Interview Jonathan Colin: Readability, Silhouette and Gameplay Animation Secrets

This month on AniMotion, we’re diving into a fundamental topic for any demo reel: readability and silhouette.

How an intention can be understood at a single glance, how a pose guides the viewer’s eye, how a silhouette stays clear even when the gameplay kicks into high gear

To explore this theme, I invited Jonathan Colin, Lead Gameplay Animator, to share his perspective.

Jonathan has an unusually rich background: gameplay animation, martial arts, home‑made mocap, rigging, tech anim, camera work, production… And above all, an obsession we share, making action readable, even when everything moves fast

His interview is packed with concrete techniques, production insights, and ultra‑pragmatic reflections on silhouette, posing, timing, transitions, camera work, hit reactions… In short, everything you need to level up for this month’s challenge.

To begin, could you tell us about your background and what led you to the role of Lead Animator?

I studied Game Design and Art. I gradually developed a passion for 3D animation through the prototypes I created in school, as well as during internships and freelance work , notably at Cyanide on Styx: Master of Shadows and with Wolcen Studio.

Then in 2014, I had the opportunity to join Guerrilla on Horizon Zero Dawn. Coming from 3ds Max, I naturally shifted toward mocap editing in MotionBuilder , first for NPCs, then for the main character, Aloy.

Sometimes, once my animations were integrated into the engine, they appeared ‘broken’. So I started digging into the animation network in Morpheme to fix the issues myself.

That’s how I developed my expertise in gameplay animation and tech.

After four years in AAA and the early prototyping phase of Horizon Forbidden West, I moved back toward a more hands‑on, craft‑driven environment with smaller teams , first at Darewise on the MMO Life Beyond, then at Dream Powered Games on WarStride: Challenges.

It allowed me to refocus on keyframe animation, deepen my knowledge of Unreal Engine 4 and rigging, and gain much more autonomy across the pipeline.

I also discovered the managerial side of the job at Darewise, as I was responsible for mentoring an intern.

At first, I found it quite challenging , she completed her tasks much faster than expected, which pushed me to improve my organization and learn how to delegate more effectively.

Then in 2020, I joined Counterplay Games , first as a Senior, then as a Lead , working on Godfall and Armatus.

I was managing between four and eight animators, while also supporting the rigging team and the animation engineering team.

There was up to a nine‑hour time difference with the team in the United States.

During the day, I could focus on my own tasks, and then concentrate all communication in the evening.

With remote work, it became easy to record my screen so I could revisit feedback sessions or trainings later on

Since 2024, I’ve been working with Brain Jar Games.

We’re developing Dead as Disco, a beat’em up where every hit automatically lands on the pulse of the music.

I get to draw on both my martial arts background and my experience in music.

Between the technical challenges and the creative ones, I genuinely feel like I’m exactly where I’m meant to be.

When you build an animation, how do you make sure the intention is readable at first glance?

The most obvious thing is to get into an anticipation pose as early as possible.

To save a few frames, I make sure there’s already a delta from the idle on frame 0.I might keep one foot in common with the idle, but move the other, lower the hips or shift them to the side, tilt the torso, etc.

I often start my animation at frame -3 (which is 0.1 seconds at 30 FPS). I apply my idle on that frame, and at frame 0, I work on the offset pose

Once in Unreal, when I place the sequence inside a Montage, I set the animation to start at 0.1 seconds.

That way, the Montage begins directly on the ‘delta’ pose, which counteracts the blend (since blending can make things feel a bit too soft).

“This contrast in posing contributes to the responsiveness of the animation, which goes hand in hand with readability.”

Timing plays a crucial role in the readability of an animation. How do you work with contrast and rhythm to reinforce that clarity?

Whether it’s keyframe or mocap‑based, what I’m looking for above all in timing is contrast: slowing down the quiet moments (holding key poses) and speeding up the fast ones (moving from one pose to the next).

This is especially important in Dead as Disco, because the playback speed of the animations varies a lot to match the pulse of the music.

An animation with stronger pose‑to‑pose definition stays clearer when sped up (compared to one where the poses are ‘diluted’ in a more monotone timing)

Our PlayRate adjustment system is non‑linear.

If an attack is slowed down, we can apply more slowdown to the anticipation while keeping the strike itself fairly fast (to preserve the energy release).

We also use Hit Stops, which freeze the animation on the attack pose.This makes that key pose visible on screen for a little longer.But you have to use them sparingly, because a Hit Stop that’s too strong will kill the energy of the animation.

In fast‑paced gameplay, poses can transition very quickly. How do you keep the silhouettes clear during those transitions?

I mentioned holding the anticipation pose as much as possible, but player attacks still need to feel responsive.

So instead, I focus on emphasizing the ‘follow‑through’ pose.

If the character throws a punch, I keep the arm extended, with the body leaning into the front leg.

If they throw a kick, I keep the leg lifted.

Even though in martial arts you do train the way you place your leg back down (with control, not by collapsing into it ) in reality you wouldn’t actually keep your leg in the air or your arm fully extended.

You’d naturally want to return to guard quickly so you don’t leave yourself exposed.

But for the sake of readability in Dead as Disco, we intentionally break that principle.

Our character holds long ‘post‑hit’ poses; they don’t return to idle right away (though the player can move again without waiting for the animation to finish).

Our combo system selects the next attack based on the current pose ( whether the left or right foot is forward, whether the torso is hunched or upright, whether the horizontal energy is rotating clockwise or counter‑clockwise ) using a tag‑based system.

The ‘post‑hit’ pose of attack 1 becomes the anticipation pose for attack 2.

When I try to understand what affects readability , especially in fast‑paced gameplay , I capture a video and review it frame by frame. It’s often an accumulation of small visual inaccuracies that ends up affecting the whole: a character slightly mis‑oriented on a single frame, a blend that lasts a bit too long…

We’re also working on an IK system to ensure the attacker’s fists and feet connect as cleanly as possible with the targeted area.

The technical side is essential to adapt the animation in real time as effectively as possible

Readability also depends on how the enemies react. How do you build hit reactions that reinforce the player’s understanding of the action?

To respond to the player’s basic attacks, enemies have a library of hit reactions.

These reactions contribute enormously to the readability of the attacks.

After all, if an action is too fast to really see, showing its consequence is a valid solution.

For hit reactions, we take into account the targeted body part, the direction (for a ‘free’ area like the head, we have up, left, right, diagonals, and center; but for a more constrained area like the legs, which are anchored to the ground, it’s only inward or outward), and the intensity (‘Flinch’ as an additive reaction with no displacement, ‘Stagger’ as a full‑body reaction with RootMotion, Knockback with a fall to the ground, etc.).

We also have reactions for the surrounding characters.If the player performs a particularly violent takedown on one of their buddies, the nearby enemies will cover their eyes, or bounce in place (in rhythm) as a sign of amusement.

You have a solid background in martial arts. How do you use it to create readable poses and strong silhouettes, even when reality would be less clear?

It helps a lot for motion capture, which I record in my backyard.

I try to exaggerate my postures and the kinetic chain ( how the energy travels from the grounding in the feet up through the hips as they ‘screw’ into the motion, then rises through the upper body and into the arms; the ‘wave’.)

I even record counters, where the player and the enemy are animated as a pair.I first perform the attacker’s role, then the defender’s.

Having practiced these techniques with a partner , both applying them and receiving them , is just as important as performing them, especially when you need to reenact both roles for mocap.

"However, you have to be careful not to reproduce martial‑arts training literally.

In self‑defense especially, everything is designed for efficiency and surprise.

Strikes are direct and untelegraphed (no anticipation).

If someone grabs you, you immediately try to ‘guard’ the opponent’s hand , meaning you pin it against your body.

That lets you work with your whole body weight instead of relying on arm strength.

But this creates major readability issues in animation."

So we cheat on the poses and create more distance between the characters, even if that makes the techniques less accurate, or we exaggerate the timing so we can quickly transition into a clear striking moment (atemi).

It’s not always easy to make the animation readable from every angle.

The camera has a huge influence on the silhouette and on how an action is understood. How do you factor it into your work?

I tend to prioritize the most likely viewing angle.

For the hack’n slash Wolcen: Lords of Mayhem, the camera was top‑down and distant.For Godfall, I worked over the player’s shoulder, and for enemies, mostly from the front.

There are also cases where the camera is animated specifically to make an action more cinematic, like during a takedown.These actions are short so they don’t take control away from the player for too long.

To avoid disorienting the player, we try to keep the camera on the same side of the action from start to finish ( if it begins in the “South‑West”, we avoid orbiting all the way to the “North‑East”).

In Dead as Disco, we actually prepare three or four camera animations, one for each cardinal diagonal (NW, NE, SW, SE).When the takedown starts, we pick the animated sequence that’s closest to the camera’s current position.

We also prefer long takes over cut shots, and we avoid moving the camera too quickly.

“Because beyond the readability of the animations themselves, there’s the readability of the gameplay as a whole.

The player needs to maintain an understanding of the combat situation (are there enemies behind them?) so they can plan their next actions effectively.”

On Armatus, we had the option to define a delay after which the player was allowed to rotate the camera (the camera animation would keep playing, but the player could orbit around their character on the yaw axis).

On Dead as Disco, when we want to highlight an action without taking control of the camera, we simply use camera modifiers.They let us adjust the distance or height relative to the player, or tweak the FOV.

Most of the time, the camera can rotate 360° around the character. So I constantly rotate the view in my 3D scene to keep a fresh eye on my work.

The opposite would be animating everything from the same viewpoint and, once you think you’re done, rotating the camera only to realize the pose looks really strange from another angle.

Beyond the animation itself, which external elements (VFX, UI, art direction…) do you think help reinforce overall readability?

We rely heavily on VFX to finalize the animations and make the impacts feel more powerful.

For attacks with multiple hits, the sound design literally makes the animation ‘sing’, because our system ensures that each strike lands on specific musical beats (for example: quarter note, two eighth notes, quarter note).

You can really feel the rhythm of the attack

The game’s interface also helps the player a lot, but if it’s too busy, it ends up blocking the animation.

For example, in Dead as Disco, enemies have indicators when they attack.

By default, these indicators are placed at a fixed offset above the capsule. But in some attacks, enemies jump. To prevent the upper part of their body from being hidden behind the indicator, we use an animation curve in Unreal to shift the indicator’s height during those attacks, with a very smooth interpolation, so the player can follow it without any jitter.

However, the UI should never replace clear animations.Our combat designer often tests the game with the indicators turned off to verify the readability of the animations.

Art direction also plays a major role in readability.

On Dead as Disco, we’re lucky to have characters with very clear, easily ‘decodable’ designs , whether it’s the hero, the enemies, or the bosses. The lines are simple and give a strong rhythm to the silhouettes, and each character has a limited color palette with strong contrasts and contour effects

To wrap up: what advice would you give animators to keep a fresh eye on their poses and improve the readability of their animations day‑to‑day?

One of the biggest challenges in our work is keeping a fresh (and self‑critical) eye on our animation.

Very often, we don’t refine the animation until it’s correct , we refine it until our eye stops noticing the flaws.

“There’s also a health aspect to it.

We spend hours watching the same movement over and over again.”

So I’d like to broaden the topic to the readability of the 3D scene itself, and to visual comfort for us as animators.

Here are a few tips.

-First, work on an entire ‘anim set’ rather than on a single asset.For locomotion, for example, I create all directional starts at the same time (mocap selection, timing, applying the desired speed in m/s, root‑bone animation, applying the idle and cycle poses, polish), instead of finishing one and then moving on to the next.

With MotionBuilder, it’s easy to manage many takes within a single scene.

-You can also do a quick first pass, put it in‑game, and let it ‘rest’.

That way, designers can give feedback before you move on to a second pass.

-In traditional painting, one trick is simply to step back from the canvas so you can see it smaller and evaluate the whole more clearly.

In animation, I just zoom the camera out.

When I’m polishing finger contacts, for example, if I zoom in too much, I’ll naturally over‑detail things and spend too much time on them , even though three‑quarters of those details will be invisible once the animation is in‑game.

By zooming out, I avoid unnecessary work, and the time I save there can be reinvested elsewhere.

-Similarly, there’s the squinting technique , or in 3D, switching the viewport to a flat‑shaded mode, which helps you check the silhouette.

-To avoid visual fatigue, I try to keep my setup as clean as possible.

My viewport takes up most of my main screen.

My timeline spans the entire width.Most tools sit on the second screen.

I enlarge the gizmo so it hides less of the model and lets me make bigger, more relaxed movements instead of tiny, tense mouse adjustments.

The picker is an essential tool.

Personally, I hate seeing the rig. Of course, I bring it back when I need to select a specific object, but the less visual clutter on my model, the better.

I honestly don’t understand how rigs with huge circular controllers surrounding the whole model ever became popular , especially when they don’t correspond to anything anatomical.

At least with HumanIK, the controllers follow the skeleton’s structure. They make anatomical sense. Cascadeur works the same way.

-Finally, to avoid visual fatigue and let my eye focus on the movement itself, I work with the most neutral model and lighting possible.

Dark grey background, a ground checker (with 50 cm or 1 m squares), and character textures turned off.

I also avoid hair cards. Instead, I add a simple geometric shape behind the head so I can clearly read its orientation without being misled by hair flow or a hood. Head direction is crucial for conveying a character’s intention or the effort they’re putting into a movement.

Some people prefer having maximum visual fidelity in their DCC , or even animating directly in‑engine.

Personally, I couldn’t work with overly realistic lighting, post‑process effects, or full textures. It’s just too exhausting over time.

On the other hand, I love that moment when I import the animation into the engine, with the final character and all the real‑time deformations (twist bones, cloth physics, etc.).

Once again, I ‘rediscover’ my animation.

Some flaws I had stopped noticing suddenly jump out at me , or, on the contrary, an animation that felt too exaggerated in the DCC might turn out to be perfectly appropriate once it’s in‑game.

" In any case, checking your animation in its final context is essential."

A huge thank‑you to Jonathan for the time he dedicated to this interview, and for the precision with which he shares his experience.

His journey , spanning AAA, indie studios, rigging, tech anim, martial arts, and pure gameplay , gives him a rare perspective on what makes an animation truly readable in‑game.

What emerges from our conversation is the vision of an animator who has learned to make intention, technical constraints, rhythm, camera, enemy reactions, and production realities all work together, without ever losing sight of the essential: guiding the player’s eye.

For him, the silhouette isn’t just an outline , it’s a tool for understanding, a way to make every action instantly decodable, even in the fastest contexts.

His approach is deeply grounded in the real: bold pose contrasts, timing treated as a language, transitions that become anchors, a camera that serves readability, and constant attention to visual comfort , both for the animator and for the player.

A beautiful lesson in how readability is built: not by adding flourishes, but by orchestrating every detail so the silhouette already tells the story.

To learn more about his background or follow his work:

Website : https://www.jonathan-colin.com/

ArtStation : https://www.artstation.com/joco_anim

LinkedIn : https://www.linkedin.com/in/jonathan-colin-52907532/