I have posted at least a couple of times here saying that good testing of VR systems would require a standard shaker table programmed with waveforms developed by having a large number of photographers shoot with instrumented cameras. I then further opined that this would be prohibitively expensive and, hence, manufacturer claims for their systems could not be compared or trusted. Well, much to my surprise the Japanese "Camera & Imaging Products Association" (CIPA) has done exactly that! Shaker table, test data, .. the whole deal. Check out CIPA DC-011 Measurement and Description Method for Image Stabilization Performance of Digital Cameras: Home. Quite impressive, actually. The specs are long and tedious, so I do not claim to have read them in detail. One factoid jumped out, though: "Camera shake has components in six directions (yaw, pitch, roll, x, y, z) in total. However when the shooting distance is about 20 times the focal length, yaw and pitch are dominent, and the other directions can be practically negligible. Accordingly the waveform with only two-axis components is adopted. [for the testing protocol]" So my takeway is to watch to see whether manufacturer claims and data are based on the CIPA tests and to be skeptical about the value of some of the more complicated VR schemes out there.