[Anaraxes]

Immediate Preparation Required is the usual choice when you need longer than Takes Extra Time (though you're still talking about a few seconds).

The mismatch is that the limitations is that the preparation is a hindrance because you can't do anything else during that time when you're doing your ritual, or in this case, song. If you could do other things, then it wouldn't be a limitation. In this case, the bard can do other things, but only of a particular kind -- just playing other songs. That's still plenty useful, but it's still a small limitation. I might start with halving the value of Takes Extra Time / (Immediate) Prep Required, at least if the medley sequence is declared in advance or fixed. And I wouldn't let the accumulated time overlap for multiple spells. (That is, imagine casting a 1-second spell, a 2-second-spell, a 3-second spell, and a 4-second spell... it's a bit much to say that takes only four seconds. That's an Enhancement, not a Limitation!)

For the stated goal (meaningful samples of real music), though, I'd just ignore how long the real samples were. It sounds like fun, so just don't insist they be 1-second snippets. Much like you can say more than one second's worth of speech in one turn (as long as you're being reasonable), just let more than one second worth of sample represent the spell. You don't have to act out your turn in real time.