Arco Íris De Três Cores

Three-color separation in screen printing is trickier than people think

You grab a design file, open your RIP software, set three colors, and hit output. Then you pull the first squeegee pass and suddenly your gradient looks muddy, the edges are bleeding, and the yellow is fighting the magenta for position on the mesh. This is what happens when people approach arco íris de três cores like it's just a reduced version of four-color process printing. It's not. The tolerances are tighter, the decisions are fewer, and every mistake compounds faster because there's no black plate to hide behind. I spent a couple years doing this work for a couple of local brands. We'd take a full-spectrum rainbow asset and try to reduce it down to cyan, magenta, and yellow. The first dozen runs were garbage. Not because the technique was wrong, but because nobody on the team understood how the halftone angles interact when you only have three plates instead of four. I learned that the hard way, wasting about forty meters of polypropylene film before it clicked.

Why arco íris de três cores is not just CMY minus K

The most common mistake I see is people taking a four-color process separation and simply deleting the black channel. That leaves you with two problem areas. First, the midtones collapse because black was handling the depth. Second, the halftone angles you're used to from CMYK don't apply anymore. With three plates, you need to rethink the angle arrangement entirely to avoid moiré patterns that show up as rainbow banding across the print. The standard angle setup for three-color process is cyan at fifteen degrees, magenta at seventy-five degrees, and yellow at zero or ninety degrees. Yellow has to be isolated because it's so light that even a slight angle change creates visible texture. If you put yellow at forty-five degrees like you would in CMYK, the halftone dots will create a distracting grain on the light areas. That was my first proper realization after a client sent back a run where the skin tones looked like they were made of sandpaper.

Setting up the separation correctly

Start with a spot color layer for each of your three inks. Don't rely on the RIP to auto-convert from RGB. RGB values don't translate predictably to ink coverage, especially in the midtone range where the three colors overlap. I always manually build the separation curves for each channel, checking the dot gain on the specific mesh and ink combination I'm about to use. If you skip this step, your oranges will look brown and your teals will come out green. Here's a specific scenario that caught me off guard. I was working on a print for a festival poster and the client wanted a smooth gradient from yellow through red to blue. The yellow-to-red transition is where magenta and yellow overlap, and that combination is extremely sensitive to registration drift. On our first proof, the magenta plate was shifted two tenths of a millimeter and the red came out as a muddy orange-brown stripe along the edge. I spent an hour checking the screen tension, the squeegee angle, and the mesh count before I realized the issue wasn't the press at all. It was the exposure time on the magenta screen. I had underexposed it by about eight seconds during vacuum exposure, which caused the emulsion to spread slightly during washing out. That tiny expansion translated to positional drift at print time. Re-exposed the screen at the correct time and the registration held perfectly.

Mesh count and ink selection matter more than you'd expect

For a three-color rainbow, you want a finer mesh than you'd use for a standard multi-color design. I typically run one hundred twenty threads per inch for cyan and magenta, and eighty for yellow since yellow ink needs to lay down more deposit to stay visible. If the mesh is too coarse, the halftone dots break up and your smooth gradient turns into a dotted pattern. This is especially noticeable in the transition zones where all three colors overlap to create white or near-white areas. The ink viscosity is another thing people ignore. Yellow ink for this kind of work needs to be slightly thinner than magenta and cyan. If all three inks have the same viscosity, the yellow won't flow through the finer mesh properly and your highlight areas will be too light. I keep the yellow at about one hundred eighty to two hundred thousand centipoise and the other two inks around two hundred fifty to three hundred thousand. Measure it with a viscometer. Guessing doesn't work here.

Common failures and how to fix them

Registration drift is the number one problem. With only three plates, every registration error is visible because there's no black anchor point to stabilize the image. I solve this by using a registration crosshair system on each screen and checking the crosses against a master target before every print run. Takes about three minutes and saves about forty-five minutes of rework. Color bleeding between adjacent halftone dots happens when your ink tack is too high. If the magenta tack is stronger than the yellow tack, the yellow won't transfer cleanly onto the substrate and you'll get color contamination in areas that should be pure cyan or pure magenta. Keep the tack values within fifteen percent of each other across all three inks. Use a tack meter to verify, don't eyeball it from the container. Moiré patterns appear when your halftone angles are wrong for the mesh frequency. I've seen this happen when someone uses the same angle setup for eighty mesh and one hundred twenty mesh without adjusting. The rule of thumb is that finer mesh requires slightly different angles to avoid interference patterns. There are online calculators for this, but they only give you a starting point. Always do a test print on the actual mesh you're using before committing to a production run.

One thing worth noting: arco íris de três cores works well on light-colored substrates like white polyester or cream cardstock. On dark substrates, the yellow layer disappears almost entirely because there's no white underbase to reflect light back through it. If you need this effect on black material, you have to print a white underbase first, which doubles your setup time and introduces another registration variable. In that case, four-color process with a white layer underneath is usually more efficient, even though it defeats the purpose of using three colors in the first place.

Workflow optimization

After you've dialed in your separation and registration, the actual print time for a three-color rainbow on a manual press is about eight to twelve minutes for a medium-sized job, depending on drying time between colors. On an automated table-top press it drops to roughly four to six minutes. The bottleneck is always the yellow layer because you need to verify it separately before moving to the next color. I keep a magnifying loupe on the press at all times and check the yellow coverage after each board before proceeding. If you're looking for a faster way to set up the separations without rebuilding from scratch every time, there are Photoshop actions and scripts that automate the three-color process. I don't use those for final production work because they don't account for your specific mesh and ink variables, but they're useful for quick mockups and client presentations. The actual separation should always be done manually with your press conditions in mind. The skill in this area comes from understanding how light interacts with thin layers of translucent ink on top of each other. Cyan plus yellow makes green. Magenta plus yellow makes red. Cyan plus magenta makes blue. All three together makes a dark neutral. The math is simple. The execution is where everything can go wrong.

I keep a reference card taped to my press with the halftone angles, mesh counts, and ink viscosities for the current job. It prevents me from reverting to old settings when I'm rushing. Three color separation doesn't forgive complacency, but it rewards attention to detail in a way that four-color process doesn't because there's nowhere to hide mistakes.