July 2023 Animation Challenge - RETROSPECT (Patreon)

Do you like reading?

In this post I'm just going to break down how I went about each animation in the challenge to provide more insight for those who are interested.  Now that I've done this a few times, I might have a different perspective or approach than what I had a few years ago.

DAY 1 - Mercy Riding Pharah

As usual, the first day is where I try to do something simple to ease myself into the challenge.  Since much of this challenge was to stress-test my Overwatch models, I decided to open with Pharah and Mercy.  Like I said in that post, I'm trying to get better at working with butts so this seemed like a good pose.

I really didn't encounter many technical hiccups here, so my difficulties centered around the butt physics.  How stylized do I want it, and more importantly how do I achieve that style?  Well once I got the core movement down I started by animating the butt purely with the bones in the rig and tried to keep the principle of squash and stretch in the front of my mind.  After that I refined the movement more, then added a few shapekeys that animated with the butt.  Refined the movement again, and then I added ripple with a plane.

Finally I was at a pretty decent spot.  However every animator gets to the point where they have to decide how far they want to push something.  With a timed project like this, you don't want to spend any time on something that isn't going to pay off.  I figured since this was just a daily animation I would risk it and try to stylize even more.  I hooked up a physics cage to Mercy's butt via mesh deform and after some trial and error I ended up with a pretty decent simulation on top of all the work I already did.

Once I went to render however, I started second guessing myself.  I feared that the motion of her butt on the downward thrust was way too much, so I went for an even riskier second render without the cage deform.  I optimized in any way I could, lowering the resolution, motion blur passes, etc.  With that render barely done in time, I comped just the top of the butt on the downward motion over-top of my original render and got it out on time.

Phew, that was a long explanation but it just encapsulates how much I can go back-and-forth on a single aspect of an animation.  All-in-all though, it's honestly one of my favorites.  It wasn't very challenging, but it was satisfying to figure out.  A lot of experimentation without much mindless work.  Plus, I don't do poses like that very often.

DAY 2 - Mercy Blowing Ana

I did a poll for this one, so a lot of the pose came together pretty naturally.  I just had to figure out where I wanted to put the hands and camera.  However later down the line I started to run into the drawbacks for how I adapted specific details of the pose.  Mainly, the angle of Ana's hips compared to Mercy's face.

You see, animations like this are one big magic trick.  When you're working with the proportions that I'm working with, doing poses as acrobatic as this, it's not as simple as moving the hips back and forth.  Meshes need to be bent, bones broken, ugly spots covered up by angles.  The set-up might determine the difference between 1 day's work and 1 week.

So once I was pretty far into the animation process it dawned on me that Ana was coming down in such a way that was convenient proportionally, but placed her lower abdomen/hip area on Mercy's forehead rather than her mouth.  This put her nose and hair right in the line of fire and made for some difficult maneuvering.  Now had I made the two more perpendicular by rotating Ana's hips left and Mercy's head right, I might have avoided a lot of this clipping.  Unfortunately that's hindsight for you.  No going back now.

Suffice to say, a lot of this animation was about managing clipping and trying to make it look half-way reasonable.

DAY 3 - Tracer Licking Widow

Usually around day 3 and 4 is where I get into the big challenges for the week.  Instead of just upping the runtime or character count, this time I wanted to get more technical.  I looked at a bunch of reference and found something that would be more interesting to render.

This animation was structured in a similar way to the one on day 4, but that one is more complex so I'll cover all of that stuff there.  For this one I'll take the time to go over the sound.

It was sort of a hidden goal of mine to do sound for each one of these in addition to everything else.  I always had a little bit of free time during renders near the end of each day, and I figured I could use that time to do all the audio work.  I've been doing this kind of thing for a while now, so I was confident a few hours was enough time... as long as it was for something simple.

This animation was not simple.  It required a lot of sounds that I don't have great assets for and a lot of things I haven't really tried yet.  When I approach the sound for an animation, there are a variety of places I get my assets from: sound libraries, voice actors, foley work, game audio, and ripped audio from videos online.  I've built up a pretty decent library over the years, but there's always something that I end up needing.

When that happens under tight time constraints, it can be an all-out frenzy.  What sound does a dick make when it's sliding/rolling across skin?  What sound does a tongue make on skin at that distance?  What if it sounds too much like the other thing?  Do I have anything of a woman breathing with her tongue out (onto a surface directly in front of her)?  Can I make this man doing the same thing sound like a woman?  Balls don't make a sound... but what if they did?  Is the noise floor for this breathing too high?

Pandora's box of questions.

There were no other outstanding problems but I did have to give up some jizz strands on Widow's body because the screen space reflections required to make them look right was fucking with Tracer's lips and I couldn't afford two separate render passes.  This unfortunately resulted in diminished readability for the overall post-orgasm situation on Widow's body, but hey that's the challenge for you.

Thankfully everything else came together pretty well for this animation.

DAY 4 - Mizora Kissing Shadowheart

The big one.

I knew I was going to do something like this part-way through the week and I had envisioned what it was going to look like pretty closely the night before.  Thanks to that, posing and lighting this one was fairly easy.  Animating it was not.

When it comes to animation, there are two ways to go about it: straight ahead and pose-to-pose.  Straight ahead means that you're just animating each frame roughly one after the other, with no significant structure in place.  Pose-to-pose usually consists of placing down keyframes at key spots in the animation, then filling the spaces in between later.

Now most animation you're going to see is pose-to-pose, but the straight ahead method has its advantages when trying to create fluid and unpredictable animation.  A lot of the kissing animations I do are just totally off-the-cuff, no reference, straight ahead animation.  Unfortunately that wasn't a great option here.  When you have to loop, it's better to have much more structure.  Pose to pose gives you structure.

Despite all that, I still didn't want to use any reference.  So similar to the last animation, I decided to lay things out using multiple 'mini-loops' inside of one big looped animation so that I could pose everything out without wasting a bunch of time.  Within this kissing animation there are essentially three main loops: the french kissing, the licking, and the sucking.  Altering a base of animation that's already there is much faster and loops better than trying to come up with front-to-back original animation.

I developed the loops first and then roughly blocked out the animation.  French kissing into licking into sucking into french kissing.  Now it wasn't quite as intimidating.  From there it was all about layering.  Layering in the spine, neck, head, tongue, hands, hair.  I wasn't too worried about clipping or how everything was lining up yet, I just wanted to get down the natural push and pull that you find in these kinds of poses first.

The thing about kissing animations is that they're just... a lot.  There's a perpetual feedback loop of one character pushing into the other character who's also pushing back.  If you look at something like a thrust or a punch, it can easily be broken down into affector and affectee, but with something like kissing it's all sort of interconnected.  Instead of refining the core movement first you just sort of have to feel it out and find the emotion and interest somewhere along the way.

Once I ended up with something I was okay with, it was time for the most dreadful part of any animation.  Cleaning up.  Un-clipping the lips, snubbing the noses, pushing around the tongues, moving the hair aside.  No fun.

I liked the animation though.  Aside from a compositing issue in editing that I caught like 3 minutes after I posted the thing, I really liked how it turned out.  Probably my most challenging animation challenge animation yet, technically speaking.

DAY 5 - Junker Queen Fucking Tracer

The poll for this one was pretty damn close, but the junker bros won out in the end.

There isn't too much to say about this one.  I got to try out a new slightly different high-key lighting style, Tracer's hips didn't totally break, Junker Queen was big enough for the pose.  One of the smoother full nelsons I've done.

So I might as well take this time to hyper fixate on something that vexed me in the end.

Motion blur is tricky.  Normally it's not an issue, but when you get really close to the action it can distort shapes and shift focus.  Junker Queen's balls were looking pretty good when I animated them, but after the render there was just something off about them.  The motion blur had essentially blurred out all but one frame in the loop, which was when they were compressed against Tracer's butt.

So why does this matter?  Well now it tricks the eye into thinking the overall shape is different than it normally is.  If you really focus on it, you can see through that and it's not a big deal, but for most people it just isn't going to read well.  Plus on top of that, the motion blur just wasn't interacting properly with the ripples, which affected readability there as well.

None of this is the end of the world, but it might just provide some insight into the kinds of things I fixate on sometimes.

Final Thoughts

I explained this in one of the other posts, but this year's challenge has made me realize just how complicated my animation process has gotten.  A lot of the changes on the outside are granular, but under the hood there's much more going on.  At the same time, it's not so complicated as to slow me down too much.  It's just... a lot to keep track of.

The models have really withstood the stress test.  With each battle my model base grows even stronger.  Now I may be one of the few people able to actually operate the damn things, but they really help me save a ton of time.

Finally, doing audio for each animation made me realize how much that free time I used to have during render helped.  I was particularly exhausted throughout this challenge.

Anyways thanks for reading!  I'd definitely recommend something like this to other animators out there.  Honestly I don't think it burns you out at all, as long as you don't do it that often.  It's the long projects that will get you.