I click on the microphone and say "R", it then jumps to the playlists that start with R.

I like that idea very much... but you would have to take it a step further and have it recognize certain keywords as well.

Just saying "R" wouldn't help the empeg unless it "knew" what you had in mind: Do you want to search? Do you want to change the volume? Do you want to skip to next track? Do you want to Tweak Order? And you would need a "master keyword" to turn on VR. Otherwise the first time someone says "Hey, how are you doing?" your playlist would jump to the "R"s. :-)

So, maybe with a limited VR capability you would say:

VR (turns on voice recognition mode)
SEARCH (Turns on Search mode)
INSERT (Sets to Insert rather than Append or Replace)
ARTIST (Sets search tree to Artist)
R (Goes to "R" in the artist tree)
U (Narrows down the search...)
S (By now it has probably found "Rush")

etc.

DO IT (Executes the search and turns off the VR mode.)

In other words, each voice command would mimic a single remote keypress. Hmmm... would it be a problem that the letters m,n,o (for example) are really "6" on the remote?

Y'know, this could work. There would probably have to be about 50 key words, chosen to not sound like single letters of the alphabet, and you would want to avoid anybody saying "Hi, how are you" while you are spelling things... but I like it.

It doesn't fulfill my dream of being able to say "Empeg. Play. Rush. Red Barchetta." and have Geddy Lee start singing about his Uncle's Ferrari... but it would be a lot better than nothing at all.

tanstaafl.