Fold a Supplied Packaging Dieline Into a Product Mockup for a Treatment

The most convincing wrong mockup is a box.

Someone reads the net, counts four panels, builds a clean rectangular carton in a modeler, drops the artwork on by eye, and renders it at a handsome three-quarter angle. Front panel correct. Sides plausible. The back panel is upside down, the bottom flaps would occupy the same plane, and the stripe that should run around the front-right corner stops dead at it.

None of that shows in the hero view, which is exactly why the hero view is the wrong place to stop. Your job is not to build a box. It is to make this net fold, then walk around the result and look at it.

That is a small, finishable job. Keep the supplied net as the structural source. Name the lines. Choose a panel to fold around. Declare a thickness. Fold, inspect every face, and keep the editable structure separate from the picture you show. What you produce is a provisional object for checking panel relationships and artwork orientation — not a newly engineered package, and not evidence that the approved net can be manufactured.

Read the drawing before you interpret it

A dieline, a net, a blank: different names for the flat development of a folded object. Whatever it is called in the file you were handed, treat it as a record of decisions someone else made, not as a suggestion you are free to improve.

Four things to settle first, and all four belong in a message back to whoever owns the file:

• Identity. Is this the version the team approved, or a near neighbour? File names lie. Ask for the version and the approval state in writing, because a mockup built from the wrong revision is a very expensive kind of confident.
• Units. Millimetres or inches, and whether the drawing is at 1:1. A net drawn in inches and read as millimetres produces a mockup the size of a box of matches, and someone will notice only when they try to photograph it.
• An untouched copy. Duplicate the original before you touch anything and never overwrite it. Everything you do happens on a working copy, so that when a question comes up you can answer it by looking at what was actually supplied.
• The intended package. Four panels and a tab could be a carton for a phone case or a carton for a pair of socks. The physical object the artwork was designed for is not inferable from the geometry alone.

Then start naming lines. Cut lines and crease lines are the whole grammar of a net, and the trouble is that you often cannot tell them apart from the drawing's decoration: some producers use dashed lines for folds and solid for cuts, some reverse it, some use colour, some use a layer, and some deliver a flattened PDF where the convention has been thrown away. Colour is a hint, never a fact. If you cannot establish a given line's role from the supplied information, that is a question for the source owner, not a coin toss. Guessing here is how a builder ends up cutting through the only thing holding two panels together.

A fixture you can actually fold

Everything below uses one deliberately plain carton. It is fictional, invented for this exercise, and you should make your own version rather than modifying anything real.

The net is 200 mm around, 90 mm tall, open at the top, and printed on one side. Reading left to right across the flat net:

Millimetres throughout. The bottom opening is 60 mm across the front and 40 mm front to back. The tuck flap is 40 mm deep because that is the distance it has to cover, and the extra 12 mm is a tab that turns up inside the tube against the back wall. The two dust flaps cover 28 mm each across the 60 mm opening, which leaves a 4 mm gap down the middle. That gap is in the supplied geometry. It is not a cutting error, and it is not yours to trim away.

The 4 mm gap is also a neat demonstration of the first idea worth internalising: flat adjacency is not folded adjacency. On the sheet, those two dust flaps do not touch. Folded, they are neighbours in the stacking order at the bottom of the box, and so is the tuck flap that lies over both of them. Meanwhile P1 and P2, which are joined by a crease, never touch as surfaces at all — they meet at a right angle and stay there.

Two arithmetic checks are worth running before you cut anything:

• Front panel 60 + side 40 + back 60 + side 40 = 200 mm around. Measure that distance across the body panels on your printout. If it reads 197 or 205, the print is scaled and every panel relationship in your mockup will be wrong by that factor. Print at 100%, never "fit to page."
• The net with its tab spans 215 mm wide and 142 mm tall including the tuck flap, so it sits on A4 landscape with a comfortable margin.

Assign the fold structure without repairing the source

Pick a base panel — the one that stays put while everything else folds around it. Use P2, the front, because it also anchors your reading orientation and keeps the artwork's "up" pointing up. Changing the base panel later should change only the pose of the model in space, not the structure. If a different base panel produces a different fold result, suspect the tool, not the net.

Now assign a direction to each vertical crease. For a tube folded print-side out, all four turn the same way. There is a cheap check on that: going once around a closed loop, the turn at each hinge must sum to 360° for the panels to meet again. Four 90° turns in the same sense give 360°. Reverse just one and you get 90 + 90 + 90 − 90 = 180°, and the loop cannot close. The panels would have to pass through each other.

That is the first planted error, so let's plant it. Assign the crease between P1 and P2 as a reverse fold while the other three fold normally. On paper you find out immediately: the tube twists open, and the glue tab and the back panel fight for the same space. In software the fold angle is just a number, so the error survives quietly until someone inspects it. The turning sum tells you which hinge is wrong without folding anything.

While you are at it, run the cheapest check in the whole method: count the pieces. This net is one connected piece before folding. If cutting leaves you with two, you cut a crease instead of scoring it, and you have just destroyed the panel relationship you were trying to inspect. One piece, every time.

Then resist the temptation that follows. If the net fails to fold, the fix is to find the misassignment — not to let the tool auto-correct a drawing inaccuracy, and not to snip the paper until it lies flat. The supplied net is an approved artifact. Correct it only on a copy that is explicitly labelled illustrative, or after the owner confirms the change and tells you what the corrected version is.

The same decisions in a folding tool

If you are working digitally rather than with card stock, the decisions are the same ones and only the vocabulary changes. Esko's version 12 documentation for Studio Toolkit for Boxes describes a sequence that distinguishes cut and fold lines, deals with drawing inaccuracies, establishes a base panel, sets board thickness and fold angles, and then saves a structural model (the version-12 page).

Read that as a description of the problem, not a map of menus. The page carries a 2012 copyright and describes version 12; interfaces since then have moved, and I have not run a fold in a current version, so I am not going to hand you a sequence of clicks that may have stopped existing. Check the documentation for the version actually installed on your machine. What carries across is the shape of the task: identify the lines, choose the base, declare the material, save the structure — and never let an automatic correction quietly alter a drawing someone approved.

Declare the material, then fold

Thickness is the assumption that does most of the work here, and it is the one most often left unsaid. A zero-thickness model is a useful fiction — it makes the fold topology easy to see — but it has a consequence worth naming out loud: the two dust flaps and the tuck flap all end up in the same plane. They overlap perfectly, which is impossible with real board.

So declare a number. If the source owner has given you a board caliper, use it. If not, say what you used instead: "paper mockup, 0.2 mm card, thinner than the likely production board." Then fold in a stated order — left dust flap, right dust flap, tuck flap, tab — and notice where the material stacks. When the flaps compete for the same plane in a physical mockup, you are looking at your assumption rather than a defect in the net. Write it down and move on. Never trim a flap to make a paper model lie flat; that is a silent edit to an approved structure, and it is precisely the kind of correction that hides itself until the first real sample arrives.

Three more assumptions belong on the same list: that a paper tuck tab will not actually retain anything (you will tape it); that the 4 mm gap at the bottom is intentional; and that a mockup says nothing about whether the tab clears the back panel's edge in production board, whether the carton opens without tearing, or how print registration lands on the finished surface.

Keep the top of the fixture open. It costs you nothing and it gives you a direct view of the single most common error in the whole exercise: print on the inside.

Put the mark where the mistake would show

Artwork is where a structurally correct mockup still goes wrong, because the structure can be perfect while a panel's content is rotated, mirrored, or assigned to the wrong face.

Two conventions make those failures visible. Add them to your mockup copy, not to the approved artwork file.

A tick in one corner of every panel. A 6 × 6 mm bar, 8 mm in from the top-left corner of each of P1, P2, P3 and P4 as each is laid out in the flat net. For a rectangular tube folded print-side out, the panel that sits at the left in the flat net appears at the left when you look at that face from outside. So viewed from outside, every face should show its tick at the top-left. A tick at the top-right means that panel is mirrored; at the bottom-right, rotated 180°; at the bottom-left, flipped vertically. Confirm this convention against your own net before you rely on it — it holds for a convex tube, not for every construction.

A band across two creases. A 6 mm-tall stripe whose bottom edge sits 30 mm above the bottom crease line, running 10 mm onto P2, the full 40 mm of P3, and 10 mm onto P4. Folded, it should run around the front-right corner and the back-right corner of the box at a constant height, unbroken, with no jog at either join.

Now plant the second error: place the artwork for P4, the back panel, rotated 180° in your mockup.

The flat view catches it immediately. The band's 10 mm tail, which should join the band at the P3|P4 crease at 30 mm above the bottom, instead appears on the wrong end of P4, against the seam with the glue tab, at 54 mm above the bottom. Rotating a panel about its centre maps a height h to 90 − h, so the tail is 24 mm too high and 40 mm to the wrong side. The tick moves to the bottom-right. Any text on the back panel reads upside down and running the wrong way.

The folded view confirms it. Stand behind the box and look at the back face: upside down text, tick in the wrong corner, band ending in mid-air beside the join. Then look at the box from the front and the right and notice that nothing whatsoever is wrong there — which is the point. A single hero view would have passed this mockup.

Two things to keep straight about the repair. If the rotation happened in your placement, fix your mockup and re-run both checks. If the supplied artwork itself is laid out that way, do not silently fix the source file: report it, with the flat and folded views as evidence, and let the owner issue a corrected version. Those are different problems with different owners.

Inspect a list of views, not a hero view

Work through the box face by face, and check specific things in each view rather than admiring it:

• Flat net. Every panel's tick at its top-left. The band continuous across both creases at one height. Girth measuring 200 mm. One piece.
• Front and back. Text upright and reading normally in both. Tick top-left in both. The band's height matching the flat position.
• Left and right. Same, and check that the artwork intended for a side is on the side, not on the other side with the panel mirrored.
• Bottom, from below. Where the dust flaps and the tuck flap actually stack, and whether the printed side faces outward or has ended up hidden against the box's interior.
• A view into the open top. Which surfaces are printed. If the printed side is facing inward anywhere, every hinge direction is suspect.
• Three-quarter view. Only after everything above passes. This is the picture you show the treatment team, and it earns its place by being the last thing you check rather than the first.

Keep the structure, not just the render

Save the editable structural model with its hinge directions and its declared thickness, save the artwork placement, and save the view list you checked. Then export one clearly provisional reading image, named so that nobody mistakes it for a finished asset, together with the source file's version and the date you folded it.

Keep the open questions next to it rather than in your head: the board caliper you assumed, the tuck tab you taped, the 4 mm gap, the fact that nobody has confirmed whether the tab clears the back panel's edge in production material. A short list of unresolved assumptions is the most useful thing a provisional mockup carries, because it tells the next person exactly where to apply pressure.

What you have at the end is a folded object whose every panel placement can be compared against the supplied structure, whose artwork orientation has been checked from the outside on all four walls, and whose structure is editable the moment the source owner answers a question. What you do not have — and should not claim — is a certified package, a colour proof, a fit test, or a box anyone can manufacture. That is the correct place to stop for a pitch, and it is a much more useful place to stop than a beautiful render that happens to have its back panel upside down.

Frequently asked questions

Why is a clean rectangular box render not enough for a dieline mockup?

A hero view can hide real problems: the back panel may be upside down, bottom flaps may occupy the same plane, and a stripe that should run around a corner may stop dead at it. The job is not to build a box but to make the supplied net fold, then inspect every face and keep the editable structure separate from the picture shown.

What should be confirmed before folding a supplied dieline?

Confirm the file's identity and approval state, the units and whether the drawing is at 1:1, an untouched copy so the original is never overwritten, and the intended package the artwork was designed for. Also establish which lines are cuts and which are creases. If a line's role cannot be established from the supplied information, ask the source owner rather than guessing.

What does flat adjacency versus folded adjacency mean in the example carton?

On the flat sheet, two dust flaps do not touch, but folded they become neighbours in the stacking order at the bottom of the box, along with the tuck flap that lies over both. P1 and P2, joined by a crease, never touch as surfaces at all; they meet at a right angle and stay there. The 4 mm gap at the bottom is in the supplied geometry and is not a cutting error to trim away.

How can artwork orientation be checked across a folded mockup?

Add a 6 × 6 mm tick in one corner of every panel and a 6 mm-tall band that runs across two creases. For a rectangular tube folded print-side out, each panel viewed from outside should show its tick at the top-left; a tick elsewhere means that panel is mirrored, rotated, or flipped. The band should run around the front-right and back-right corners at a constant height, unbroken. If the back panel is rotated 180°, the band's tail appears on the wrong end and too high, and the tick moves to the bottom-right.

What can and cannot be claimed from a folded dieline mockup?

It can let every panel placement be compared against the supplied structure and artwork orientation be checked from the outside. It cannot be claimed as a certified package, a colour proof, a fit test, or a box anyone can manufacture. Unresolved assumptions should travel with it, including the board caliper assumed, whether the tab clears the back panel's edge in production material, and how print registration lands on the finished surface.

Further reading

• Build a Rough Cardboard Set Model for a Commercial Pitch — Read this for a physical card model instead of a digital fold: it sets a scale, records assumptions, and states that an exploratory model is not authority.
• Make a Simple 3D Location Blockout for a Film Pitch — This builds a plain volume model from a footprint and labels every dimension by how you got it, so a provisional model is not read as a survey.