The system for measuring latency has remained unchanged for a year. We use this test all the time so as to see how far latency diverts from the manufacturer's claimed values. Up to now, everything has worked fine.

But before getting into too much detail about our measurements, we're going to explain exactly how we go about it, even at the risk of being copied by our competitors. But this is a risk we're prepared to take on board, because the transparency of our approach is an absolute necessity. Indeed, the actual latency of an LCD panel is never found to match that claimed by the manufacturer. We know that and so do they. But it would be impossible to take our measurements and then point an accusing finger at certain manufacturers' practices without explaining clearly what we did to obtain our readings. So in September 2004 we chose to open the method to everyone.

The complete system is composed of:

• An optical probe, which converts the intensity of the brightness to an electrical voltage
• a clean laboratory-standard power supply for powering the probe
• a digital oscilloscope
• a simulating software suite.