Dipping audio when someone is talking

11 replies [Last post]
tomtailford
Offline
Joined: Oct 17 2007

My commentary audio is slowly getting better... which is good, as I have realised how decent sound improves the production level on a whole =D

I want to render out a .mov file with 4 audio channels.

Channels 1 + 2 would be commentary
Channels 3 + 4 would be ambience + additional microphones for MCs

I would like to dip the audio in 3+4 if channels 1+2 are talking - is this possible? Or will this make my sound really dodgy?

I've noticed parts where the commentators are talking (ch 1+2), and the music is blasting (ch 3+4) and it's hard to make out what they're saying.

I 'm only asking this as I remember when trying to DJ at school the equipment was able to do this when talking into the microphone... so surely there is a way of doing this in Adobe Audition?

Z Cheema
Offline
Joined: Nov 17 2003

Not sure about Adobe, but it is sometimes called (audio) Ducking

In DJ equipment it is done electronically using the mic input as a bias for the music output.

In Sony Vegas there is a script that looks at the channel were the commentary is on and then "Ducks" the assigned music track by a preset amount with options of fade in/out durations.

You might be able to link (or send) audio plug-in chains to what is done electronically, by adjusting the volume depending on the input from another audio plugin

http://forums.adobe.com/thread/792215?tstart=0

MAGLINK
Offline
Joined: Mar 8 2007

You would usually dub such content so you end up with a master track of how you wish it to be, I presume you are summing all four channels to get this final mix so why not spend the time doing it properly rather than just slapping some plug in on it and hoping that it ends up OK.

It may also be that you need the sound fx and MC on a sep track so that people can listen without the commentary but this is getting into advanced dubbing and you would not normally just dip the second pair of tracks as if you are listening to them alone the dip would be out of context. This is into M+E (music and FX) mixing and you would have a mixed commentary with dipped fx on tracks 1+2 and then an un-dipped FX only on tracks 3+4.

There are ways of doing this using compressors and I have done lots of live sport where you use multiple compressors to make sure that the commentary is always heard above the crowd fx, I spend lots of time setting all this up to make sure that the levels are manageable and it is also possible to do the same in post but you really need to know a lot about how compression works.

You can also set up something called side chain compression where the sound fx compressor is driven by the commentary and it makes the levels dip when someone speaks.
I am not an expert on adobe but this plug in may assist: http://www.digifreq.com/digifreq/download.asp?ID=86
Also do a search for adobe side chain compression and read up more about it.

All quite advanced stuff and just hoping that something will do it auto style may not get good results so it may be better at your level to do it more manually in your edit system adjusting the levels where required.

One thing you could try though is applying a compressor plug in to each of the pairs of tracks and adjusting the threshold on them so that you always end up with the commentary above the fx etc.

Huge amounts of compression are sometimes applied to live audio and at times there may be only a 2-4db dynamic range, but that is the style of the audio that is wanted so that you get the effect of the commentators pushing through the crowd fx, the commentators also like this as it gives them something to work against.

As a final note getting a proper sound dub done will also help but it may be that you don't have the budget to do this but in doing it within an edit system you need to be careful not to make your audio tracks worse than they already are.

tomtailford
Offline
Joined: Oct 17 2007

Thanks guys,

I've looked into the sidechain compression and just paid for the VST plugin, it looks good!

Budget is a key issue, but sound is so important - I currently leave a field recorder unattended and do a sound check at the start of the event... which isn't the best set up in the world.

Maybe I need to explain the importance of sound to my customers as it really does make a difference.

Thanks again for all the information, very helpful!

Tom

MAGLINK
Offline
Joined: Mar 8 2007

Hi Tom I think you do right to get that plug in and look at side chain compression to see if it can improve things for you, as you already know the type of location recording you are doing can be very challenging and sorting it out in post may be the best option for the moment as you can manipulate things better than at the time on location.

look at compression ratio's of around 10-1 and have a play with the levels within the plug in to find the best balance between the main commentary and the dipped fx.

When doing this live I use a 1k line up tone to set the compressors up so that no matter what I do mix wise the commentator is always around 4db's above the crowd fx and that the overall levels never go over broadcast peaks.

Hard to describe it in words as I would also be mixing things as well but the side chain way in post is a possible quicker way to achieve the same thing in a more automatic mode.

Good luck and more reading here: http://en.wikipedia.org/wiki/Dynamic_range_compression

Rob James
Offline
Joined: Jun 26 2001

All this, while it can work, is no substitute for using your finger(s) on faders. If that isn't an option then "rubber band" gain adjustments are still preferable to automatic ducking. If manual ducking is done well, the audience will be completely unaware that any interference has taken place. Leaving it to automatics risks "pumping" and other unnatural artefacts.
I know I'm preaching the ideal but it has the virtue of cheapness (no plug-ins required) and practicing it will help you understand about intelligibility and sound mixing.

Rob The picture is only there to keep the sound in sync

Ben Longden
Offline
Joined: Jun 10 2009

In Vegas you can eaily create "duck Audio", so the voiceover ducks under the music track.

As an insurance, have a second person come into the cut suite and have a listen. One of the biggest sources of complaints in the industry is dodgy audio mixing where the voice gets muddied with the music.

MAGLINK
Offline
Joined: Mar 8 2007
Rob James wrote:
All this, while it can work, is no substitute for using your finger(s) on faders. If that isn't an option then "rubber band" gain adjustments are still preferable to automatic ducking. If manual ducking is done well, the audience will be completely unaware that any interference has taken place. Leaving it to automatics risks "pumping" and other unnatural artefacts.
I know I'm preaching the ideal but it has the virtue of cheapness (no plug-ins required) and practicing it will help you understand about intelligibility and sound mixing.

Steady on Rob you might encourage people to employ dubbing mixers!:D

Potentially non PC in these days of self shoot video monkeys and FCP X has an auto audio fix, I am sure it will give an auto mix and dub function in the very near future!

Rob James
Offline
Joined: Jun 26 2001

I have nothing against algorithmic mixng when conceived and executed properly. However, just like auto-grading it is caught out easily and is still no substitute for skill.
If you don't have the skill or the time or the money to employ somebody with the skill then fine. Just be aware of the compromises you are making.

Rob The picture is only there to keep the sound in sync

MAGLINK
Offline
Joined: Mar 8 2007

I am a sinner myself when self shooting as I always have my panasonic P2 camera's switched to auto with the limiters on, but as you may guess I make sure my gain structure does not break too many rules and wreck things beyond repair in the dub!;)

Rob James
Offline
Joined: Jun 26 2001
Gary Nattrass wrote:
I am a sinner myself when self shooting as I always have my panasonic P2 camera's switched to auto with the limiters on, but as you may guess I make sure my gain structure does not break too many rules and wreck things beyond repair in the dub!;)

I'm really not a purist about this. If smarts help you to get shots which would otherwise be missed or unusable technically then why not use them?
I've always used limiters as insurance against the unexpected when shooting and also to tidy up annoying peak overs in post but I very rarely use compression and certainly not when shooting. If I do use it, then it is never more than 2:1 with very gentle settings.

Rob The picture is only there to keep the sound in sync

dominicwitherow
Offline
Joined: Apr 2 2006

I have always found that using the Channel EQ in Soundtrack Pro is an excellent tool for ducking. That and the rubber band always seem to do the best job for me. I'm sure there's a simple equivalent in every track laying application (probably called 'Channel EQ' too).