I shot the four hand-held images sequentially top-to-bottom with Image Stabilization OFF, ON, ON, then OFF. All shots were at full zoom (85mm - equiv to about 135mm on a 35mm film SLR).
The subject is a watercolor print, shot indoors under poor lighting conditions from about 12 ft away. The images were zoomed to better highlight the blur, then jpeg compressed before uploading onto the Pbase site. I used the ISO 3200 setting (therefore, quite a bit of noise is apparent). Even so, the low lighting required 1/20th of a second at f5.6, so in the top and bottom shots, my inability to hold the camera steady is obvious. However, with the IS turned on, the middle two shots show a lot less blur, proving out the functionality of the IS feature. It looks like I should be able to gain an f-stop in certain conditions using the IS.
Then there's the whole issue of repeatability. In the non-stabilized shots, it looks like I improved at holding the camera still the second time (fourth shot vs first shot). My results without IS are obviously subject to my minute-to-minute ability to hold the camera steady. This variability is eliminated with the IS, as the two middle shots are not only sharper, but show less variability in degree of sharpness between them. My suspicion is that I can input varying degrees of (minor) unsteadiness and the output is relatively similar sharpness. Not bad.
