Friday, May 02, 2014

Klein K10A photometer for TV colour cablibration

Had a splendid day today getting to grips with the new Klein K10A photometer and testing it out on a couple of Sony broadcast monitors; an LMD-series LCD and a PVM-series OLED.

For many years we've been relying on the PM5639 colour probe - originally made by Phillips and now badged by DK. It's the right hand probe in this picture and they have served us well through hundreds of calibration sessions. However - they are photometers and as such are tied to the display technology that they are intended to be used with. The CRT one is only to be used on old-school tube'd monitors and the LCD head is for newer style broadcast monitors. The reason for this is down to a phenomenon called metamerism (sometimes metameristic failure)  where you attempt to faithfully capture all colours from a wide-band source (sunlight etc) with only a tri-stimulus detector (TV camera, colour probe, human eye) you run into the inadequacies of the source-detector response. Here is the spectrum for a typical phosphor-based colour CRT;

You can see that the blue and green emissions are kinda what you'd like but the red is all over the place! Red was always a challenge with CRTs but the folks who manufactured the calibration probes knew this and they arranged the bias and gain of the photo-diodes in the detector to match the characteristic of the phosphors; this allowed them make good measurements off the front of a CRT but since LCDs and other display types have quite different spectral graphs the CRT probe would struggle to accurately read the colour off the front of a newer style monitor. Metameristic failure.

This image shows the Klein and the DK probes pointing at the same OLED monitor. If you can't read the values;

DK - x:0.314, y:0.312, Y:102.2 Cd/m2
Klein - x:0.329, y:0.313, Y: 98.9 Cd/m2

With the attendant colour temperature difference.

So which is correct? Why the difference? Well, the DK probe is, as mentioned, an LCD probe and so can't be trusted on display types that have a different spectral distribution. How can you then trust the Klein with multiple display types? It's a photometer after all; isn't it built to match the spectral distribution of a certain monitor type? Is it only good for OLEDs? No - they took a different approach. Rather than matching to the spectrum of a certain monitor type they have tweaked/weighted their photo-receptors to match the human eye; so the probe suffers metameristic failure for all display type but it's the same metamers as your eye fails to see; and that's what counts. If I'm matching two monitors it's only important that they look the same.
We all know that CRTs, LCDs, Plasmas have different spectral distributions but so long as to the eye you've got two displays matched they are matched. The Klein allows you to load profiles into the probe head to provide for better matching but it really is sub the last 1%. I followed their instructions for making a new OLED profile and it allowed me to get the white point a few kelvins closer to 6504k but it wasn't that much of a difference. Apparently the 2nd gen. Sony OLED TriMaster monitors have a new formulation in the blue pixels.
Anyway - we're very pleased with the K10A; it's very easy to use, the ChromaSurf software is clear and makes calibrating a breeze.
Anyway - I wrote some notes - colour for TV 101 kind of thing. Grab them here.

No comments: