One technique to consider for dialogs is just using multiple portraits over multiple events
For example, let's say you are having an actual dialog between two characters. You would have an image of each set up for a theatric conversation - by this I mean character A would be facing the front left and character B would be facing the front right. You would show the appropriate picture of each when their dialog is going on.
For monologs, you use a similar strategy of two portraits of the same character that you cut back and forth between as the various points of the dialog are reached.
For both of these techniques, a text event is used for each time a picture shows. This means you don't have to worry about timing like you would with an animation.
I wish I could say I thought of this, but hans used both of these in various demos he has done for FRUA. I think they looked really good.
