ssssssssssssss n de-esser

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
matthewisgrand
KVRer
3 posts since 23 Feb, 2015

Post Tue Jul 16, 2019 9:26 pm

Hi I am processing some vocal for a friend n the sss sound is extended in his voice when processed
I tried using a de-esser n he said it sounds like he has a lisp
I took the de-esser off and he said its just the same
any ideas?
is it his end or my fault? he seems to think its not in his vocal

imrae
KVRian
740 posts since 2 Jul, 2010

Re: ssssssssssssss n de-esser

Post Tue Jul 16, 2019 11:11 pm

Use the de-esser with more moderate settings.

User avatar
an-electric-heart
KVRAF
1812 posts since 13 Jun, 2008 from Napier,New Zealand

Re: ssssssssssssss n de-esser

Post Tue Jul 16, 2019 11:32 pm

Yes, what imrae said. You don't just slap a de-esser on, you need to spend time dialling it in, too much compression and it'll get rid of the "ssss" all together, you don't want that, you're just trying to turn it down a bit.
What plugin are you using? Some do a better job than others of visually showing what's happening, I quite like the T-Racks one for that very reason.

User avatar
Unaspected
KVRian
797 posts since 4 May, 2012

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 1:49 am

Depending on the signal, you might require more than one de-esser with very narrow bands or to side-chain an EQ with the required peak filters to remove the sibilance.

If side-chaining, don't forget to place another filter in the side-chain so it only passes those high frequencies that you want to suppress - otherwise you'll have the filter throughout the duration of the signal, rather than just on the sibilance. A band pass filter will likely be better than a high pass for this job - but is always best to experiment.

sjm
KVRAF
1886 posts since 17 Apr, 2004

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 2:13 am

Try automating the volume pre-compressor. These sounds are all focussed in the higher frequencies, whereas a standard compressor reacts to the entire frequency spectrum. That means that these sounds are not affected as much by the compressor as most of the rest of the singing with a fuller frequency spectrum- and it's this that makes them stick out after compression. The other bits got quieter, these bits stayed more or less the same volume.

If you automate the volume of the audio before the compressor, you can make the compressor's job easier, and that tends to reduce the issue quite a bit. The issues also tend to get exaggerated if your compression is over the top. So make sure you aren't compressing the vocals to hell and back. You want the vocals to be roughly the same volume throughout, which means taming the peaks and generally levelling everything. You don't need to squash the entire take. Two compressors in series is my standard approach to vox. One for reducing the peaks and some levelling, and one on a more gentle setting that simply smooths everything out. The first one will generally have more GR and jump about a fair bit; the second one has less GR and is more consistent.

As well as a de-esser with the right settings, I also like using dynamic EQ on vocals. Dynamic EQ also lets you target more than the sibilance. In my case, I find it a great way to reduce the boom and boxiness at the lower end of the spectrum too. I have a deep voice, and this allows me to take the edge of the lower frequencies too when I got more excited during the performance.

I find a combination of volume automation, de-essing and dynamic EQ normally works. I've never got the results I wanted using just a de-esser on its own.

This takes quite a bit of work, so don't expect immediate results. You'll get a lot quicker at automating the volumes once you've learnt to recognise the waveforms of different sounds, but even then, it's a thankless task. The end results mean it's generally worth it though. Of the de-essers I have tried, I've had the best results with the T-Racks one mentioned above. You can also use a multiband compressor and just use 1/2 bands on the area where the sibilance is and leave the rest untouched.


Btw, you can check whether it's the take itself by disabling all FX. If he's still lisping, then it's in the take; but it's almost definitely your processing chain.
Voted KVR's resident drunk Robert Smith impersonator (thanks Frantz!)

https://soundcloud.com/steevm/

imrae
KVRian
740 posts since 2 Jul, 2010

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 8:45 am

Sorry I didn't quite understand the post now I re-read it. If you think the take without de-essing has too much sibilance, and your singer thinks it has too little... you may have a disagreement that can't be fixed by technology. Are you using the same listening environment?

jochicago
KVRian
544 posts since 26 Feb, 2018

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 8:56 am

> the sss sound is extended in his voice when processed
Review your processing. One of the reason most of us collect dozens of compressors is because not everything goes well with anything. Some compressors that make things worse, others make things better. Reviewing your FX chain and start turning things off then try to replace the offenders with different plugins.

> I tried using a de-esser
Sounds like you are not taking your time with it. Try a few different de-essers but take your time to play with the settings and really hear what's happening. Also as was said, when things are tough it often helps to do 2 (or more) layers of moderate de-essing.

If you are still stuck with this after trying for a while, download and test Tb Sibalance 3 (free demo). That's what I use 99% of the time. The demo is fully functional it just won't remember the settings when you save the project.

Also tell your friend to back off the mic a bit.

User avatar
Unaspected
KVRian
797 posts since 4 May, 2012

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 9:04 am

Another point which might have been missed is the order of processing. You're going to want to clean the audio before any processes which raise the noise floor and then possibly consider cleaning again to correct issues. So de-ess ahead of compression.

I figure that might have something to do with the issue as the vocalist suggested that when you took the de-esser off, the issue was still there.

jochicago
KVRian
544 posts since 26 Feb, 2018

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 10:19 am

Unaspected wrote:
Wed Jul 17, 2019 9:04 am
You're going to want to clean the audio before any processes which raise the noise floor and then possibly consider cleaning again to correct issues. So de-ess ahead of compression.
Very good point. In this case probably a de-esser before compressor, then another one after. Gentle on both as to not mangle the audio.

User avatar
donkey tugger
Boss Lovin' DR
5989 posts since 15 Mar, 2002 from the grimness of yorkshire

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 10:45 am

Good suggestions so far.

As per usual, getting the recording right at source is of paramount importance. Most singers like too get to close to the mic, so stick a pop shield in front - not so much to cut any plosives, but to keep the f**ker at a reasonable distance.

As others have mentioned it's rare that any issues can be wholly solved just by sticking one plug-in on and leaving it. You might need to look at the overall picture - is there a compressor (either on the track or on the master) with an attack too short? Do you need to alter the order of the effects, or use two compressors/two de-essers in stead of one..etc. Even after that you might need to go in and automate the level of the de-esser manually for certain syllables - as one does when one gets an bit enthusiaSSSStic with the Liam Gallagher impression.. :scared:

Some good plug-ins mentioned - another one is Air windows DeEss, which works slightly differently Not sure I understand the science malarky, but sometimes it's done a job for me recently where others have not been so effective.

User avatar
Unaspected
KVRian
797 posts since 4 May, 2012

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 10:58 am

donkey tugger wrote:
Wed Jul 17, 2019 10:45 am
Good suggestions so far.

As per usual, getting the recording right at source is of paramount importance. Most singers like too get to close to the mic, so stick a pop shield in front - not so much to cut any plosives, but to keep the f**ker at a reasonable distance.
Boosting the level of the foldback mix also works wonders. I was told about that and was amazed how well it worked the first time I employed it. I was making the vocalist move his head backwards and forwards from the mic in direct proportion to the pot between my fingers. Fantastic stuff. Won't help with headphone bleed though.

donkey tugger wrote:
Wed Jul 17, 2019 10:45 am
Even after that you might need to go in and automate the level of the de-esser manually for certain syllables - as one does when one gets an bit enthusiaSSSStic with the Liam Gallagher impression.. :scared:
+1 for manual editing.

User avatar
donkey tugger
Boss Lovin' DR
5989 posts since 15 Mar, 2002 from the grimness of yorkshire

Re: ssssssssssssss n de-esser

Post Wed Jul 17, 2019 11:14 am

Unaspected wrote:
Wed Jul 17, 2019 10:58 am
donkey tugger wrote:
Wed Jul 17, 2019 10:45 am
Good suggestions so far.

As per usual, getting the recording right at source is of paramount importance. Most singers like too get to close to the mic, so stick a pop shield in front - not so much to cut any plosives, but to keep the f**ker at a reasonable distance.
Boosting the level of the foldback mix also works wonders. I was told about that and was amazed how well it worked the first time I employed it. I was making the vocalist move his head backwards and forwards from the mic in direct proportion to the pot between my fingers. Fantastic stuff. Won't help with headphone bleed though.

Hehe, remote control - I like it. Being an old rocker bastard I don't generally have nay issue with headphone bleed - belt the bloody things out. Can see how it would be a problem with one of these whispery, breathy female vocalists which are all the rage these days though. "Bit louder please love..." :hihi:

User avatar
Unaspected
KVRian
797 posts since 4 May, 2012

Re: ssssssssssssss n de-esser

Post Thu Jul 18, 2019 8:39 am

donkey tugger wrote:
Wed Jul 17, 2019 11:14 am
Unaspected wrote:
Wed Jul 17, 2019 10:58 am
donkey tugger wrote:
Wed Jul 17, 2019 10:45 am
Good suggestions so far.

As per usual, getting the recording right at source is of paramount importance. Most singers like too get to close to the mic, so stick a pop shield in front - not so much to cut any plosives, but to keep the f**ker at a reasonable distance.
Boosting the level of the foldback mix also works wonders. I was told about that and was amazed how well it worked the first time I employed it. I was making the vocalist move his head backwards and forwards from the mic in direct proportion to the pot between my fingers. Fantastic stuff. Won't help with headphone bleed though.

Hehe, remote control - I like it. Being an old rocker bastard I don't generally have nay issue with headphone bleed - belt the bloody things out. Can see how it would be a problem with one of these whispery, breathy female vocalists which are all the rage these days though. "Bit louder please love..." :hihi:
Hehe. I quite like some of that intimate style. Though this was a stoner rock band and the vocalist was definitely half deaf (judged by the level he needed in his headphones) - technique still worked. A little bleeding never upset anyone.

Return to “Production Techniques”