If you are recording the VO part, why would you need to use time compression?
I would record the vocal speaking part without any reverb or compression.... in other words dry.
Then I would use just the slightest amount of reverb to give it a bit of space. Just enough compression to level out the lows and highs a bit for a very consistant vocal level. Assuming you recorded it correctly, you should not need much. Since the vox is the main thing here...get it sounding good without the BGM playing.
EQ.... for spoken word, you obviously want the words to be clear and articulate. I'd cut the lows below 80 to 100 hz, and add or cut the mids and highs according to the needs for the voice. Each voice is totally different and what works for my voice might not work for yours. Cutting the lows as I mentioned is a pretty safe bet for most voices. The rest will depend on the voice.
As far as levels between BGM ( music) and the voice.... listen closely to what other radio spots do. Use envelopes on the volume for the tracks.... I'm assuming you have maybe one music track and you are recording one track for voice and maybe another for the sound FX.
Be careful with the special sound FX... do not over do it. Same goes for the voice fx and reverb. most spots have a clear, clean voice. FX can easily distract the listener from the message, so be careful in that area.
Envelope each track for volume. set the starting levels. When the speaker is speaking, pull the music level down so the voice is clear. I would think you would set the voice as the main aspect of this project. Get it sounding right (eq/comp) and set it's level. Aside from slight enveloping it will probably stay the same for the entire spot. only envelope the noise from the vox track.
The music and FX will be enveloped up and down as needed, around the vox. You do not want the BGM to sound like someone is riding the faders however. It should be at a smooth natural BGM level.
I hope this helps..... if it was my project, that's how I would approach it.
It sounds like the client might be a bit picky... trying to play engineer after the fact... voice too loud or not loud enough...... so I'd set it where it was clearly heard, and every word was understandable. If the client then commented the voice was too loud, I would reply in a friendly manner that "the message is the important thing....you are playing for the radio/TV time, and you need to have your message heard clearly, your goal is not to pay hundreds of dollars for playing music but to get the message to your customers" then let them decide.....another option........ You could also make several copies with varying vocal levels if you think that would be more to the customer's liking. Let them choose the one they like best.
post edited by Guitarhacker - 2010/05/17 08:47:02