A friend (Janet) asked about "...a talking headshot..." for web use.
This is a simple process. We'd need, in no particular order:
An actor - with a short piece committed to memory;
A location;
A camera, some lights, and a microphone.
First, we set up some nice flattering lighting, hopefully relevant to the material being performed. Then, a little bit of camera and sound check* to make certain things look right and can be heard clearly. Maybe a very little bit of blocking practice if there's any walk-in involved. If not, record.
Then we sit down to edit, and encode it to whatever format the host wants. Ask for specs.
Then, push it up to the desired server for storage and playback.
* Adobe On Location makes this very simple, and will record direct to hard drive in the bargain. Do I really have to tell you why this is a good idea?
(Disclosure:
I am not compensated by Adobe in any way. Their program just makes my life easier.)