I need some advice on how to approach a project. I received a color management assignment from an engineering company. They heat treat metal and are in the process of preparing some polished steel discs (semi-matte, not a mirror finish), then heat treated. As the heat treating temperature changes, so does the final color of the heat treated sample, apparently quite reliably. They want me to determine the Pantone color that comes closest to each heat treatment temperature. Their final goal is to hand out printed samples that reliably reflect the temperature at which a sample was heat treated. Sounds simple. My first challenge will be getting reliable L*a*b* readings for a decent match. Just to experiment, I tried sampling the matte aluminum surface on the back of my iPhone directly using my i1Pro spectrophotometer, and got a Lab reading of 77L, -1.1a, 1.5b using D50/2° (5 different samples were all very close). Using the color picker in Photoshop (whole Lab numbers only), it says that Pantone Cool Gray 4C (Pantone Solid Coated Library) is the closest match. If I use the Pantone Premium Metallics Library, it gives me 10103C. They want me to print a few dozen accurate color charts on my Epson printer and semigloss paper, so I will have to use the Pantone Solid Coated Library. I'll probably create the file in Lab color space using the equivalent Lab numbers for each Pantone color. Here's my problem… When I hold the printed Cool Gray 4C from my Pantone swatch book against my iPhone, the Pantone sample looks a lot darker than my iPhone. Of course, if I tilt the iPhone so it receives less direct light, it gets darker and approximates my 4C sample. I'm confident the Lab values are correct. The a* and b* color data is probably quite usable, but I suspect getting a suitable brightness level that prints so it looks like the sample is going to be difficult. It's like photographing any shiny metallic object (like silverware)…I can make the object appear pure white, pure black or any shade of gray I want, depending on my lighting and the family of angles. Hmmmm. If reading directly with a spectrophotometer doesn't work well, I guess I can photograph the samples under controlled lighting, but I am faced with the same dilemma, because I can make the metal as bright or as dark as I wish depending on the lighting setup. Does anybody have experience and advice on how to proceed? Thanks, Lou Dina