around the Eververse (the in game storefront and source of greedy post-sale monetisation many are so upset about) comes out right as an organised online campaign, #RemoveEververse is firing up.

Destiny is nobody’s favourite franchise right now (which is a shame, because it was for a brief, glorious moment) and Bungie, of all developers, are nobody’s favourite people.

So what gives?

The details of what made Destiny’s four year development so turbulent have been covered pretty extensively by journalists, including an entire chapter in Jason Schreier’s excellent book Blood, Sweat, and Pixels. Suffice it to say that the Bungie that came out of Destiny’s development in 2014 looked a hell of a lot different than the Bungie that went in in 2010.

Jaime Griesemer - a major designer at Bungie for 13 years, Marcus Lehto - with Bungie also for 13 years and the creative director of Bungie’s last Halo game, Marcus Lehto - A lead designer at Bungie for 10 years, Adrian Perez - a programmer at Bungie for 9 years, Vic Deleon - a senior environment artist, Marty O’Donnell - composer and sound director at Bungie for 12 years, Harold Ryan - the motherfucking president of the company,

Over the course of development, all of these noteworthy figures at Bungie would either quit or be pushed out due to frustrations in Destiny’s development, not always on very good terms from the sound of it.

I’ll be frank but, I feel, realistic, in saying that while a studio’s artistic identity is a conglomerate of all of what each and every member brings to the table, some members’ input counts for a larger chunk of that identity than others. Some are more replaceable than others in terms of how much sway they have over the creative vision that the company will all work to bring to its audience, and what’s worrying was not that so many people left Bungie during Destiny’s development - people leave and join studios all the time - it’s that so many veterans, the kind of people who make up the majority of their studio’s identity left in such a compressed period.

Whether this is a sign that the project was mismanaged (it was), and that those who left should have something to feel betrayed over, or just a natural part of Bungie’s transition from one major chapter in its life as a Microsoft subsidiary to the next chapter as a free bird with new responsibilities and a new direction to chart, the point made by Theseus’ Ship is very relevant here. How much of the unique creative energy we know as “Bungie”, remains?

People have been calling the last couple of years “The Downfall of Bungie” or claiming that Destiny is the game that destroyed Bungie. Are they that wrong when considering that changing to a new identity does technically involve the destruction of the previous?

How many members, and which members of a group must leave and be replaced before that group can no longer be said to still be, even if the corporate entity known as “Bungie” is still well and truly operational?

“What’s happened to Bungie?”

Simple. It’s died. And was replaced by a new Bungie. It happens all the time (though usually more gradually). Change is inevitable. To try and resist this constant shifting of identities is foolish. After all, Bungie changed a lot to become a Halo factory we all loved for 10 years. The important question is what has Bungie changed into? We’re looking at a new Bungie. But is it a change for the worse? If 2017 is anything to judge by, I fear it very much is.

While the story of Destiny 1′s development from its initial release through to the Taken King expansion can be read as a developer releasing a... ‘not great’ game and working to make it better, Destiny 2′s post launch development so far is the story of a developer releasing a much better game and then doing everything they can to poison people’s experience with it.

Happy holidays and such.

I’ve been puling my hair out over this rig, specifically how the shoulders are behaving. See I’m determined for this rig to A) operate entirely on joints, and B) be able to lift its arms overhead without the shoulders collapsing like so many joint based rigs currently employed in vijyagames. But before I can even think about that, I’m having trouble getting bellow-head shoulder movements performing properly.

This is the base pose:

Pretty standard relaxed M pose. From here the clavicles should be able to move forward about 45°, back about 17°, up about 30°, and down about 16°. Don’t even need to think about twisting at this point, such as the clavicles even can.

The up and down motion is fine. The skin weights are by no means final, I plan to have a lot of helper joints around the trapezii and pectorals. But look at the forward movement...

Ew! I don’t like it! Get it out of here! As the shoulders move forward, they move outward and end up way too broad. I understand what’s happening, the pivot points for the shoulders were placed right at the front of the chest, where the clavicles attach to the sternum, since this is where the shoulders really do pivot from in reality. But as you can see bellow, this means the shoulders become broader at the clavicles reach their forward-most positions:

This would seem to be correct. Geometrically and skeletally the shoulders would indeed broaden slightly as the clavicles approach creating a combined 180° angle. So why does it look so obviously wrong? The arms clearly do not move out that far from the body! Looking at joint rigs extracted from various games, nobody seems to have this problem. But that’s because those rigs are far less anatomically accurate (My rig is meant to be more anatomical in bone placement, the mesh is still highly stylised) and the shoulder pivots are usually placed deep inside the shoulder rather than at the front, so forward backward movement doesn’t lead to broadening. A tiny but of narrowing if anything.

The reason for this is that rigs in games rarely if ever treat the scapula as separate from the clavicle. And why would they? The scapula’s movement only offsets significantly from the clavicle when the arm lifts above the head, which videogame characters rarely need to do. That’s why I’m in this mess!

But my rig is meant to be based more closely on anatomy, with the shoulder pivoting from the clavicle/sternum connection, joining onto the scapula which carries the arm nearby. But this shouldn’t matter! Real world shoulders don’t broaden as they rotate forward! Or so I thought... The problem, as it turns out, is in my base pose. This shoulder placement, the same placement used in thousands of 3d characters’ resting poses, is actually impossible. Don’t believe me? Go stand in front of a mirror. Take your shirt off. Turn side on. Your resting posture probably had your shoulders and arms in the middle of your torso mass, equal distance from your front as from your back. Like this, you shoulders are as broad as in my model’s resting pose, but notice in the birdseye image above that my model’s shoulders are pulled back to align with the back. So try that. Pull you shoulder back so they line up with your back. Holding this pose, turn to face the mirror and notice how much narrower your shoulders are now.

Boy it’s been a while since I’ve had time to work on my joint-driven deformation rig. When I left it last I was pretty unhappy with how the wrist was working, so I overhauled it after studying the way the wrist bones actually operate.

This is how I laid out the joints...

So first thing’s first. The wrist can’t actually twist. Twisting happens in the forearm by the radius and ulna circling eachother. The wrist can only rotate up and down and side to side. So to make things simple for the animator and keep all these movements tied to one control for all of the wrist’s motion, I encapsulated the wrist joint inside another joint occupying the exact same location and orientation. The axis of the hand control that controls twisting is actually tied to this WristHolder joint, carrying the entire wrist and hand joint hierarchy with it (and of course, the twist distribution joints along the forearm are automated to make sure the twist deformation of the forearm is smooth). The remaining two axes are tied to the actual Wrist joint, inside WristHolder, they control the bending up and down and left and right. I set the rotation order so the middle axis is the one controlling side-to-side, because if you check out your own hand, the wrist has a very limited range of movement in this axis, so there is no risk of passing that dreaded 90 degree threshold and triggering Gimble Lock. This would have been a huge problem if the wrist was rotating in all three axes, but remember, it’s only rotating in two - the twist axis is happening one level up, to WristHolder.

Next. If you take a look at this video, you’ll see that the bending of the wrist is actually distributed over two points of articulation throughout that daunting cluster of bones between the forearm and metacarpals.

And what’s more, they don’t rotate equally. Posterior rotation has the joint A (the one closer to the forearm) doing 75% of the work, while anterior rotation has joint B (closer to the hand) doing 75% because joint A runs up against the radius. I’ll try and follow this more strictly in a future rig, but for now I simplified. Behold.

I plugged the up-down rotation axis of the hand control into a condition node, and told the node that if the received value was negative, to use that value to drive wrist joint A, but if the value was positive, to use that value to drive wrist joint B. So rather than a 0.66%-0.33% joint effort (hyuk! hyuk! hyuk! hyuk! hyuk!) one joint takes over completely depending on whether the wrist is moving in the anterior or posterior direction.

Previously I had done this all with just one wrist joint and while posterior rotation looked fine, anterior rotation was clearly not right. Because in reality, due to this two joint bending distribution, the pivot point of the wrist moves around all throughout the up-down movement.

But there was one last problem. Without using any corrective blendshapes (which I’m determined no to do for this rig) there’s some terrible collapsing in the shape of the wrist. In joint based rigs under real-time rendering constraints the rigger usually has to do the best they can with skin weighting, usually resulting in some of the hand’s surface penetrating the upper wrist. The silhouette is preserved this way, but the texture looks like its eating itself.

So I made some corrective joints. For a long time, processor limitations meant real-time productions didn’t have the joint count to spare on all the little corrective joints needed to avoid typical collapsing during character movement. But that’s less of a problem now. Now that kind of calculation is peanuts!

The light grey joints in the wrist a couple of images ago are colour-coded so because a) they’re not skin weighted, they exist purely to drive other joints that are skin weighted, and b) these are the corrective joint Holders. Why the need for holder joints? Why not parent the corrective joints directly to the wrist rotation joints? Well see here!

I take the up-down rotation axis of a given wrist joint (A or B), plug it into a MultiplyDivide node, multiply it by -0.5 and plug the result into the corrective joint Holder contained within that wrist joint. This means that the Holder’s rotation will move to cancel out the parent joints’s rotation, but only by half way. The result, as seen in the image above, is that the holder joint’s orientation will always be smack between that of the hand and the neutral position (neatly visualised by the direction of the forearm). And since the actual corrective joints are parented to and being carried by that holder joint, it now only has to worry about translating in one direction. The dynamic rotation of the Holder joint will ensure that the corrective joint’s relevant axis is always pointed in the right direction.

Now all we need to do is tie the rotation value of the wrist joints to the translation value of the corrective joints. Don’t worry about how far out they move, you can compensate by how strongly or weakly you skin weight the relevant surface to them.

But there’s one last problem. One of the directions in the up-down movement dips into negative values, remember? Positive rotation value = anterior movement while negative rotation value = posterior movement. And it’s that rotation value that is used to drive the translation of the respective corrective joint. What do we do to take into account that that value can be positive or negative? You have two options. One is to just think ahead and make sure that the negatively charged direction’s corrective joint is oriented in such a way that the resulting negative translation still gives you translation in the outward direction you want, or, if you didn’t think of that, you can do what I did: Before plugging in the wrist’s rotation value, run it through some nodes to return the absolute value. “Dafaq is an absolute value?” I would have asked a year ago, before Khan Academy fixed me up. An absolute value is a simple scalar, meaning it measures magnitude but not direction. Translation? Absolute values disregard questions of positive or negative. All absolute values are treated as positive. The absolute value of 6 is 6. The absolute value of -6 is also 6. So we can use this to mitigate the possibility of a negative value when we really need a positive one. We don’t just want to use a reverse node, because while that would turn a negative value positive, it would also turn a positive value negative, and we’d be right back where we started. Absolute values solve our problem, but Maya doesn’t come with a node to return absolute values :( We’ll just make our own! We can do it with just two nodes. Two MultiplyDivide nodes. Plug the troublesome rotation value into the first MultiplyDivide node. Set it to ‘Power’, and set the amount to 2, so we’re squaring our rotation value, and a square is always positive, because a positive number times a positive number gives us a positive number, and negative number times a negative number also gives us a positive number. We take the result, plug that into the second Multiply divide node, also set to ‘Power’ and set the amount to 0.5. A number to the power of 0.5 gives us the square root of that number. So we’ve essentially undone the effect of our first MultiplyDivide node but this time starting from a guaranteed positive value, so the square root will also end up positive. So with just two nodes we’ve made our own function that returns whatever value we plug in, but if it’s negative it will be turned positive, and if it’s positive it will come back unchanged.

Plug that result into the outward facing translation axis of the corrective joint, skin weight the necessary surface to it, and now as the wrist rotates, the corrective joint pushes outward to counter the collapsing effect. It’s the same process for the side-to-side rotating as well.

I remember after seeing Batman v. Superman predicting a similar fate would befall it; that it would be doomed to be appreciated for the least flattering reason - as a film that so spectacularly and thoroughly demonstrates how not to tell a story that it would become an invaluable tool for film students for decades, doubly so since the question of how to and how not to set up an interconnected film continuity is becoming more important.

And it seems one of my favourite internet personalities and film critics, Moviebob, has had the same idea.

His usual program is ‘Really That Good’, where he takes groundbreaking films of the past and helps an audience who wasn’t there in the moment of their phenomenon to understand why they were and still deserve to be a big deal. This time mutated into a one time ‘Really That Bad’ to take a look at why all the many... MANY things about Dawn of Justice that don’t work. That there are so many geek personalities online with their own takes on what BvS’s fundamental problem is is not what’s telling. What’s telling is that they all appear to be right.

I like rigging. I’ve gotten better at it since I’ve been learning all the maths I maliciously forewent in high school. The goal of this rig is to stick to game compatible joint systems but to try and sort out some crappy deformations I always notice in games.

All four limbs can be blended between IK and FK control depending on the animator’s needs in a given motion. Unlike previous rigs where I used a similar IK/FK blending system on the spine, I just have the IK spine controls constrained to the FK spine controls, with the IK controls controlling the offset. That way the animator can still choose their preferred method, but this saves on a whole mess of nodes needed when using the blending method.

I’ve wanted to include metacarpal joints for a while, it lets you get way more natural looking hand poses. I could sit and play with hand controls all day.

This one really is overkill, but I wanted to try it out anyway. The jaw’s motion is more complex than people think. The hinge actually slides forward as the mouth opens, and it was a simple thing to tie the jaw bone’s rotations between two specified angles to a translation of the parent bone using a remapValue node.

Okay this is a big one for me: I’m so sick of seeing characters’ shoulders collapse in on themselves like dying stars whenever they lift their arms over their heads. To try and combat this, I’m bringing the scapula into consideration. In most joint rigs, the clavicle connects the chest to the arms, but in reality the clavicle connects to the scapula which connects close by to the arm. The extra bone assists with twist distribution.

But it’s tricky to get the scapula’s movement fully automated without using a whole web of set driven keys which opens the shoulders up to Euler flipping. So I included controls that can offset the scapula from their automated positions if needed. But this is hardly the end; I’ll need to make use of cone readers to drive corrective joints if I want to retain volume without muscle objects or blend shapes (though blend shapes are totally viable in high end games these days if used sparingly). I might have similar challenges around the hips and crotch, although I’ve seen in other rigs people getting pretty good deformation just with well placed skin weights and extra bones to average the orientation of the buttocks between the pelvis and thighs.

This might sound weird, but I’ve had a very genital last couple of days. I’ve been making some base meshes - one male one female - that I can wheel out whenever I’m working on a human character in future so I don’t always need to start from scratch. Good fingernail topology and all that, planning hopefully for any possible future need, including detailed genitals. The whole package.

I’ve been finishing the female mesh first, which involves staring at women’s crotches on the internet for longer than all other times in my life put together (honest). The topology down there is a nightmare. The human body is gross and wonderful! And at the same time, my mum and I (great way to start the next sentence) listened to a new podcast: Ladies, We Need To Talk with Yumi Stynes, whose first episode was about the lady bits and stupid taboos surrounding them. And on top of that, just browsing social media lead me to reading about the works of Megumi Igarashi (the vulva canoe lady) and her work pushing against the stooooopid censorship imbalance in Japan around female as opposed to male genitals.

So I am just all about female genitalia all of a sudden! - Sam Leheny, 2017 -

I wouldn’t have bothered posting about this were I not inspired by the cosmic forces clearly at play, so here is the vulva I’ve been modeling!