A buyer-side guide to decorating glass covers — silk screen, UV inkjet and vacuum metallizing compared by artwork, cost structure, durability and what each one does to a wireless charging signal.
Three suppliers quote the same pale pink cover at three different prices, and all three are honest. One is screen printing two passes of ink, one is inkjetting a white underlay plus colour, and one is printing a base coat, metallizing it and tinting the top coat. The surface looks similar in a photo and behaves differently in a pocket. This guide is about choosing the route on purpose — and then writing an approval and test plan that matches the route you chose.
1. Why the Same Colour Has Three Different Prices
Buyers tend to treat decoration as a single line item: "print it pink". In production it is three different businesses, because the processes do not share equipment, setup logic or failure modes.
Silk screen pushes a thick ink film through a mesh, one plate per colour. It produces strong, opaque colour with crisp edges, and it charges you for a plate and a setup run per colour. UV inkjet sprays droplets and cures them instantly under ultraviolet light. It has no plate, handles gradients and photographic images, and charges you per unit of ink and per pass. Vacuum metallizing deposits a thin metal film in a vacuum chamber, then protects it with a top coat. It produces the mirror and depth effects that printing cannot, and it charges you for chamber time, masking and yield.
A matte pastel cover such as TR-022 — a light pink panel with a matte finish, unique camera cut-outs and customisable colour and logo, in a 163.2 mm by 74.1 mm by 0.55 mm format with a fine-sand surface and selectable 50% or 75% transmittance — can legitimately be made by any of the three routes. The price difference is not a markup. It is a different number of operations.
2. The Three Routes, Side by Side
| Criterion | Silk screen | UV inkjet | Vacuum metallizing |
| Setup cost | Plate per colour | Low; digital file | High; chamber run and masking |
| Best at | Solid colour, white, fine text, thick opaque coverage | Gradients, photographs, many variants, small lots | Mirror, brushed, deep gloss, colour-with-depth effects |
| Weak at | Photographic images and smooth gradients | High opacity in one pass; visible droplet texture if over-driven | Cost at low volume; conductivity (see section 6) |
| Ink / film thickness | Thick — microns per pass | Thin, built up in passes | Nanometre-scale metal, plus base and top coat |
| Colour matching | Pantone mixing, very repeatable | ICC profile dependent; drift across heads and batches | Dependent on base coat and film thickness |
| Per-unit cost curve | Falls with volume; setup amortised | Nearly flat; volume helps only through yield | Falls sharply with chamber loading |
3. Match the Process to the Artwork, Not to the Price
The fastest way to overpay is to send a flat brand colour to an inkjet house with a large minimum, and the fastest way to underdeliver is to send a photographic gradient to a screen printer. The matrix below is the decision most buyers actually need.
| Artwork type | Sensible route |
| One or two flat brand colours, solid coverage | Silk screen — lowest unit cost at volume, best opacity |
| Fine text, small logo, sharp edges | Silk screen if the stroke width suits the mesh; otherwise UV print |
| Gradients, photographs, many colourways | UV inkjet — no plate, and variants cost almost nothing to set up |
| Mirror, chrome, brushed metal look | Vacuum metallizing, with a clear top coat for abrasion |
| Deep colour under a glossy clear coat | Metallized base plus tinted top coat, or print plus clear coat |
| Matte pastel, single tone | Either screen print or a tinted matte top coat; decide on volume |
One caution: the route determines the texture as well as the colour. A screen-printed surface carries the mesh pattern faintly, an inkjet surface carries the droplet pattern, and a metallized surface carries the base coat's texture. If the brief says "soft matte with no visible pattern", that requirement eliminates routes before price is discussed.
4. What Silk Screen Does Well — and What It Cannot Do
Screen printing remains the workhorse for glass covers because it lays down a thick, opaque film in one pass. White is the clearest example: a single screen-printed white pass covers a dark substrate in a way that inkjet needs several passes and a dedicated white channel to approach.
Its limits are geometric. Gradients are produced by halftone dots, which reads as a pattern at close range. Photographic images need four or more plates and careful registration, and registration error on a 0.55 mm part with a camera aperture is visible. And every extra colour is a plate, a setup and a chance to misregister.
The process is also sensitive to what is underneath it. Printing over a fine-sand texture is not the same as printing over polished glass: ink sits on the peaks, coverage is uneven, and the apparent colour shifts with the texture. If the surface has a textured finish, the ink and the texture have to be approved together. The differences between texture processes themselves are covered in our article on laser etching, AG treatment and nano coating — this article is about what goes on top of that surface.
5. Where UV Printing Wins — and Where It Quietly Fails
UV inkjet wins on variety. Ten colourways, a gradient, a photograph, a serial number: all the same setup cost, which is close to zero. For a brand testing a palette across a small launch, that flexibility is worth more than a lower unit price at volume it may never reach.
It fails in three predictable places. Opacity: without a white underlay, colour on clear or dark glass reads thin and changes with the background. Adhesion: UV ink on glass usually needs a primer, and a primer that is skipped to save a step shows up as edge wear three months later. Texture: droplets build a surface, and on a glossy cover that surface is visible as a fine stipple under raking light.
There is also a colour-management trap. Inkjet colour depends on the profile, the head condition and the ink batch. Two runs of the same file can differ enough to fail a ΔE check that screen printing would pass without effort. If you choose inkjet, agree the measurement method and the light source at the same time as the colour.
6. Vacuum Metallizing: The Mirror Finish and Its Hidden Electrical Cost
Metallizing delivers effects printing cannot: a true mirror, a brushed-metal depth, a colour that appears to sit under the surface. It works by depositing a very thin metal film in a vacuum, usually over a base coat that controls the final texture, then sealing it with a top coat that carries the colour and the abrasion resistance.
The cost that buyers miss is electrical. A continuous metal film is conductive, and a conductive layer on the outside of a phone interacts with antennas and with wireless charging fields. On a cover that sits over a charging coil, a full-area metallized layer can attenuate the field or trigger foreign object detection; the usual fix is a masked window over the coil area, which adds a process step and a visible design constraint. If the product supports wireless charging, the finish and the coil window have to be designed together. Our guide to wireless charging back cover technology sets out the electrical side.
Metallizing also has a durability profile that differs from print: the failure mode is corrosion and adhesion loss at edges and apertures rather than ink abrasion. Salt spray and humidity testing belong in the plan, not just a rub test.
7. Does Decoration Weaken the Glass?
Ink and coatings are thin films on the surface; they do not carry structural load, and a well-run print line does not change the strength of the part. Two process details do matter, though.
The first is cure temperature and handling. Thermal cure steps and repeated handling introduce new opportunities for edge damage, and edge damage is what actually breaks covers. The second is any process that removes glass — laser engraving, etching, sandblasting — because those cut into the compressive layer created by strengthening. On a chemically strengthened part, an engraved logo is a locally weakened area, and its depth and edge quality become strength variables.
Ask a simple question of any decorated sample: was any material removed after strengthening, and if so, how deep and how was the edge finished? The answer separates decoration from decoration-plus-damage. The wider process sequence is described in how back cover manufacturing differs from screen protectors.
8. Why Colour Approval Is the Most Expensive Argument in the Programme
Colour arguments are expensive because they are usually conducted without a shared instrument. One side compares to a screen, the other to a memory, and both are confident. The remedy is procedural and cheap.
Approve against a physical master, signed and dated, with one copy retained by each side. Measure with a spectrophotometer under a stated illuminant, and agree a ΔE tolerance before the first lot — commonly around 1.5 to 2.0 for a commercial match, tighter for adjacent parts that sit side by side. Then add the things a meter cannot capture: gloss level, texture direction, and how the colour behaves under warm indoor light versus daylight.
For metallized and effect finishes, add one more requirement: the approval sample must be a production-representative stack, not a bare glass chip with ink on it. A mirror colour read through a top coat is not the same colour as the same ink on bare glass, and the difference is large enough to reopen a closed approval.
9. What "50% or 75% Transmittance" Means When You Print on It
Some covers in this range are offered with selectable transmittance — the fine-sand TR-022 platform, for example, lists 50% and 75% options. That is not a decorative nicety. Transmittance decides how much light passes through the decorated panel, which matters whenever there is a sensor, an indicator or a lighting effect behind the glass.
The important point for buyers is that decoration and transmittance multiply. If the substrate passes 50% and the printed layer passes half of that, the effective transmission is a quarter, and a sensor that was specified against the substrate number will underperform. Specify the transmittance of the finished decorated part, at the wavelength the sensor uses, not the transmittance of the raw glass.
The same logic applies to infrared: a stack tuned for visible light can behave very differently at 940 nm, which is where depth sensors and proximity sensors operate. If the phone has one, put the requirement in the drawing as a number with a wavelength attached.
10. The Tests That Actually Predict Field Failures
| Test | What it catches | Most relevant to |
| Cross-hatch adhesion | Ink or coating lifting from the glass | All three routes; the first test to run |
| Abrasion (Taber or equivalent) | Wear-through in a pocket or bag | Screen print and metallizing; inkjet gloss |
| Solvent rub | Attack by hand cream, sanitiser and cleaning agents | All; especially low-cost ink systems |
| Humidity (85/85) | Blistering, haze and adhesion loss | Metallizing in particular |
| Salt spray | Corrosion at edges and apertures | Metallized finishes and metal logo badges |
| UV ageing | Fading and yellowing of the top coat or ink | Pale and pastel colours, clear coats |
| Colorimetric check (ΔE) | Batch-to-batch drift, head drift on inkjet | Any programme with more than one lot |
Two notes on how these are usually misapplied. Adhesion is often tested on a flat, clean coupon rather than on a production part with a chamfer and a camera hole, where real failures happen. And abrasion is often tested on the printed area only, when the vulnerable area is the edge of the print where the film terminates.
11. Registration: Why the Camera Hole Is Where Print Looks Worst
A printed mask has to stop at the aperture, and the aperture has its own tolerance. The two errors combine: if the hole moves one way and the print the other, the visible result is a crescent of missing ink on one side and ink creeping onto the chamfer on the other.
The fix is a specified keep-out and a specified acceptable ring. State the maximum width of unprinted area around the aperture, whether ink on the chamfer is a defect, and the magnification at which it is judged. Then inspect the aperture in the same operation as the colour, because it is the feature a customer photographs.
This is one place where a supplier's process claim becomes concrete. "Dust-free silk screen" is a marketing phrase until it is attached to a defect list: particles under the print, ink bridging a small hole, and halo at the aperture. Ask which of those three the line is set up to catch.
12. Small Lots, Many Colourways: Where the Cost Really Sits
Decoration cost is driven by changeovers, not by ink. A screen-printed programme with six colours and four variants can carry twenty-four setups, and each setup consumes time, material and a first-off inspection. An inkjet programme with the same variety carries almost none, which is why it can be cheaper at low volume and more expensive at high volume.
The practical question to ask is not "what is the minimum order" but "what is the minimum order per colourway". Those are different numbers, and the second one is the one that hurts when a brand launches eight colours and sells three.
A second question is whether the supplier prints in-house or subcontracts. Subcontracted decoration adds a transport step, a second handling risk on a thin glass part, and a quality boundary that nobody owns. For a decorated cover, the decoration and the glass should ideally be under one quality system.
13. What to Send With the Artwork
Vector artwork at final size, with strokes above the minimum the chosen process can hold, and text converted to outlines.
Colour references: Pantone numbers for flat colours, an ICC-tagged file for anything photographic, and a physical reference where the colour is critical.
White underlay instruction: where opacity is needed, and where the glass should remain translucent.
Keep-outs: apertures, sensor windows, antenna and coil areas, and any bonding land that must stay unprinted for adhesion.
Finish and gloss target, on a scale the supplier already uses, plus a statement of whether a visible mesh or droplet pattern is acceptable.
Durability requirements, named as tests with acceptance values rather than as adjectives.
Product-level examples are useful here: a cover like TR-022 is a matte pastel panel with camera cut-outs and custom colour and logo; a platform like RT-L3001 lists more than twenty colour and texture options including AG, holographic and sky mirror. Those two briefs would not be sent to the same process by default, and that is the whole point of this article.
14. Where Buyers Lose Money on Decoration
Choosing a process from a photo. The photo shows colour; it does not show the mesh, the droplet pattern or the depth of a metallized layer.
Approving colour on a screen. The most common and most avoidable argument in the programme.
Ignoring conductivity on a metallized cover. A mirror finish over a charging coil is an electrical decision disguised as a cosmetic one.
Specifying transmittance on raw glass. The printed part is what the sensor sees.
Testing adhesion on coupons. Real parts have chamfers, apertures and edges, which is where coatings fail.
Launching eight colourways without per-colour minimums. The inventory cost lands after the launch, not in the quotation.
Letting decoration and glass sit under different quality systems. When a defect appears, nobody can say which process caused it.
15. Frequently Asked Questions
Which decoration process is cheapest? It depends on colour count and volume, not on a fixed ranking. Silk screen wins on flat colour at volume, inkjet wins on many variants at low volume, and metallizing wins only when the chamber is well loaded. Ask for a price at your actual mix, not at a nominal quantity.
Can you print a photograph on a glass back cover? Yes, by UV inkjet, usually over a white underlay. Expect a visible droplet texture on gloss unless a clear coat is added, and agree the ΔE tolerance because inkjet drifts more than screen print.
Does a metallized finish block wireless charging? A continuous metal film can attenuate the field or trigger foreign object detection. The usual solution is a masked window over the coil area, designed together with the finish rather than added afterwards.
Will printing change the strength of the glass? Printing itself does not. Processes that remove glass after strengthening — engraving, etching, blasting — do, because they cut into the compressive layer.
Do you print and finish in-house? Colour and logo customisation, CNC engraving and dust-free silk screen are part of the production scope described for these covers. For a specific programme, confirm the process route and the quality ownership in writing before tooling.
Decorating a Glass Cover for Your Brand?
Send the artwork, the colour reference and the durability requirement. We produce decorated glass back covers with silk screen, print and metallized finishes, in matte, fine-sand and effect surfaces, and we will tell you which route fits the look you want — and which one will cost you more in the field.