Peter,

Your best bet may be to create a slideshow in Mathematica and intersperse your voiceover between each slide (or each group of cells) by creating speaking buttons such as:

Print[Button["Slide 1", Speak["Introduction"]]];

You can then hide all the creator cells leaving only the buttons. Now you can start the slide show and click on the buttons to do your speaking. You can then use a screen capture/record program to capture the entire video with audio and cell evaluations, etc.

If you want it to be more seamless (depending on how much time you will spend doing this) you can create special formats for your slideshow (like have an Author mode and a presentation mode -- in presentation mode it hides the button creators) . You probably can make a program that auto plays the slide show and the speech buttons but that is more complicated.

Regards

Well. You'll have to break the problem down and ask more specific questions. There's a lot to unpack here. You'll want to start with small steps like making the audioless video you want. And then learning how to generate speech from text and then figuring out how to line up the audio with the video and then learning how to add the audio to the video. Etc. Each one of those is a small project to figure out.

Well. The description of what you want is vague. Maybe I or they were confused about what you wanted exactly. Are you looking to do Text-To-Speech? You'll probably want to use someone's API to generate the audio. And then, overlaying audio on top of a video isn't something that the Exporter for AVI or MOV does. You can look at the documentation for AVI and MOV. There are no options for adding audio to an animation that is exported.

You should begin by learning how to make the movies using standard video editing software and then experimenting with how much you can move over to Mathematica. No doubt it's completely automatable even if you have to call some third party utilities. I just don't think it'll be straightforward. You'll have to implement a fair number of features.

I have been told that this can be done.

Who said this? They probably mean that it's possible, which is true. But that doesn't mean it's the best way of doing that.

Most certainly you could use Mathematica for part of this. But I would use video editing software to make the actual voice overs and do actual video editing.