Paul,
The conform process is for outputting HDV back to the HDV camera or to an HDV deck. In my limited experience doing that, it seems to take about six times realtime. If you are editing native HDV on the timeline, and outputting that to HDCAM via HDSDI, you should probably render the entire sequence out to HDCAM in advance, simply to avoid the posibility of dropped frames. There are complex calculations going on in real time when playing back HDV, and I think you actually save time by rendering, because you’ll be confident that everything will be right when going to tape. If someone else thinks the render is not necessary please feel free to make that point…
DRW