In the video, from the time you answered the call, until the voice started, it took 5 seconds. If that's normal for the system, I'd be worried my clients would hang up before hearing appointment reminder.
back
1 comments
So just between us geeks: Twilio retrieves your MP3 files when the call starts (and caches them), but my Capistrano tasks effectively bust caches on deployment, and since doing that video took a couple of takes with redeploys in the middle, the shot you saw was live with a cold cache. Actual users will almost certainly have a warm cache for that default script, and the voice will start playing almost immediately. (That said, I agree, if I had a team to do video for me it would have been more polished.)
I've been doing a demo video last week, and you can get a long way with powerpoint, windows movie maker, camstudio and audacity. I found I got significantly better results by writing my script, doing the voice over, then the screencast, import all into movie maker, and then edit the screencasts to make it all fit.
('better' when compared to the 'do it all in one go' approach)