What's new

Ducking music volume to make way for voice narration - is there a better way than side-chaining ?

pderbidge

Senior Member
Couldn't the same thing be done with multi-band compression, which is built into most DAWs?
Yep, or a dynamic eq like the TDR Nova which has a free version but I think a multi band compressor would work better in this case since a dynamic eq would need some broad eq curves that might be hard to dial in without messing with the audio. Always more than one way to skin a cat but you're right multiband should've been my first suggestion as it makes the most sense for this scenario.
 
Last edited:

Fredeke

Senior Member
Usually, dimming the freqs around 2khz in the music makes room for clarity of speech to come through. That way, you don't need to duck the music so much. (If you have a way of sidechain ducking just that band - I never tried but I'd be curious to know how it goes)
(answering myself ;))
In addition, it could be worth trying reducing the bass and/or low-mids in the speech, so that music and speeck are each prominent in different parts of the spectrum.
 

Jerry Growl

Composing Music in the Plastic Dark Ages
There are of course several reasons why audio engineers in post-production just don't leave it at ducking for mixing VO:
(besides the fact that it's just plain lazy):
  • It's predictable and boring and sounds mechanical
  • Ducking does not relate to the loudness of the music
  • Especially in longer productions this method sounds cheap, inapt and unprofessional
Best results and most natural transitions are usually drawn with a fader (no not your mouse), in a process called fader riding. You don't need an array of 64 motorized faders to achieve this, just any one simple fader controller will do.
 
Last edited:

Nick Batzdorf

Moderator
Moderator
Best results and most natural transitions are usually drawn with a fader (no not your mouse), in a process called fader riding.
What you're saying about ducking is what I wrote above, only you're saying it better and with more feeling. :)

However, after having a HUI for years, as well as a digital mixer that spoke HUI protocol to DAWs... that is, years of having them and never bothering to use them when I was in the throes of an actual project... I had to admit that the mouse is my weapon.

Faders are great, no question. They come from consoles, which are musical instruments for a lot of people. It's amazing watching a great engineer play one, and it was even more amazing back in the pre-DAW days when they would do all kinds of interesting stuff with the SSL automation.

But for me, at my speed, it's mouses.
 

benmrx

Senior Member
FWIW, I do more audio post than composition work and my day job is working at a multi room facility with 3 PTHDX rooms. We do tons of tv, radio, web, film, etc. 99.99% of the time I’m not using any kind of side chain/ducking etc. to duck music. It’s just faster, and more accurate to ride a fader (or use the mouse) because it all depends on the tempo of the VO, the tempo of the music, the length of gaps, what the VO is currently ‘talking about’, lyrics/vocals, music edits, where ‘the beat is’ when the VO starts/stops, SFX, etc.

Plus, usually there’s more than one spot to work on, so you’re attack/release times might work for the 15sec. spot, but not the 30, etc.

And..., just my opinion, but you don’t want the music to duck/raise the same amount and at the same speed each time as it can sound repetitive and predictable.

Sometimes I’ll have ‘heads/tails’ track for the music at the very beginning and end of a spot so I can treat the music during the main body of the spot more drastically.
 
Last edited:
OP
ManicMiner

ManicMiner

in the Skylab landing bay
I did go for a very subtle side-chain dynamic EQ-ing, and I generally EQ'd the music with a node around the 1k mark with a generous Q (see image below). I think 1k is the middle of where the voice frequency sits and the EQ made a nice subtle hole for it.
I'm happy with the result of how it sounds for my particular project. My music is not busy music that needs definition, its a soft warm bed.
I might start to look into learning to fade with the mouse or an external fader to see if I can get even better results.
 

Nick Batzdorf

Moderator
Moderator
I think 1k is the middle of where the voice frequency sits and the EQ made a nice subtle hole for it.
Spoken voice? Usually female vowels are somewhere around A and E below and above middle C, and males are maybe a 5th lower - and that's where you have to be careful under dialog/narration. Consonants/plosives are in the 4K range, and that's where the intelligibility is. (4 - 8K is also where a lot of people, I believe especially men, lose hearing as they age and why older people can have a hard time understanding speech.)

I haven't heard your project, of course, but my guess is that your wide Q caught the area an octave below 1K while removing some harshness in the 1-2K area.
 

Dietz

Space Explorer
Usually female vowels are somewhere around A and E below and above middle C, and males are maybe a 5th lower
Good point - but keep in mind that this is also a socially shaped phenomenon. Especially in an Anglo-american environment it seems to be social expectation that men have markedly virile, deep voices, while women there usually tend to have a decisively higher-pitched fundamental. This difference might be less pronounced in other cultures and languages. (... I read the abstract of a scientific study some time ago which underpinned this thesis, but I don't have the citation at hand, sorry.)

... but I digress. 8-)
 

Fredeke

Senior Member
Good point - but keep in mind that this is also a socially shaped phenomenon. Especially in an Anglo-american environment it seems to be social expectation that men have markedly virile, deep voices, while women there usually tend to have a decisively higher-pitched fundamental. This difference might be less pronounced in other cultures and languages. (... I read the abstract of a scientific study some time ago which underpinned this thesis, but I don't have the citation at hand, sorry.)

... but I digress. 8-)
Sure. Once I lamented than French movies don't telephone-eq dialogs like American ones do, and it makes French dialogs more difficult to fit into a busy soundtrack. (The problem usually arises when dubbing Hollywood films in French, and the worst solution is sometimes taken, which consists in lowering the international track's volume as a whole.) But I was answered by a colleague that that was probably due to a difference in how the languages sound: American English (especially Californian) is more nasal than most French accents, and can probably benefit from narrower EQing.

...but I digress too.
 

Nick Batzdorf

Moderator
Moderator
Good point - but keep in mind that this is also a socially shaped phenomenon. Especially in an Anglo-american environment it seems to be social expectation that men have markedly virile, deep voices, while women there usually tend to have a decisively higher-pitched fundamental. This difference might be less pronounced in other cultures and languages. (... I read the abstract of a scientific study some time ago which underpinned this thesis, but I don't have the citation at hand, sorry.)

... but I digress. 8-)
Oh, I like virile women and higher-pitched men too.

I've actually plonked out where people's voices are on the piano before writing music under them.
 
Top Bottom