Draw a Character Model Sheet That Another Artist Can Use
Draw a Character Model Sheet That Another Artist Can Use
You have one drawing of your character that works. The proportions feel right, the gesture reads, the face is appealing. Then you try a second view and the head grows, the shoulders drop, and the nose climbs the cheek. Nothing about the character changed between the two drawings. You just tried to draw it twice.
The short version of the fix: a model sheet is not a gallery of finished art. It is a set of measurements somebody else can draw from. Choose the few views your particular design actually needs, align meaningful landmarks across all of them, and build each new view by turning simple volumes in space rather than copying the outline you already like. Then treat expression as a deliberate change to a stable construction, and hand the sheet to one other artist to discover what it failed to say.
That last step is where most pitch-stage sheets get better, and it is usually the step people skip.
Choose the views your design needs
Start with the uncertainties, not the convention. Ask what another artist would have to guess if you handed over your single drawing. If your character has a pouch strapped across the body, the guess is where the strap crosses the back and which hip carries the weight. If the character has a long nose and a small skull, the guess is how far the nose projects when the head turns away. Those guesses become your view list.
For most pitch work that means a front, a three-quarter, and a side, plus one extra view only when a specific feature refuses to resolve in those three. A rear view earns its place by clarifying a strap, a tail, a back seam, a hairstyle, or a pair of shoulder blades that will matter in a pose. It does not earn its place because sheets are supposed to have one.
Two artists who have written about building turnarounds arrive at the same practical point from different directions, and neither is handing down a studio standard. A 2019 walkthrough from Thea Studio describes several views built with aligned guides and then compared in layers, and notes that a one-sided detail has to be redrawn rather than mirrored. A 2025 first-person tutorial from Carlianne Creates draws a related distinction: lining up heights is not the same as keeping volume consistent, and every new view should be checked against a drawing you already made. Their preferred number and order of views is their own habit, not a requirement — which is exactly why you should pick views by uncertainty instead of by count. Both posts are represented here by their text; the images and embedded video that accompany them were not examined, so nothing below depends on how they draw.
A pitch-stage sheet is also not a production asset specification. You are not delivering a full set of mouth charts and hand studies for a series that has not been green-lit. You are delivering enough spatial information that a storyboard artist, a cleanup artist, or a collaborator can draw your character in a pose you never supplied.
Establish landmarks that survive rotation
Pick levels and attachments that mean something in this design. Eye line, jaw hinge, the widest point of the skull, shoulder, pelvis, the top of the pouch, the base of the nose, the point where an ear meets the head — choose four or five, not twelve. A landmark you cannot locate confidently in every view is not a landmark; it is a guess with a line through it.
Draw them as light horizontal guides running across the whole sheet, so a level in the side view visibly matches the same level in the front view. Then keep the pose convention consistent: same stance, same weight distribution, same basic angle of the head, differing only in rotation. If the front view stands square and the side view leans back on one hip, your guides are comparing two different poses and will lie to you.
Here is the trap. A horizontal line across a sheet fixes a height. It does not fix depth. Your shoulders can sit at exactly the same height in a three-quarter view and a side view while the nearer shoulder is a hand's width closer to the viewer — which changes where the arm reads, how much of the chest is visible, and how the collar behaves. Height alignment is necessary and it is not sufficient.
So carry at least one depth measurement, and keep it simple. How far does the chest sit in front of the spine? How far does the nose tip project past the back of the skull? How deep is the head front-to-back compared to its height? A rough overhead sketch of the head and torso — what used to be called a plan view, and which nobody has to see in the final sheet — settles these questions faster than staring at two finished drawings and hoping they agree. When a new view disagrees with the others, the overhead sketch usually tells you which one drifted.
Turn the volumes, not a copied silhouette
Your character is made of forms arranged in space. For the worked example in this article, take an invented design: an egg-shaped head, a compact capsule of a torso, a long nose, and a pouch hanging on the character's left hip. Treat the following as a paper plan you can run yourself.
The head is an egg. The nose attaches to that egg at a specific point, and the tip of the nose sits a fixed distance out from that attachment. When the head rotates, the attachment point does not move on the skull. It just travels around the curve of the egg, and the tip follows it. In the front view the nose projects toward you and foreshortens to almost nothing. In the three-quarter view it projects partway. In the side view it projects fully. Same nose, same attachment, three different silhouettes, because the silhouette of a volume changes with the viewing angle even when the volume does not.
Now the classic beginner failure. You draw the side view and place the nose where it looks right on the page, which is usually somewhere above the eye line and floating a little. The nose attachment has risen. The character now has a nose that migrates up the face as the head turns, and any artist drawing your character in profile will inherit the mistake. The correction is not tolerance; it is repositioning the nose back onto its attachment point on the egg, and letting the tip land wherever that puts it.
The torso works the same way. A compact capsule has a front, a back, and two sides, and its waist and chest dimensions are fixed measurements, not a contour. When you draw the side view, the outline you draw is the front-to-back depth of the capsule. It will not match the front view's width, and it should not try. The front view tells you how wide the character is; the side view tells you how deep. Tracing the front contour into the side view produces a flat card with a character painted on it, which is why the resulting figure looks strangely thin and brittle from any angle but one.
The practical tool here is the overlay. Put the new view in its own layer, scale it so the eye line and pelvis land on the same guides as the previous drawing, and turn it on and off. Every place the outline grows or shrinks between views is a question. Sometimes the answer is correct — the pouch is genuinely hidden behind the hip in the side view — and sometimes the answer is drift, and your eye will catch the difference faster by flipping layers than by measuring. The 2019 walkthrough describes comparing drawings in layers this way, and the 2025 tutorial's insistence on checking every new view against one you already made points at the same habit, which is probably why the method keeps getting reinvented by people working alone.
Track asymmetric features and controlled expressions
Label left and right as the character's own, the way a stage direction does, and label them on the sheet itself. This sounds fussy until the first time you flip a drawing horizontally as a shortcut for a three-quarter view.
Do that with our invented character and the pouch, which hangs on the character's left hip, lands on the character's right hip. Worse, the strap now crosses the opposite shoulder, and that diagonal is one of the most visible things about the whole design. A mirrored pouch is not a small error; it changes which side of the character reads as weighted in every pose. One-sided features get redrawn, not mirrored — a point the 2019 walkthrough makes directly, and the kind of thing that only annoys you once you have already shipped the flip.
The rear view does its best work here. If the strap's path across the back is genuinely unclear after front, three-quarter, and side, that is your reason to add a fourth view — a specific unresolved question, not a quota. If the strap reads fine in three views, skip the rear view and spend the page on something the sheet cannot currently answer.
Expression is the other half of this. A broad smile deforms the mouth, pushes the cheeks up, narrows the eyes, and may lift the jaw slightly. It should not change the size of the skull, the height of the eye line, or the distance from the eye line to the chin. So draw the smile on a head that keeps the same construction guides visible underneath, and the reader learns something repeatable: this is what my character's face does, and this is what stays put while it happens.
The failure to avoid is subtle. If you draw the smiling head slightly smaller than the neutral head — a common slip, because a wide grin changes the proportions you are eyeballing — the reader cannot tell whether the character shrinks when happy or whether you drifted. One clear expression over a stable construction separates intended deformation from accidental proportion change far better than three expressions over no construction at all.
Check the sheet through a bounded handoff
You cannot test a model sheet by looking at it. You know what you meant; that is precisely the problem. The check is another artist, one pose, and an agreed brief.
It can be brief. Ask one person to draw the character in a three-quarter view looking back over one shoulder — a pose that appears nowhere on your sheet, and which forces them to reason about both the nose's attachment and the strap's path. Tell them to keep the construction lines if they work that way, to work at the same scale, and to mark anything they had to guess. Offer credit if the drawing might travel with the pitch, and get a clear yes before using it anywhere.
Then compare three things against your sheet: where their nose attachment landed, which hip carries the pouch, and which shoulder the strap crosses. If their answers match yours, the sheet communicated those. If any of them drifted, you have not found a bad artist — you have found the annotation or the missing view that your sheet needed. Fix the sheet, not their drawing. The urge to correct their version back into your version is strong, and it destroys the only information the test produced.
One handoff is one data point. It shows that a specific artist, with a specific brief, resolved a specific pose one way. It does not establish that your sheet is universally interpretable, and it does not prove that a different artist, a different brief, or a more ambitious pose would go as well. If a second person later reads the strap differently, that is not a contradiction; it is the same finding, twice, about the same ambiguity.
End with the construction, not the gallery
The shape of the finished submission is unglamorous on purpose. Keep the guide lines visible on at least one version of the sheet. Show the faulty pass where the nose attachment rose and the pouch flipped sides, and overlay the corrected construction so the repair is legible. That pairing does more work than six polished poses, because it tells a collaborator where the model is fragile and what you already checked.
The genuinely finished sheet is the compact one — front, three-quarter, side, a rear view only if a real question demanded it, one expression over a stable head, and a short note on anything still uncertain. What it hands over is not your drawing. It is the ability to make the next one without you in the room, which is the only thing a pitch ever needed from it.
Frequently asked questions
How should you choose which views a model sheet needs?
Choose by uncertainty, not convention or count. Start with what another artist would have to guess from your single drawing: how a strap crosses the back, which hip carries weight, how far a long nose projects when the head turns away. For most pitch work that means front, three-quarter, and side, plus one extra view only when a specific feature refuses to resolve. A rear view earns its place by clarifying a strap, tail, back seam, hairstyle, or shoulder blades, not because sheets are supposed to have one.
Why is aligning heights not enough, and what depth information should you carry?
A horizontal guide fixes height, not depth. Shoulders can sit at the same height in a three-quarter and side view while the nearer shoulder is closer to the viewer, changing where the arm reads, how much chest is visible, and how the collar behaves. Carry at least one simple depth measurement, such as how far the chest sits in front of the spine, how far the nose tip projects past the back of the skull, or the head's front-to-back depth compared to its height. A rough overhead sketch of the head and torso can settle these questions.
How do you keep a nose from migrating up the face when the head turns?
Treat the head as an egg and the nose as attached at a fixed point on that egg, with the tip sitting a fixed distance out from the attachment. When the head rotates, the attachment point does not move on the skull; it travels around the curve, and the tip follows. In the front view the nose projects toward you and foreshortens; in three-quarter it projects partway; in side it projects fully. Do not place the nose where it looks right on the page in the side view, or the attachment rises and inherits the mistake.
How should asymmetric features and expressions be handled?
Label left and right as the character's own and label them on the sheet. Do not flip a drawing horizontally as a shortcut for a three-quarter view: a pouch on the character's left hip lands on the right, and the strap crosses the opposite shoulder. One-sided features are redrawn, not mirrored. For expression, draw the change on a head that keeps the same construction guides visible. A broad smile deforms the mouth and cheeks but should not change skull size, eye-line height, or the distance from eye line to chin.
What does a bounded handoff test, and what does it not prove?
The check is another artist, one pose, and an agreed brief. Ask them to draw the character in a three-quarter view looking back over one shoulder—a pose not on the sheet—and mark anything they had to guess. Compare where their nose attachment landed, which hip carries the pouch, and which shoulder the strap crosses. If their answers match yours, the sheet communicated those. If they drifted, fix the sheet, not their drawing. One handoff is one data point; it does not prove the sheet is universally interpretable or that a different artist or more ambitious pose would go as well.