Depth of Field Reducer for DV Cameras

I've built this adapter for my Sony PD-150 video camera that will let me attach old Canon FD lenses to my camera, and consequently exploit the narrow Depth of Field attribute of 35mm lenses.  I've posted a few still frames here for comparison and created a comparison video file of shooting with and without the adapter.  I'll try to get some better footage on here soon, and to clean my adapter a little so that it provides a cleaner picture.  Keep in mind that these examples suffer from some JPEG and MPEG artifacts and are primarily for the purpose of showing the difference in Depth of Field.  Here are the samples:

(click on image for larger version)

Plain PD-150 Lens	PD-150 w/DOF Reducer		Plain PD-150 Lens	PD-150 w/DOF Reducer

Notice the difference in the focus of the mirror in the first sample and the radically different look of the trees and traffic in the second sample.  Both of these samples are completely unmodified.  Here I've included a video comparing the normal footage with the DOF Reduced footage (top left and top right frames).  Just for fun I've included footage of some color modified, de-interlaced, 24fps footage in the bottom right and left frames... another attempt to achieve the so-called film-look that DV-lovers have come to detest.  In my opinion the DOF reducer provides the video with not only more of a film look, but a much more layered (as opposed to flat) look, which helps the viewer pay attention to what the film-maker wants them to, as opposed to struggling with noisy backgrounds and other visual distractions.  Anyway, here's the MPG (1.09MB): Depth of Field Comparison

• The picture comes out upside down and must be flipped back over in NLE (or could use some sort of prism)
• Camera requires much more light (about 2 times as much) to create equivalent DV image
• Image sharpness is sacrificed (could maybe be fixed by using a different light transfer medium)
• Built to use Canon Still Camera FD lenses (could easily be adapted for EOS or any other line of lenses)

So, does this seem interesting to anyone? Any comments or suggestions? E-mail me at okaplan@ucla.edu.