Matte or Gloss: How the Finish Decides Whether a Printed QR Code Scans

Matte or gloss: which finish should a printed QR code have?

Specify matte for any QR code the reader cannot re-angle, and gloss only where the object is held in the hand. That single test settles most jobs, because glare is an angle problem rather than a brightness problem: a gloss surface returns light at the mirror angle, and whether that angle points into a camera depends on where the light, the code and the phone are - three things that are fixed on a wall sign and free to move on a business card.

The finish is a scanning decision before it is a design decision. A QR code that measures 3 cm, holds 0.5 mm modules and prints at 21:1 contrast is a good code, and a gloss laminate under a ceiling spotlight can still make it unreadable for everyone standing in the one place people naturally stand. Nothing in the artwork file predicts that, which is why the finish belongs on the print spec rather than in the design review.

Gloss is not the wrong answer everywhere, and treating it as one leads to worse packaging. A gloss film typically deepens the dark modules by a visible step and resists scuffing better than matte, so on a carton picked up in a shop aisle - diffuse overhead light, reader free to tilt - it is both the more durable and the higher-contrast choice. See QR codes on product packaging for the rest of what a pack asks of a code.

Where the answer is genuinely ambiguous, specify matte. The cost of matte on a job that would have tolerated gloss is a slightly lighter black and a slightly softer edge, both of which a black-on-white code has enormous margin for. The cost of gloss on a job that needed matte is a code that fails for a whole class of readers in a way nobody discovers until the print run is on the wall.

Why the finish changes a scan at all

A finish changes a scan because it changes what fraction of the light arriving at the print is sent back towards the camera in one direction. Specular reflection is light bouncing off a surface at the same angle it arrived at, which is what puts a bright patch of glare over part of a printed QR code. A matte surface scatters that light across a wide cone instead, so no single viewing position collects it all.

Glare does not dim the code. It adds the same amount of white light to the dark modules and the light ones inside the patch, which raises the black level and compresses the ratio between them. The decoder measures contrast locally, in small blocks, so the modules under the bright patch lose their difference while the rest of the symbol is untouched - a partially readable code, which is a code that does not read.

That is why the diagnosis is a movement rather than a measurement. If tilting the phone 20 or 30 degrees makes the code decode instantly, the fault is specular and the fix is the finish or the mounting angle. If the code fails from every angle, the fault is size, contrast or the quiet zone, and changing the laminate will not help. The contrast scan test sets out how to separate the two on a printed sheet.

The working contrast targets are 4:1 on a screen and 7:1 in print, and a black code on white paper sits near 21:1, so it has room to lose some. A finish spends part of that room: gloss spends it only at the mirror angle, matte spends a little of it everywhere by lifting the darkest reflectance. A brand-coloured code designed at 7:1 has nothing left to spend either way, which makes finish and colour one decision rather than two.

Matte also costs a small amount of edge definition, because the scattering surface that removes the glare slightly diffuses the boundary between a dark module and a light one. On modules at or above the print floor this is invisible to a decoder. Below the floor it compounds with ink spread, which is the case where a matte laminate and a small code fail together and each looks innocent on its own.

The finishes compared, and what each one costs

Every finish trades glare against durability and black density. This table is the trade in one place, with the QR-specific verdict rather than the print-buying one.

Soft-touch is the finish that surprises people. It is the least reflective surface on the list and therefore excellent for scanning, but it lightens the black more than a standard matte film and it holds fingerprints, so on a dark or coloured code that was already marginal it makes a small problem visible. On a black code at full size it is the safest premium finish available.

Spot UV and foil are the only entries here that are categorically wrong rather than situational. A mirror-bright patch inside a QR code does not carry a reflectance value the decoder can bin as dark or light: it carries whatever is in front of it, which is usually a ceiling light or the phone. Keep both outside the symbol and outside its quiet zone.

Coated or uncoated: the stock decides more than the finish

Uncoated stock is the bigger scanning variable, because it changes the geometry of the modules rather than the light coming off them. Dot gain is the spread of printed ink beyond the size it has in the file, which thickens a QR code's dark modules and narrows the light gaps between them. Coated stock holds the ink on the surface; uncoated stock lets it wick into the fibres, and ISO 12647-2 sets different tone value increase targets for the two for exactly that reason.

The practical consequence is a module floor, not a percentage. Keep printed modules at 0.4 mm or larger on coated stock, and 0.5 mm or larger on uncoated, kraft or recycled stock. A code sized from those floors survives the spread; a code sized from the space left in the layout finds out on press. The arithmetic that turns a floor into a width is in the guide to the minimum size for print.

Measure the spread on a real sheet rather than trusting a number. Scan the printed proof on a flatbed at 1200 dpi, measure the width of a light module against the nominal module size, and halve the difference to get the growth per edge. Anything that consumes more than about a quarter of a light module is a code that needs to be larger, not a code that needs a different finish.

Coated stock and a matte laminate together are the combination that scans best in the widest range of conditions: the coating keeps the modules geometrically correct, and the film keeps the light diffuse. That pairing is worth specifying by name on menus, signage and anything mounted, because a print quote that says only "matte" may still arrive on uncoated stock.

CMYK: build the dark modules from 100% K

Build a printed QR code's dark modules from black ink alone - 100% K, with cyan, magenta and yellow at zero. A four-colour rich black needs all four plates to register perfectly, and typical sheetfed registration tolerance is around 0.1 mm, which on a 0.4 mm module is a quarter of its width picking up a coloured fringe on every edge in the symbol at once.

The light modules must be paper, not a light tint. Knock the background out to 0/0/0/0 rather than screening a colour back to a pale percentage: a 5% tint is a halftone dot pattern inside every light cell, and it lowers the lightest reflectance the decoder measures at exactly the scale where it has the least room.

KoloQR, a free QR code generator with circular and custom-shaped QR codes, exports SVG, PDF, EPS and PNG in RGB, so the conversion happens in the layout rather than in the download. For a flat black code that conversion is a single step: place the vector file, select the dark artwork, and set it to 100% K before the file goes out. For a coloured code, agree the CMYK build with the printer rather than letting an automatic conversion decide - the SVG vs PNG guide covers which file to hand over.

• Dark modules: 100% K only. One spot colour is the other acceptable answer where the brand demands a colour, because it is still one plate.
• Light modules and quiet zone: paper white, no tint, no varnish pattern, no image behind them.
• Total ink limit: irrelevant for a K-only code, and the reason to keep it K-only. Printers commonly specify a limit near 300% on coated stock and lower on uncoated, which a rich black approaches.
• Overprint: switch it off for the code. A dark module set to overprint a coloured panel picks up that colour underneath and loses density.
• Colour profile: convert once, in the layout, with the printer's profile. Converting twice is how a black becomes a rich black without anyone choosing it.
• Varnish and die lines: keep both off the symbol and off the four modules of quiet zone around it.

Check the swatch after conversion rather than trusting the appearance on screen. An RGB black of #000000 converts to a four-colour build under some profiles, and it looks identical in the layout window - the difference only appears on press, as soft edges across every module in the symbol.

Glass, curves, foil: surfaces where the finish decides the outcome

Four surfaces make the finish decision more important than usual, because each of them adds a second reflective layer or removes the reader's ability to move.

Behind glass

A code behind glass has two specular surfaces in front of it, and a matte print only removes one. Position matters more than finish here: mount the code away from the wall the opposite light source sits on, and check the view from where a reader actually stands on the pavement, at their eye height, at the time of day the shop is busiest. Window and shop-front codes are also the case where size is usually wrong at the same time - see the sizes on the retail stores page.

Curved and cylindrical labels

On a cylinder, a gloss surface always produces a bright band somewhere across the curve, because every angle from grazing to head-on is present at once. That band moves as the reader turns the bottle, so it is escapable in the hand and unavoidable on a shelf. Specify matte for bottle, jar, tube and cup labels, keep the code narrower than about a quarter of the circumference, and add roughly 20% to the width.

Laminating pouches and encapsulated menus

A 125-micron gloss laminating pouch is the single most common cause of an unreadable restaurant code, because it is thick, very glossy, and sits directly under pendant lighting. Matte pouches cost the same and remove the problem. Where a venue is already committed to gloss, moving the code from the middle of the sheet towards a lower corner often takes it out of the reflection of the lamp above the table - the printing details for that case are on the restaurant menus page.

Foil, metallic ink and mirrored substrates

A metallic or foil surface has no fixed reflectance, so it cannot serve as either the dark or the light half of a QR code. Metallic dark modules read as bright at the mirror angle and as dark elsewhere; a mirrored background does the same in reverse. Print the symbol in flat ink on a flat area, and use the foil in the surrounding design, where the creative treatments belong anyway.

How to test a finish before the print run

1. Ask the printer for a finished sample. Request the code printed at final size on the production stock with the actual laminate or varnish applied. An unlaminated proof tests none of what this page is about.
2. Print both finishes where the choice is open. Ask for one matte and one gloss sample of the same artwork. The comparison takes two minutes and settles the argument permanently for that venue or that pack.
3. Mount it where it will live. Fix the sample at the real height, on the real wall, in the real frame, or lay it on the real table. A sample held in the hand tests the one condition the finished job will not have.
4. Turn on the real lights. Use the lighting the code will live under, at the hour it matters - the restaurant at dinner service, the aisle under fluorescents, the shop window at dusk with the street lights on.
5. Scan from three positions, standing still. Scan straight on, from the left and from the right, without tilting the phone to help it. A mounted code has to work for someone who does not know that tilting helps.
6. Find the glare angle deliberately. Move until the bright patch lands on the code, then try to scan. Note whether that position is somewhere a reader would naturally stand. If it is, the finish is wrong regardless of how well the other angles read.
7. Check the modules with a loupe. Look at the light cells between the dark modules on the printed sample. If the gaps have visibly narrowed against the artwork, the stock is spreading and the code needs to be larger, not differently finished.
8. Repeat on a second phone. Use one iPhone and one Android handset, at least one of them a few years old. Older sensors need more light, which means more exposure, which makes glare worse.

The glare-angle step is the one people skip and the one that catches the failure. Every code scans from somewhere; a mounted code has to scan from where people stand, and that is a fact about the room rather than about the artwork.

What to put on the print spec

A finish decision is only real if it reaches the printer in writing. These six lines cover everything on this page and fit in an email.

1. Stock and coating, named: for example "300 gsm silk coated" rather than "card".
2. Finish, named as film or varnish: "matte lamination, both sides", not "matte".
3. The dark modules built as 100% K, overprint off, no rich black.
4. The light modules and the four-module quiet zone left as bare paper - no tint, no varnish, no image.
5. No spot UV, foil or embossing within the symbol or its quiet zone.
6. A finished sample at final size before the run, on the production stock.

Send vector artwork alongside that spec so the code can be placed and scaled without softening, and hand over the destination URL in plain text so nobody rebuilds the code by retyping it. The generator exports SVG, PDF and EPS with no watermark, and the resolution decisions for a raster export are in the guide to QR code resolution. The finish is the eighth of nine checks a print run has to pass; the print checklist has the rest in the order they run.

Common mistakes with finish and coating

• Approving an unlaminated proof. The proof scans, the finished job does not, and the difference is the one variable the proof did not include. Ask for a finished sample or accept that the finish is untested.
• Treating glare as a size problem. Enlarging a code that fails under a spotlight produces a larger code that fails under a spotlight. Tilt the phone first: if that fixes it, no amount of extra width will.
• Specifying matte for a code that is already marginal on contrast. Matte lifts the darkest reflectance slightly. A black code has room for that; a mid-grey or brand-coloured code at 7:1 does not, and the answer there is a darker code rather than a different film.
• Letting spot UV follow the brand pattern across the symbol. A varnish plate drawn from the layout rather than by hand routinely runs a gloss stripe over part of the code, which is invisible in the PDF and fatal on the shelf.
• Using a gloss pouch because it was in the drawer. Encapsulated menus and mounted signs are the two jobs where gloss reliably fails, and both are usually laminated in-house without a spec.
• Assuming uncoated is safer because it is matte. Uncoated removes glare and adds ink spread. It is the better light behaviour and the worse module geometry, which is why the module floor rises to 0.5 mm rather than staying where coated stock leaves it.
• Converting the file to CMYK twice. One conversion with the printer's profile is controlled; a second one turns a K-only black into a four-plate build without a visible change on screen.

Finish, stock and size fail together more often than any one of them fails alone. A code that clears the module floor on coated stock, prints as flat 100% K, and carries a matte film is a code with margin in all three directions - and margin is the only thing that survives a print run nobody gets to repeat.