This page, though updated a bit, is still something of a work in progress. However if you find anything really broken or wrong about it, please tell me about it, using the 'Contact" menu off the home page.
This page will display and convert RGB (8 bit) values to CIE Yxy and XYZ coordinates in real time. I use it to dial in monitors bias and gain settings when I'm calibrating them with a colorimeter running from a remote machine. Most colorimeters I've come accross, expensive ones aside, will read out in Yxy space, thus leaving you to wonder which color to boost or reduce on the monitor. No more! Just mix RGB until you Yxy coords match, and see which way to go.
It is also hand as a learning tool, as you can finally see in an intuative way the relationship between CIE Yxy and XYZ colorspace and RGB, our earstwhile CGI friend. Though this appears somewhat obscure, it's actually quite relevant, as CIE XZY is the 'grandfather' of all other colorspaces. Yxy is another way to state XYZ, and the two are interchangeable. Yxy make a clear seperation of Luma (Y) and chroma (xy), which you can see in the xy chromaticity plot on the right side of the page.
The current iteration of the gadget uses only one whitepoint, though it's important to know that different whitepoints affect the math that makes the conversion to RGB. One day soon I hope to add a dropdown box to select from a few popular whitepoints. Right now it is fixed at 6500K, and also assumes a 2o Standard Observer, using the 1931 chromaticities. While there have been improvements to the chromaticity plot over the years, 1931 still holds strong as the most useful, though not the the most accurate.
Some interesting things to note: Set the background to a dark grey, and see the impact of moving and color. Now start with almost white and notice that the crosshair in the chroma plot moves quite a bit slower. This reinforces the expirience (of frustration) when setting hardware to work properly in the darker regions of it's gamut. Also notice how adjusting blue clears the crosshair way out of the RGB gamut alltogehter when in the darks. I have no idea why this happens, and perhaps a closer look at the math is required, though the coords are largely correct for this model. I'm open to suggestions ;-) -----Beak, 27 May 2005