I can think of dozens of reasons to split a single performance into segments and render them seaprately. Any long project is MUCH easier to deal with if you do it in segments -- hence the nested VEG capability in Vegas 6. Unfortunately, neither the design of Vegas nor DVDA 3 have completely come to grips with the reality of how people work on long projects (something the programmers themselves surely must understand, since every programming environment for the last decade has dozens of tools designed to let programmers work on small modules and then stitch everything together).

To solve the problems, DVDA 3 must provide a facility to actually stitch MPEG-2 and AC-3 files together prior to preparation. Otherwise the gaps between the VOB files, or the necessity of putting each segment in a different titleset, will result in the navigation problems reported in this thread.

"Are you splitting live performances into individual clips and then "stitching" them back together via music compilations? May I ask why? I mean, if they're from live performances, surely they were recorded in one sitting -- why not just render that performance as one complete video?"

Actually, since it is a CD-to-DVD transfer, the performance is already split into different tracks. Besides, working with individual tracks is much easier whan making music compilation because you don't have to worry about setting the length of accompanying visual (still pictire in my case) - DVDA does it automatically and sets appropriate encoding bit rate with minimizes the overall size of the project while maintaining quality output. I tried authoring music as a regular video track with still images instead of motion video but it doesn't come out as good and the file size is larger than when using cusic compilation option. The other option with working with audio range rather than individual track is making still-image slide show but than the langth of each and every slide has to be set manually, whit the end result - at least in my experience - never being as precise as auto-setup. Plus, going slide-show route prevents you from inserting more than one audio track so it there is a need to compile several recordings at once one has to use external audio editor to stitch them together and then treat them as one big audio file. Again, the issu of manualy setting shapters and transitions becomes a bit problematic and time-consuming at this point.

Regarding the inability to have seamless playback with live recordings, I don't think it is so much a matter of 'stitching engine' itself as much as the way DVDA assigns and handles navigation commands. In other authoring packages such as the old but IMHO still unsurpassed DVDmaestro if one assigns 'go to next chapter' command at the end of previous chapter there will be the same fraction of a second pause, even with standard video material (resulting in a brief image freeze at chapter-transition points). If, however, the 'go to ...' command is ommitted (as it should be because - if there is no command - DVD player automatically advances to the next portion of the track) you get the seamless playback because there is no need for the DVD machine to waste farction of a seccond 'reading' the command assigned. So, ultimately, the crux of the problem, is the way DVDA auto-command scripting is done by the programmer, most likely having to do with 'go to next chapter' command being auto-inserted at the end of each track. So the solution, most likely, is to remove it and have DVD player make natural transition from one track to another. Needless to say, this needs to be done on designers' end since on user's end there is not that much options for controling navigational/playback command assignements.