I just tested this workflow by recording some narration in PowerPoint, then saving the deck and importing into Captivate.
I have to admit that I was totally shocked at the low quality of the audio PowerPoint gave me. It came into Captivate without issue when I created the project. But when I listened to it in the Captivate preview, it was full of static. I wondered if something about the import process was at fault. So I returned to PPT and previewed. There was static there too!
I returned to Captivate and recorded. HUGE difference in quality! My voice was crystal clear when recorded using Captivate.
Note that I'm not a Power User of PPT, so there may be some settings I failed to change because I didn't know about them. If they exist, they might have allowed me to achieve a different result.
Just thought I'd toss out my own findings. Hopefully others will chime in to offer their own experiences too!
Helpful and Handy Links
If the PPT audio is giving you fits, there's always the option of letting the SME record in a third-party program like Audacity, which gives you extensive control over audio quality and is easy to import into Captivate. Audacity is also open source, so there aren't massive licensing hassles with multiple installs.
In my own adventures in audio recording, I've notice that environment makes huge difference. A real studio is obviously the best (I assume; I've never actually had one available.), but there's also huge differences between one meeting room and another because of racket from outside traffic, heating and cooling systems, induced "buzz" from flourescent lights, etc. "You never realize how noisy the world is until you try to record."
It will pay to do some scouting around yourself to find the best location before bringing your SME's in for a final "take."