Introduction
In a recent virtual panel discussion, we tackled one of the most persistent hurdles in digital product creation for fashion: how to achieve accurate and consistent colour across the supply chain, from screen to sample to store.
Moderated by Michael Ratcliffe (PI Apparel), the session featured Olaf Koelling (DMIx by ColorDigital), Sylwia Szymczyk (FashionINSTA.ai & Former Timberland), Michaela Jauk (Odlo), and Gabrielle Shiner-Hill (Bureau 555), four experts deeply embedded in the world of DPC, colour data, and material digitisation.
The panel explored six core questions shaping the future of digital colour accuracy, from the root causes of colour mismatch between digital and physical products, through to AI-driven colour correction and real-time rendering.
🎥 Watch the full video below!
Don’t have time to watch the full video? Scroll down for a summary of key takeaways and noteworthy quotes.
🔑 Key Takeaways
🎯 1. Colour Accuracy Depends on a “Single Source of Truth”
The foundation of digital colour consistency is accurate data. Without a precise, standardised starting point – often defined by spectral measurements or LAB values – brands risk misalignment across photography, 3D renderings, and physical samples. One trusted, calibrated environment is essential.
💡🖥️ 2. Display Technology & Lighting Still Skew Perception
From uncalibrated monitors to unpredictable daylight shifts, colour perception changes drastically depending on the viewing context. Even with well-managed data, brands must account for the human and hardware variables that distort how colour is seen at different points in the value chain.
📊🌍 3. Spectral Measurement Enables Remote & Reliable Approval
Using spectral data for colour verification allows teams to reduce back-and-forth sampling, approve materials remotely, and ensure consistent outcomes across global suppliers. The technology is now precise enough to trust over subjective visual checks, but adoption still requires a cultural shift.
🤝 4. Cross-Functional Alignment Is Crucial
Designers, DPC teams, eComm, fabric technologists, and manufacturers all interpret colour differently. Without clearly defined ownership and shared standards, brands face breakdowns in communication. Successful companies are assigning colour ownership, often to colour designers or technologists, and formalising approval workflows.
🧵 5. Material Simulation Affects Colour Representation
Accurate colour in 3D design depends not just on the colour data itself, but how materials behave under lighting. From roughness and reflectance to transparency and specular maps, every channel influences the final look, making lighting setups and consistent rendering environments non-negotiable.
⚙️ 6. AI Has Potential But Isn’t a Silver Bullet Yet
While AI is showing promise in areas like image enhancement and lighting simulation, it’s not yet a reliable fix for accurate colour transformation or correction. However, panellists agreed the space is evolving rapidly, and success will depend on large, diverse training datasets.
➡️ 7. Change Management Is the Biggest Barrier
Technology isn’t the issue—people are. Long-established habits, siloed teams, and lack of visibility into successful case studies are the real obstacles to progress. Teams must invest in onboarding, communicate value clearly, and create feedback loops that reinforce adoption.
💬 Noteworthy Quotes
💚 Olaf Koelling, ColorDigital GmbH
“If your data’s not right, you can adjust what you want but it’ll still be a mess.”
“Sometimes we had four different versions of a single colour, depending on who viewed it where.”
“You can produce the same fabric in Turkey or Vietnam—but only if your colour data is standardised.”
“AI is improving quickly, but right now, algorithms are still better for colour control.”
💛 Sylwia Szymczyk, fashionINSTA.ai
“The questions we had about colours and calibration most often came from the eComm team. They wanted the 3D colours to match the ones from the photoshoots. But the photos were edited, not true to life. So we had to reverse-engineer things in Lightroom to get them closer.”
“Companies are gatekeeping their success stories. That’s holding the industry back.”
“Everyone who worked digitally had this huge PDF that explained exactly how assets had to be delivered. It didn’t matter who created them—freelancers, suppliers—everything looked the same. That consistency made a huge difference.”
💜 Michaela Jauk, Odlo
“Material simulation isn’t just texture – it’s about how the surface reacts to light.”
“We realised that if we gave the retoucher the 3D render, the physical sample, and the photoshoot image all at once, it was actually the render that aligned most closely with the real product. The photos were often heavily edited, and lighting conditions varied, but our render pipeline was consistent. That’s when the team began trusting 3D as a reliable reference—not just a visualisation.”
💙 Gabrielle Shiner-Hill, Bureau 555
“We couldn’t control every screen, so we controlled the capture process instead.”
“It took three months to set up the spectral colour measurement system, but once it was running smoothly, the colourist said, ‘I’d never go back.’ He was finally able to approve colours remotely without endless physical sampling — it completely changed his workflow.”
A huge thank you to all of our panellists for taking part in such a great and important conversation. We look forward to meeting some of you in person at next week’s PI Apparel Europe 2025 event on Lago Maggiore, and on that note, if you are yet to grab your ticket for our 12th European edition, you can do so now by clicking on the image below!🥂🍕
Part 2 – How Do We Better Align People, Process, and Technology to Tackle Limitations & Misconceptions? – coming soon! 🔜
