The synchronisation issue is likely because Captivate was still getting things sorted out at runtime by the time you had specified the voiceover audio should start.
It's generally regarded as good practice to NOT put voiceover that requires synchronisation on the very first slide. I usually have an initial blank slide of a second or two duration to allow Captivate to get set, then I have a title slide with a little intro music that goes for about 7-9 seconds where I display info about the name of the course, department, module, duration etc. Then on the third slide I start my voiceover and CC text. This seems to work well and the users are getting all the info they need without synchronisation issues.
