top of page

Film Positive Drift in Screen Printing Prepress: Why Your Color Separation Grows Out of Register

3 days ago
8 min read

A separation can be signed off in the browser and still fail on press. Between the digital file and the exposed stencil sits the last analog step in prepress: a plastic sheet. Each color is carried by its own film positive, and those sheets can change dimension, distort locally or fail to hold the intended halftone. Accurate separation is necessary, but it cannot compensate for unstable film or inconsistent output.

How much film drift actually matters in screen printing registration

Registration error is usually treated as a press problem: a micro-registration tweak, a platen adjustment, a nudge on the squeegee. But there is a version of it the press cannot touch. If one positive in a set has grown or shrunk relative to the others, the colors get progressively further apart the farther they sit from the center of the image, and no amount of press adjustment pulls them back together. That is a film problem wearing a press problem's clothes.

The arithmetic: what a tenth of a percent does at press size

The table shows the change in length if a positive expands or contracts uniformly. It is a simple way to put small percentages in press-size terms.

Percentage change

Run length

Error in inches

Error in millimeters

0.1%

10 in

0.010 in

0.254 mm

0.1%

12 in

0.012 in

0.305 mm

0.1%

14 in

0.014 in

0.356 mm

0.25%

10 in

0.025 in

0.635 mm

0.25%

12 in

0.030 in

0.762 mm

0.25%

14 in

0.035 in

0.889 mm

In multicolor apparel work, a shift of a few tenths of a millimeter is often enough to be visible at a print edge. A two-tenths-of-a-percent length change on a 14-inch print is 0.028 inch, or about 0.71 mm, already a problem on fine edges or halftone detail. Published tolerances vary by job: simple spot-color work may hide movement that tight halftones, small type or a narrow underbase edge will reveal.

Why polyester film positives move in a screen printing prepress room

Biaxially oriented PET (BOPET) is preferred for film positives because it is generally more dimensionally stable than acetate or polypropylene. But PET is not inert. Temperature, humidity, internal stress and the imaging process can all affect the sheet.

Humidity is the usual suspect behind a color separation that changes size

Humidity is the first thing most shops blame, and it does move film. Kodak’s polyester film technical bulletin publishes a humidity coefficient of 0.0008% per percent relative humidity for one PET base, and notes those values are averages, a general guide rather than a manufacturing specification. Run that number: a 20-point RH swing changes a 14-inch positive by about 0.016%, or roughly 0.002 inch. Small. It matters on a tight halftone edge or a narrow choked underbase, not on a bold two-color logo.

Acetate is more moisture-sensitive; polypropylene absorbs little moisture but can be more vulnerable to heat-related shrinkage. The base material, coating and supplier’s data sheet matter more than the generic polymer name.

Heat and ink load push the film much farther

Now the other direction, and it is larger. The same bulletin lists shrinkage of up to 0.06% after exposure to about 82°C and up to 0.15% at 100°C. At 0.15%, a 14-inch positive loses about 0.021 inch, or over half a millimeter, in length. That is the number that ruins a ten-color job, and it comes from heat, not from the weather.

Toner fusers and thermal imaging expose film to heat. As Screen Printing Magazine’s film-positive reference explains, heat can distort polyester, and mixed high- and low-coverage regions can create localized misregistration when shrinkage is uneven. A capstan drive or friction-feed printer can also introduce transport distortion; printer tracking accuracy is a separate variable from film stability.

Inkjet positives have a different risk: with some swellable coatings, heavy aqueous ink load can expand the coating unevenly, curl or cockle the sheet. These are material and output-system effects described in industry guidance, not InkSplit test results. Consult the film and printer manufacturers’ specifications.

Contact in the vacuum frame is a film problem too

A positive that will not lie flat never fully contacts the emulsion, and the exposure softens along exactly the edges you were trying to protect. Curl from storage, a sheet printed too close to the edge of the roll, or a positive that was rolled for transport all show up as fuzzy detail that looks like an exposure fault. Let cut sheets relax flat under a clean weight before use rather than pressing them flat with extra vacuum and hoping.

InkSplit print preview bar showing a Class of 2026 design at 2100 by 2850 pixels, 11 screens, and mesh counts 156, 190, 230 and 305

The print preview bar carries the output decisions a film inherits: finished pixel size, screen count, and the mesh set the job was built for. Inspect each channel before output; no preview verifies the dimensional stability of the printed sheet.

Six ways film positives lose registration before the screen is burned

Prepress causes of a color separation that will not line up

Six of them account for most of the drift described in shop forums and supplier troubleshooting notes. The pattern repeats: one positive in the set ends up living in different conditions from its neighbors.

Symptom

Likely source to check

Channels printed hours apart in changing room conditions

Humidity or temperature changed between output runs; film did not equilibrate consistently.

Film stored rolled, leaning or in a hot vehicle

Curl, heat exposure or uneven stress can compromise flatness and dimensional stability.

Film used straight from a cold delivery

The sheet may be at a different temperature from the output room; allow it to acclimate.

Heavy coverage differs sharply between channels

Thermal imaging or toner load may create local shrinkage; inkjet coating load may create curl or cockling.

Positive taped or tensioned onto the screen

Uneven tape placement or pulling can stretch or skew the sheet during contact.

Old film set reused after artwork changed

The new channels may not share the old set’s scale, marks or output history. Rebuild and verify as a matched set.

Film density and halation: the other half of screen printing halftone accuracy

A film can register dimensionally and still expose poorly. The halftone dot that leaves the RIP must be blocked cleanly by the positive and transferred to the screen without light scattering softening its edge.

Dmax, Dmin and dynamic range on a film positive

Dmax is the UV-blocking density of the imaged area; Dmin is the density of the nominally clear film. The difference is dynamic range. In SGIA’s film technology guide, the recommended minimum Dmax for screen printing is 3.5 D, with a warning that values below 3.0 D let light through and expose emulsion that was meant to stay open. That same guide puts acceptable Dmin below about 0.05 D. For halftones, a minimum dynamic range around 2.5 is often cited, with 3.0 or better preferred. High Dmin haze blocks UV in the clear areas too, costing exposure time and narrowing the process window. These are published targets, not universal acceptance limits: measure and test your film, emulsion and exposure system together.

Why dyed mesh changes the halftone your color separation produces

Undyed white mesh scatters more light through its threads, contributing to halation and undercutting. Yellow dyed mesh reduces that scatter and can sharpen stencil edges. General guidance often calls for 50–100% more exposure on dyed mesh, but Graphics PRO notes an LED exception: its published claim is that LED units behave the same on dyed and undyed mesh. Test your own exposure unit rather than applying a fixed multiplier. Dyed mesh is commonly recommended from about 102 threads per inch upward for finer detail.

The smallest dot your screen printing mesh can physically hold

Halftones: use mesh geometry, not mesh count alone

Published rules of thumb disagree: a dot may need to span two thread diameters plus one mesh opening, two openings plus two threads, or at least four mesh openings. They are approximations; stencil thickness (including EOM), tension, exposure, ink and press conditions affect what holds in production.

The table below gives representative values from a published NorCal SPS mesh chart. Note what it reveals: 110 and 160 mesh both list 42% open area, yet 110 has an 80 µm thread and a 150 µm opening while 160 has a 55 µm thread and a 103 µm opening. Open area alone tells you nothing about the dot a screen will hold. The thread diameter and opening do, so read the mesh datasheet rather than the count on the frame.

Mesh count

Thread diameter

Mesh opening

Open area

110

80 µm

150 µm

42%

160

55 µm

103 µm

42%

230

48 µm

62 µm

32%

305

34 µm

49 µm

27%

Worked example using the two-threads-plus-one-opening rule on 230 mesh: (2 × 48 µm) + 62 µm = 158 µm estimated minimum dot diameter. This is a geometric floor, not a guarantee that a dot will expose and print reliably.

A common starting convention is mesh count divided by 4.5 for approximate LPI: 230 ÷ 4.5 ≈ 51 LPI. Textile halftones generally sit around 40–65 LPI. The Zorofilm LPI guide documents both the convention and the differing published dot rules; validate the combination on your own screens.

Registration shift or dot gain: telling two screen printing failures apart

Both failures arrive as a complaint about the print, and they get fixed in completely different places. Chasing dot gain with new films wastes a set; chasing film drift with exposure changes wastes a day.

What the press shows

Registration drift

Dot gain

Where it appears

Edges of one color against another, growing with distance from the center of the image

Every tone printing darker and fuller, shadow areas plugging first

Behavior across the run

Consistent, and it does not change with squeegee pressure or speed

Sensitive to pressure, speed, durometer and ink viscosity

First thing to check

Registration marks on the films, measured against each other

Film dot percentages against the printed dot percentages

Second thing to check

Whether the channels were output in the same session and condition

Mesh, EOM, exposure and press settings

A tell that costs nothing: measure between two registration marks on the films themselves before they go near a screen. If the marks disagree, the press cannot fix it.

A film-handling protocol for screen printing color separation output

  1. Bring film into the output room and let it acclimate before imaging. Two hours minimum if it came off a cold truck, overnight for a full box.

  2. Keep the room’s temperature and relative humidity in a stable range rather than perfectly on a target number. A steady 48% does less damage than a room swinging between 30% and 60% around a job.

  3. Print every channel for a job in one session, using the same device, media lot and output settings.

  4. Store cut sheets flat, protected from heat, sunlight, dust and pressure.

  5. Measure registration marks across two channels with a steel rule or calipers; log the distance and conditions.

  6. If a film is out, re-image it. Patching or stretching one channel can introduce new local errors.

Film positives with visible registration marks on a backlit prepress bench beside a steel rule, calipers and hygrometer

Illustrative shop photography, not an InkSplit screenshot.

Where automated color separation fits

Color separation review before film output

InkSplit handles the digital checks upstream of film handling. Its live print preview shows the artwork on the actual garment color, so a shop can see whether the separation holds up before anyone commits a screen. Channel isolation opens each plate on its own so an operator can read its dot structure, and lets the underbase be inspected separately from the colors sitting on it. That is what catches a runaway solid white or a channel carrying far more coverage than the artwork needs, both of which turn into film and ink problems downstream.

The honest boundary: none of that proves a transparency is dimensionally stable or properly exposed. The separation can be right and the film can still be wrong.

Cloud project storage lets a shop reopen the same separation for a reorder rather than silently rebuilding it. InkSplit exports to the RIP a shop already uses; it does not replace that RIP. Shops without a preferred RIP method can use the built-in RIP option. Browser-based access also means freedom from workstation-specific installations and added IT setup. See InkSplit features.

InkSplit reports benchmarks of 1–3 hours saved and about $45 saved per job; actual savings vary with artwork and workflow. The point of Ready, Sep, Go.™ is a faster route through separation, not skipping the film and screen checks that protect registration.

Where a screen printing color separation actually ends

When colors drift on press, check the entire chain: digital channel, output device, film base and coating, storage conditions, registration marks, mesh and exposure. A mathematically sound separation is only as accurate as the positive carrying it. Stable conditions, matched output sessions and measured marks turn film handling from an assumption into a controlled prepress step.

 
 
 

Comments


bottom of page