De-Essing Best Practice

[Keith Wilby]

I usually have buses for my lead vocals and backing vocals. Where best to put a de-essing plug-in, on each individual vocal track or is it possible to de-ess an entire bus to achieve the same effect?

Many thanks.

[Lord Tim]

For any lines that are very obviously leads (including any lead harmony lines that are meant to sound like single vocal parts rather than choir parts) I tend to put the de-esser on the track itself, so it's all independently adjustable. You might find that some vocals need a bit of a different setting so you're not sounding like you have a lisp.

On the other hand, if I have stacked vocal harmonies, I clean them up at the track level first to remove noise between the phrases and tidy up the timing, then send that to a Bvox bus and I'll do all of my de-essing and EQ shaping there. We typically have 30 to 80 layers of harmonies, so treating each independently is asking for either an angry computer or walking into the ocean by the time you're done with that!

Honestly though, whatever works is the correct answer. You actually might find that you have one layer of Bvox that are particularly sibilant - there's nothing wrong with throwing its own de-esser on there to tame it and still having one on your Bvox bus. If it works, it's correct.  

Edited January 19, 2023 by Lord Tim

[Keith Wilby]

44 minutes ago, Lord Tim said:

an angry computer

Thanks Tim. This made me smile 👍

[treesha]

I have cut the lead vocal to make a clip of the section or phrase that needs deesing and just deess that section or sections vs use the deeser on the whole track. For me its just certain places that need it and sometimes they need it differently so I treat them individually. 

[Glenn Stanton]

i like to use RX on client vocals to clean up as part of prep and do some light de-essing. on my own, i will use RX and then clip gain and edits, then use Melodyne to reduce some of the sibilance, and then one or two de-essers in serial to get sss and cha and tsk which may be different sets of frequencies. each stage is generally small and may be on the VOX buss if i have some doubling, or as Tim noted, on the BVOX buss for background vocals. otherwise, it's mainly the primary vocal track. 

[Base 57]

Clip gain envelopes. 

De-essers all suck (out desirable high frequencies from non sibilant portions of the clip).

[jackson white]

goto here is Melodyne for de-essing with focus on the specific bits vs a fixed setting for the clip. 

[Keith Wilby]

14 hours ago, fossile said:

use Melodyne to reduce some of the sibilance

I didn't even know that Melodyne did this, I have much to learn. Thank you.

[Glenn Stanton]

7 hours ago, Keith Wilby said:

I didn't even know that Melodyne did this, I have much to learn. Thank you.

bottom selection of the amplitude tool.

on my own materials, i'll simple select all, then use this tool to drop all sibilance 2db or so (starting point is around -10% then up or down slightly). just enough to start the process without being obvious. on the help page they have some ideas on split tracks to adjust the "balance" of the sibilance --  https://helpcenter.celemony.com/M5/doc/melodyneStudio5/en/M5tour_ToolAmplitude_Fade_Sibilance?env=standAlone

 

[bats brew]

i use a dynamic eq as a de-esser.

so nothing gets 'sucked' out that doesn't NEED to get sucked out.

LOL

[Glenn Stanton]

technically, a de-esser is a dynamic eq 🙂 just one that has been configured for a specific purpose.  and also more like a frequency-limited band compressor... depending on the brand and model...

Edited January 29, 2023 by fossile

[bats brew]

i refer to the waves F6 plug,

a brilliant plug, with multiple uses.

 

[Jim Roseberry]

FWIW, I wouldn't want to blanketly apply de-essing across an entire vocal track... and definitely not across a vocal bus.

You're going to lose too much intelligibility/articulation in the vocal.

Isolate each offending S (or sibilant) as a separate audio clip.

You can then use your favorite De-Essing plugin to reduce the sibilant (individually) for each offending clip.

You can tailor the "reduction" of each sibilant.

This leaves the rest of the vocal completely in-tact. 

 

[Keith Wilby]

2 hours ago, Jim Roseberry said:

Isolate each offending S (or sibilant) as a separate audio clip.

Thank Jim. Having had a play with Melodyne's functionality, I wholeheartedly agree.

[Gswitz]

Dessing live helps the performance. 

I think things work best when the artist performs through the fx chain. 

It is not that hard to set up and it can change your performance some. 

[Keith Wilby]

2 hours ago, Gswitz said:

I think things work best when the artist performs through the fx chain.

Thanks, I think I need to learn how to set that up 👍

[bats brew]

yea, that F6 plug would be perfect for this.

[SVSX]

Besides Melodyne (witch is awesome for this case), i can recommend Debess from Airwindows (works only with 44.1 and 48 kHz samplerate recordings, because it's kind of filtering the high frequency samples out) and Waves Sibilance. These two are as transparent as possible. Not perfect, but very good! 

[Bruno de Souza Lino]

There's also this guy here, which costs you nothing and comes with Cakewalk....

[Keith Wilby]

Thanks @Bruno de Souza Lino, that was actually "excluded" for me for some reason. I'll have a play with it 👍