I want iTunes to bust in every n tracks and use the “Victoria” speech synthesis voice to say things like “…and that wass Meeles Dahviss with Eena Silent Mood. Right before that we heard…” and so on.

Someone must have written this plugin, surely. If not, here’s a set of scripts that nearly offer the answer; the obvious deployment is to tie it to a key-command and randomize the wrapper script, so that the voice says different things as it announces the trackname and artist. The current implementation also only draws the variables from the current track, and speaks as the track is playing (paused counts as not playing, in this instance).

This has the unfortunate effect of placing the speech right over (or, as in my case) under the audio, making it even harder to understand. I’d hack it up, but I really should be looking at fixing those comments.