Halftones that survive the press.
Vortex screens your art and shows you what the mesh, the ink and a tired screen are going to do to it. Before you burn anything.
What’s your dot gain?
Most shops don’t know. The step wedge you can buy costs $22 and comes with advice to assume 30%. The meter that would actually tell you starts at $1,450.
So you guess. Then the 50% tint pulls at 65%, the midtones fill in, the art goes flat, and you eat the reprint.
What it doesn’t know about is the emulsion that won’t hold a 3% dot. The mesh that pushes ink sideways. The squeegee running dry at the end of a long pull. The screen that’s four thousand shirts into its life and doesn’t hold what it held on the first one.
That’s the part that decides how the print comes out. Vortex works there.
Two kinds of effect. Only one goes on your film.
Get this backwards and you burn the mistake into the screen.
Geometry
A dot, a stroke, a hole in the emulsion. Real, and it ships.
When Vortex shows you a screen breaking up in the highlights, the film has that breakup in it, and so will the shirt. You can hold it up to the light and point at it.
- Halftone dots, lines, stipple, glyphs
- Dot gain and the smallest dot the screen will hold
- Screen wear, clogged mesh, pinholes, cracked emulsion
- Squeegee drag and a pull that starves
- Underbase and choke
Simulation
These happen at the press, not on the plate. They’re there so you can see what’s coming.
Put misregistration on a film and you’d print the misregistration, then misregister it on top of that. Same for paper tooth. The paper already has tooth.
- Registration drift between screens
- Optical gain and ink that isn’t fully opaque
- Paper texture, tooth, aging
- Show-through from the stock
- The composite proof
A test holds the line. Screen wear has to survive the trip to film. Misregistration has to not. If anybody moves an effect across, the build stops before it can hand a shop a bad plate.
Every effect starts from the thing that causes it.
Each one models something that physically happens on a press. That is what lets them stack together properly instead of fighting each other.
Dot gain
Ink spreads at the edge of a dot, so gain is worst in the midtones. A 50% tint has more edge than anything else on the sheet.
Vortex grows every dot by its own perimeter. It also knows ink on a white underbase spreads further than ink straight onto cotton, and treats the two differently.
The smallest dot you can hold
Emulsion won’t hold a tiny dot. It washes out in the rinse and the highlight is just gone.
Vortex checks the dot on the plate, before spread, because the stencil is what fails, not the ink. So highlights break up instead of fading out, same as they do on your screen.
Optical gain
Light goes into the paper between the dots, bounces around inside it, and comes back out from under them. The dot reads bigger than it is.
Vortex simulates the light instead of asking you to dial in a Yule–Nielsen number. You get that number out of it, not into it.
Distress with hard edges
Wear built from a noise texture always looks soft. Pile up enough random noise and it averages out smooth. That’s the mush you can spot across a room.
Vortex uses phasor noise, which keeps the phase and throws the size away. The edges stay hard, the way ink actually breaks.
Cracked ink
Ink film splits. Then the pieces split again. A new crack stops dead when it hits an older one, so you get T joints, never a clean cross.
Vortex builds crazing by splitting the same way, so the joints and the big-crack-to-little-crack pattern come out right on their own.
Photocopy edges
A copier hits an edge three to five times harder than the middle of a solid. Thin type goes fat and black. Big solids go hollow.
Vortex measures how far every pixel is from an edge and develops it accordingly. Copy the copy and it compounds, same as walking down to the corner store with a fifth-generation flyer.
Damage in the machine’s direction
Squeegee drag runs along the pull. Roller marks repeat around the drum. Mesh moiré sits at the angle the screen was stretched at.
Every defect in Vortex is drawn in the direction of the thing that made it. That one decision is most of why the output reads as a press instead of a filter.
An impression counter
Screens wear out. Tell Vortex how many you’ve pulled and the emulsion starts letting go of the small dots, the ink film thins, pinholes open up.
By 2,000 impressions the smallest dot it’ll hold has gone from around 3% to around 12%. That’s what the trade reports, and it’s what Vortex reports.
The paper, first
Foxing, tide lines, browning at the edges. That belongs to the paper, not the ink, so Vortex lays it down before the ink goes on.
A tide line has a dark ring where the water stopped and a lighter patch behind it. Every fake-vintage filter gets that backwards, and that’s how you spot one.
Two rules we don’t break.
Realism comes out of what you refuse to do. These two rules threw out more code than they let in.
A vignette that darkens toward the middle of the picture is a camera lens. The press has never heard of the middle of your picture. So it goes.
But a sheet that goes dark around the edges is real. Two wet fronts meet in the corner and the paper dries slow there. Looks like the same effect. Only one of them happened.
We deleted effects we’d already built and liked to keep this rule.
Four screens wear four different ways. Separations that all break in the same spot never came off a press.
But they all print onto the same sheet. So the foxing, the tooth, the tide line have to be identical on every plate, or the eye reads four sheets stacked up instead of one print.
Vortex won’t let you get this wrong. The code that models the sheet can’t even ask which screen it’s on.
Print one page. Photograph it. Get your real numbers.
You shouldn’t need a $1,450 meter to find out what your own shop does to a midtone.
You get a PDF, not an image
A PDF knows how big it is. Open a PNG, hit “fit to page”, and it scales. Once it scales, the ruling isn’t the ruling anymore and you’ve measured nothing. So the target comes out as a PDF and prints at exactly the size it says.
Print it on your press
Your ink, your mesh, your stock, your squeegee, your screen at whatever age it’s at today. Vortex screens the patches itself with the simulation switched off, so the only thing changing them is your shop.
Photograph it and drop it back in
Vortex reads each patch against the bare paper right next to it, not against a white square way over on the other side of the sheet. A shadow across your table doesn’t end up in the answer that way. Out comes your gain curve, plus a Photoshop .acv that corrects for it.
No meter, no charge, and it works whether or not you ever separate a job in Vortex. If you take one thing off this page, take this.
Run it again next year and the plates match.
A print run is an edition. When a shop comes back eighteen months later on a different computer, the films have to come out the same, or the reprint doesn’t match the run and your limited edition quietly stopped being limited.
Vortex’s core is compiled twice: once to WebAssembly for the browser, once as a native binary. Before anything ships, a check renders the same job both ways and compares them byte for byte. One pixel moves and the build stops.
Same file in, same plates out. That’s the whole promise, and it’s the reason for a lot of unglamorous work underneath.
Films you can actually burn.
The right size and resolution, labelled well enough to survive a light table.
Film, as PDF
One page per screen, at your imagesetter’s resolution, with a slug. The size is baked in, so nothing gets scaled by accident.
Film, as 1-bit TIFF
PackBits, no anti-aliasing anywhere. A grey pixel on a separation is a lie, so there aren’t any.
Up to eight plates
Separated against a GRACoL profile, with a white underbase and a proper choke.
Your correction curve
A Photoshop .acv built from your own measured press, not from a table of averages.
The films say what they are. A 1-bit TIFF used to carry its name only in the filename, so the second you printed it you had four anonymous sheets of black dots on the light table and no way to tell which one was cyan. Now it’s burned into the film:
Even the slug follows the rules. It’s drawn with strokes, not from a bitmap font, and there’s a printing reason for that. A bitmap font’s stems get thinner as the type gets smaller, and a stem thinner than your mesh will hold doesn’t print thin, it breaks up. Drawn with strokes, the weight is set in millimetres and stays put no matter how small the text goes.
The slug is added below the art, never around it. You register to the art, not to the edge of the film, so not one pixel of your separation moves.
Where your art goes.
Vortex runs in your browser. Loading the art, separating it, screening it, proofing it and exporting the film all happen on your own machine. Nothing is sent anywhere while you work.
If you want your projects saved in the cloud so you can pick them up on another machine or send someone a link, sign in and save them. That does upload your art, because that’s what saving it somewhere else means. It’s your call, and it’s off until you make it.
Don’t sign in and nothing ever leaves the computer. Plenty of shops handle client art under NDA, and that has to be a real answer rather than a promise in a policy.
Put a real job through it.
It’s free to open, there’s nothing to install, and you can have a film in about a minute.