It's a common scenario. The ink supplier's lab reads a spot color drawdown and reports that it's on target. The press operator reads the printed product and gets different numbers. Both instruments were calibrated that morning, so one of them must be wrong. But neither is wrong. The two instruments are looking at the sample in different ways, and on many surfaces that alone is enough to change the reading. The difference is called measurement geometry, and it's one of the most common reasons two "correct" measurements don't agree. The previous post in this series, Color Conformance: How Printers Measure Whether Color Actually Matches, showed how Lab values and Delta E turn "does this match?" into a number. It ended with the question of what counts as close enough. Before that can be answered, there's a more basic question: were the two values being compared measured the same way? If they weren't, no tolerance will mean much. For readers new to Lab and Delta E, that post is a quick read, but this one stands on its own. A spectrophotometer works by lighting a sample and measuring the light that comes back. Measurement geometry describes two angles: the angle the light hits the sample from, and the angle the instrument views it from. Two geometries cover almost every instrument in use. A 45/0 instrument lights the sample at a 45-degree angle and views it straight down, at 0 degrees. Many 45/0 instruments light the sample from a full ring around the measuring spot rather than from one direction. This setup reduces the effect of paper grain and surface direction. The key point is where the gloss goes. Light that bounces off a shiny surface leaves at the opposite angle. At 45 degrees in, the glare leaves at 45 degrees on the other side, away from the detector. It's the same thing a person does when tilting a glossy print so the glare doesn't wash out the image during evaluation. A 45/0 instrument sees the color more or less the way a person does. That's why 45/0 is the standard geometry for graphic arts measurement. Most graphic arts handhelds, scanning spectrophotometers, and inline systems in commercial print use it, and published print characterization data is based on it. A sphere instrument puts the sample against the opening of a hollow sphere with a bright white interior. Light bounces around inside the sphere and hits the sample from every direction at once. The instrument views the sample at 8 degrees, just off straight down. This geometry is written d/8 or d:8. Sphere instruments measure in two modes, and many can do both: Instrument makers don't all use the same labels for these modes. Some instruments and software say SPIN and SPEX. Others say SCI and SCE (specular component included and specular component excluded), often written with the geometry, as in "d/8 SCI." SPIN and SCI mean the same thing, as do SPEX and SCE. If a supplier's data says SCI and the instrument in hand says SPIN, both refer to the same mode. SPIN is useful because it measures the colorant with little influence from the surface finish. The same color at two different gloss levels reads much the same in SPIN. SPEX gets closer to what the eye sees, but it isn't the same as 45/0. A 45/0 instrument keeps essentially all of the gloss out of the reading. A sphere in SPEX mode doesn't always let all of it escape. On glossy samples, a SPEX reading and a 45/0 reading still won't match exactly. In graphic arts, 45/0 is the default. But some products are hard to measure consistently at 45/0, and the companies that make them often use sphere instruments for good reason: This is where one of the most common geometry disconnects comes from. The ink and the brand standard are usually built and approved on 45/0, because that's the graphic arts default. The producer, such as a can decorator or a textile printer, measures its finished product with a sphere instrument, because that's what works for its process and substrate. Both are reasonable choices. The trouble starts when numbers from one are compared with numbers from the other. When light hits a glossy printed surface, a small part of it reflects off the top without ever entering the ink. That reflected light carries no color. It's essentially white. A 45/0 instrument mostly misses that light. A sphere instrument in SPIN mode counts all of it. Adding white light to a reading makes the color measure lighter and less saturated. The effect is largest on dark and highly saturated colors, where there's little colored light coming back to begin with. On a matte sample there's less surface reflection, and the readings tend to agree more closely. This is why readings can't be reliably converted from one geometry to the other. The difference depends on the surface, the color, and the finish. A conversion that works for one ink on one substrate won't hold for the next job. A few situations make it worse: Geometry mismatches are usually not obvious. They show up as disagreements that are hard to explain: In each case, the color may be fine, but the comparison isn't. Geometry is one part of how an instrument sees a sample. The light source is another. The previous post covered optical brightening agents, which turn UV light into visible blue. How much UV an instrument puts on the sample changes how much an OBA paper glows, and that changes the reading. The graphic arts measurement standard, ISO 13655, describes this with measurement conditions: The M labels belong to graphic arts instruments. Sphere instruments handle UV their own way, usually with up to three settings: Not every sphere instrument offers all three. Some have no UV filter and can't measure UV excluded. If a reference was measured in M1 and the sphere instrument excludes UV, the readings will differ on materials that contain brighteners. Two instruments with the same geometry can still disagree on OBA stock if they use different UV settings. On materials without OBAs, the settings read much more alike. A Lab value by itself isn't a complete specification. Every color standard that is sent or received should state how it was measured: When a standard arrives without this information, the receiver should ask for it. That's the most important takeaway from this post. The measuring geometry should match the geometry the reference was measured with. For press-side QC against graphic arts references, that's 45/0. For metal, textiles, and other surfaces where 45/0 readings don't repeat well, a sphere may be the right tool. The two can coexist in the same supply chain. What matters is that every party knows which geometry each number came from and that numbers from different geometries aren't compared as if they were the same. When two parties can't measure the same way, the fix is to build a correlation on the actual ink and substrate, not to rely on a general rule of thumb. Matching geometry and UV setting gets two instruments looking at a sample the same way. It doesn't guarantee they agree. Two instruments of the same type can still read differently if one has drifted or hasn't been maintained. The next post in this series covers calibration practices that keep an instrument's readings trustworthy. Alder helps printers, converters, and brand owners sort out measurement disagreements, from choosing instruments that match their references to setting up specifications that state how color is measured, not just what the numbers are. Companies that keep getting different numbers from their suppliers or customers can contact Alder at: General Manager, Color Services, Alder Color Solutions · Great Western Ink Technical Representative, Northern California Lou holds a B.S. in printing science from RIT, including two co-op internships in the prepress group at the National Geographic Society. He was at Barneyscan as the desktop publishing revolution was getting started, then at Adobe as Color Evangelist when ICC profile support was first added to the applications that are now Creative Cloud. He earned G7 Expert certification in 2006, as the program was starting, and G7-calibrated dozens of presses through his own consulting practice before returning to corporate color roles at EFI, Pantone and X-Rite. He sits on the Print Properties Committee of PRINTING United Alliance and is Great Western Ink's technical representative for Northern California.First, a Step Back
1. What measurement geometry means
2. 45/0: measuring the way people look at a print
3. Sphere instruments: SPIN and SPEX
When a sphere makes sense
4. Why the same sample gives different numbers
Where this shows up
5. The light source matters too
6. What to write down with every color standard
Which Geometry to Use
What's Next
Measurement Help from Alder
loup@aldertech.comAbout the Author
Lou Prestia