To duck or not to duck
Posted: January 5th, 2010 | Author: andy | Filed under: post production | 4 Comments »I read recently about a technique to make a voiceover sound distinct from an underlying music bed without ducking. Ducking has traditionally been used – this is where the music bed is faded down as the voice speaks, then (un)fades back up as soon as the voice stops.
The non-ducking technique involves removing the major frequencies in the voice from the music bed, this ensures that the main frequencies in the voice are not fighting against the same frequencies in the music. Clever!
I thought I’d experiment and I’ve made 3 recordings below. The first is just my voice competing with the music, no ducking or anything – a bit of a din. The second recording utilises traditional ducking. The third uses this new technique, I like to call it frequency accommodation. My voice has prominent frequencies between 200-500Hz and again at about 5kHz, so I reduced these frequencies from the music. The differences are subtle. What do you think?
Latest word count: approx. 17,930. Yes, I know it hasn’t gone up much – I’ve been tweaking the Planning chapter.
No modification [0:40m]: Play Now | Play in Popup | Download With ducking [0:40m]: Play Now | Play in Popup | Download Using frequency accomodation [0:40m]: Play Now | Play in Popup | Download
Interesting technique. The first (straight mixing) is indeed a bit hard on the ears. Ducking is ok but it always sounds odd to me when a music track fades down/up around words.
Your Frequency Accommodation idea is good – it sounds very natural and the music doesn’t go too quiet.
How’d you do it? Are you covering it in the book?
Ian.
Thanks Ian. Yes, I’ll definitely cover it in the book. First you do a frequency analysis of the voiceover. Your’re discovering the prominent frequencies in the voice. You then lessen these same frequencies in the music bed – not too much, just a bit so the music doesn’t sound too altered. Then, when combined, those prominent frequencies in the voice don’t have to compete with the same frequencies in the music bed. It’s as if the music has accommodated the voice. Clear as mud?
Andy, I do like this approach. My concern is when you finish talking my mind is expecting the music to peak. It must be years of conditioning!
So all sounds great until your talking pauses, then it feels like your sound engineer was napping and missed the fade up.
However, that might just be this one sample and people might be more tolerant of the duck free zone when listening to podcasts
I can’t wait to play with it myself though as I am sure there is a happy level there to be found.
It may be that your approach combined with ducking might allow for subtle ducking, just enough to give the voice authority.
Steve
Steve… or ducking just the competing frequencies – now there’s an idea…