1) From what I have been led to believe, if you bring in DV over firewire, and then render it untouched to disk as DV, you will have an exact copy. True?

SonicEPM: True

2) I have been led to believe that events are rendered as a unit. If I have a long event and make one tiny little change (fade in at the beginning say), the entire event will be re-rendered and recompressed. The entire event will lose quality and not be an exact duplicate. Still true?

SonicEPM: True, but the new DV codec is extremely precise and you will find it difficult to find any recompression errors in a single pass.

3) Therefore, it is good practice to split events around any effects/transitions, compositing, etc. that you do so that anything that is untouched from the original is not re-rendered. Still true?

SonicEPM: True- this would make your final render go faster, but as above, you'll be hard pressed to see any generaion loss (and please send and example if you do- I'd love to see it). The time it would take to set this up manaully might exceed total render time if you skipped that step, so you'll have to be the judge.

4) However, I have noticed a behavior in VF2 and VV3 that (hopefully) contradicts my above assumption that rendering is all-or-nothing by event. It seems that when I render a project as DV, only the parts that are modified show being rendered frame by frame in the preview window. The rest of it, whether it's part of the same event or not, is just frozen at the last frame of the changed portion. Does this mean that there is some kind of "smart rendering" that detects changes from the source material and only renders as necessary? If this is true than this would imply that the answer to question 3 would be not true.

SonicEPM: If you put a brightness/contrast filter (for example) on a track, with levels set to "0", you will still force a recompression of the entire track. We don't consider an fx/motion/composite keyframe with a zero level to be the same as a pass-thru, since typically there would be interploation across 0, but in very few cases would you hold at 0. Maybe once in awhile, yes, but not typically (or else why add the fx in the first place)

5) If there is no such "smart rendering," is there any chance that this could happen in the future? Isn't it an unnecessarily tedious process to go through an entire project and separate out all the changes by splitting them into their own events if one wants to enjoy faster rendering and higher quality output?

SonicEPM: Intelligent rendering, as used in print-to-tape from the timeline, means that only events that need to be rendered are rendered. It does not mean that only the frames inside of an event that need to be rendered will be rendered. In Vegas 2, the entire project was rendered to a single .avi no matter what, and that is no longer the case. In Vegas 3, a "cut-list" is generated by sniffing, then rendering, events that need to be rendered. This behind the scenes cut-list is played back in sequence during the ptt process. Again, the codec keeps any generation loss to a minimum, if not zero.

If you are really concerned about render times, you could theoretically create new tracks and events for areas you knew would not need recompression- difficult to do (for you, or us).

6) Finally, I have been led to believe that the audio portion of DV is completely separate from the video portion. So that non-destructive edits to the audio portion in VV, or destructive edits accomplished through Sound Forge), will not affect the way video is rendered. True?

SonicEPM: True.

-Jeremy