This must have happened a dozen times: a client asks me to sub-title a documentary which has so much voice-over commentary I feel obliged to tell them they need to re-voice-over the commentary part [and sub-title only the interviews, pieces to camera, signage. etc]. This is standard process.

However, the client seldom manages to come up with the un-mixed multi track copy, yet gets the producer's permission to go ahead.

So-oo, I'm just wondering whether anybody has had any luck trying to erase the voice alone from mixed audio which might have a background of sync sound, maybe music and atmos,. [No, the budgets these clients propose do not extend to the painstaking re-creation of all audio tracks].

Since I'll be laying down another language voice-over, largely on top of where the original voice-over existed, I'm thinking maybe I could get away with reducing the original voice to a mere murmur and the new voice might cover up the old one.