TL;DR: Artwork files for packaging production are physical and digital assets that degrade, corrupt, and become print-unusable if stored without a defined environment protocol — and most file loss happens not at the agency but after handoff to the factory.
TL;DR: We have rejected incoming artwork packages on at least 12 occasions in the past two years due to corrupted embedded profiles, humidity-damaged proof sheets, or unreadable optical media — all preventable with basic handling controls.
How File Format Degradation and Physical Proof Damage Happen Before Press #
The majority of artwork quality failures we see on our prepress intake desk are not design errors. They are handling failures. A PDF/X-4 file that was structurally correct when the agency sent it can arrive at our RIP station with a mislinked ICC profile, a corrupted spot colour definition, or a flattened layer stack — not because anyone edited it, but because it was stored, copied, or transported incorrectly.
Here is what we track on incoming artwork packages, measured against our internal AP-12 Artwork Intake Checklist:
| Artwork Asset Type | Common Storage Failure | Observable Defect at Prepress |
|---|---|---|
| PDF/X-4 or PDF/X-1a print file | Stored on FAT32 drives; file size truncation above 4 GB | Embedded profile unreadable, overprint settings lost |
| Physical contract proof (Cromalin/Epson) | Stored >60% RH or exposed to direct light | Colour shift, surface delamination, unusable for press match |
| Pantone swatch reference card | Stored in ambient warehouse without UV protection | Measurable delta-E shift within 6–8 months of first use |
| Gravure/flexo film positive (legacy jobs) | Rolled without interleave tissue, stored >25°C | Newton rings, emulsion blocking, dimensional instability |
| USB or optical media (client-supplied) | Stored near magnetic sources or temperature cycling | Read errors, file table corruption, zero-byte files |
The data here reflects our intake log across approximately 18 months of incoming jobs. Physical proofs and optical media account for roughly two-thirds of the failures; digital files on cloud-linked transfers almost never fail at the bit level but frequently carry metadata corruption when exported from DAM systems that don’t preserve ICC tagging on re-export.
Our stance: PDF/X-4 delivered over SFTP or a validated FTP channel with MD5 checksum confirmation is the only format we treat as lossless-received. Everything else gets a secondary integrity check before RIP.
What Actually Goes Wrong — Condition, Mechanism, and What to Inspect #
The most consistent failure mode we see is humidity-driven proof damage. A physical contract proof is typically produced on a pigment inkjet substrate with a polyethylene or resin coating. When that proof is stored above 65% relative humidity for more than 72 hours, the coating begins to absorb moisture, causing micro-blistering and subtle gamut compression in the cyan-to-green range — the range most critical for food packaging brand colours. By the time the proof arrives on our colour desk, the visual reference it provides is no longer aligned to the original L*a*b* target. We then have two choices: reject it and request a new proof, adding 4–7 working days to the schedule, or proceed with the digital reference only and document the deviation. Neither outcome is clean.
The second failure mode is thermal cycling damage to digital storage media. USB drives and portable SSDs that travel internationally in checked luggage or uninsulated air freight undergo repeated transitions between hold temperatures (as low as -15°C in high-altitude cargo holds) and ambient conditions (25–35°C at destination warehouses). Repeated cycling across this range — roughly 40–50°C delta — stresses the NAND memory controller junctions in flash storage. We have received three USB drives in 2023 that passed initial directory read but produced CRC errors on file extraction. The artwork appeared accessible; the actual file data was partially corrupted.
The third failure mode is less discussed but, in our experience, the risk here is greater for brands running long SKU libraries: ICC profile drift in DAM systems. When an artwork file is archived in a digital asset management system and later re-exported, some DAM platforms remap the embedded profile to the system’s default output intent rather than preserving the original. If the original was ISOcoated_v2_eci.icc (the FOGRA39 profile, standard for European offset work per ISO 12647-2) and the re-export maps to sRGB, every CMYK value in the file is reinterpreted. The on-screen preview may look identical, but the RIP will calculate completely different ink densities. We have seen this cause 8–12% dot gain shifts on the first press proof, which is significant enough to require a full reapproval cycle. Check the profile on every re-export, not just at initial artwork creation. The right tool is a preflight check against ISO 15930-7 (PDF/X-4) using Acrobat Pro or Callas pdfToolbox before every file handoff.
Should Physical Proofs Be Shipped to the Factory at All? #
For most jobs running on our sheet-fed offset lines with G7-calibrated output, a certified digital proof file with embedded L*a*b* targets is sufficient — physical proofs add risk without adding accuracy if the receiving press is properly profiled.
Physical proofs are still appropriate for three situations: spot colour match on a new Pantone formula not previously run on our line, metallic or fluorescent ink validation where screen simulation cannot replicate visual effect, and substrate-specific approval for materials we haven’t run before. Outside those three cases, we actively discourage physical proof shipment from overseas clients because the transit and storage conditions are too variable to guarantee integrity at receipt. If a physical proof is required, it must be shipped in a moisture-barrier polybag (WVTR <5 g/m²/24hr per ASTM E96) inside a rigid carton, flat-packed, with silica gel desiccant maintaining ≤40% RH inside the package.
Specification Notes for Brand Partners #
When you brief us on a new artwork package, we need the following to begin a clean prepress intake: the print-ready file in PDF/X-4 format with all fonts embedded and all spot colours defined by Pantone code plus CMYK fallback values, the ICC output profile used at export (FOGRA39, GRACoL 2013, or JAPAN COLOR 2011 depending on your market), and confirmation of the file’s MD5 or SHA-256 checksum so we can verify integrity on receipt.
The single most common gap in client briefs is the absence of a colour profile declaration. We receive files daily where the embedded profile says “Document RGB” or is absent entirely. When that happens, our colour team has to make an assumption, and assumptions require a press proof approval cycle before production can begin.
If you are sending a physical proof for colour approval, contact us before shipping. We will provide our current warehouse humidity and temperature log (we maintain 18–24°C and 45–55% RH in our prepress storage area) and confirm whether the shipment window aligns with our intake schedule. Our standard prepress-to-first-proof timeline for sheet-fed offset is 7–10 working days from receipt of a clean, verified file. File issues extend that, sometimes significantly.
Frequently Asked Questions #
What file format should I use when sending final artwork to the factory?
PDF/X-4 with all fonts embedded, spot colours named per Pantone convention, and an ICC output profile embedded at export. We reject files submitted as native AI or INDD without outlined fonts, and we do not accept files sent as low-resolution JPEGs regardless of the claimed DPI in the document metadata.
How long can I store a physical Pantone proof before it becomes unusable as a colour reference?
It depends on storage conditions. A proof stored at 18–22°C, 40–50% RH, and away from UV light will hold its colour values within an acceptable delta-E tolerance (typically ΔE < 1.5 per ISO 13655) for 12–18 months. Expose the same proof to warehouse fluorescent lighting for 4 weeks and you may see a ΔE shift above 3.0 — enough to cause a visible colour mismatch against press output. We do not accept physical proofs more than 12 months old as production references.
Can I use a cloud storage link instead of sending files by email or USB?
Yes, and for files above 500 MB we require it. We accept SFTP, WeTransfer Pro, or Google Drive links with download expiry set to no less than 14 days. What we require in addition to the download link is an MD5 or SHA-256 checksum generated before upload, so our prepress team can verify the file transferred without corruption. Email attachments are not a reliable transfer method for print files above 25 MB.
Does the factory keep copies of our artwork on file after production?
We archive production-released artwork files for 24 months from the last production date, stored on RAID-6 redundant storage in our server room (maintained at 20°C, 45% RH). After 24 months, files are moved to cold archive and require 3–5 working days to retrieve. Clients who need faster access for repeat orders should maintain their own archive and resend files at next order.
Our agency uses a different colour profile than FOGRA39 — will that cause problems?
It depends on the substrate and printing process. For uncoated stocks, FOGRA47 or the FOGRA52 profile for uncoated paper is more appropriate than FOGRA39, and using the wrong profile will shift neutral tones visibly. For flexo printing on PE film, neither FOGRA profile applies — we work against a custom press characterisation file. When you brief us, tell us which profile your agency used. If it doesn’t match our press characterisation, we convert it through our colour management workflow using ICC-compliant transforms, but we flag every conversion to the client before proceeding because any out-of-gamut colours need explicit approval. This process adds 1–2 working days to prepress but prevents press surprises.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
