Lispy de-esser

[pascual]

Looking for clues on how to avoid my de-esser from turning the vocals sibilance into more of a lispy sound, which is kind of worse. Often this is the case when I want to tame those harsh frequencies.
My desser of choice is the renaissance de_esser.
Ideas?

[rasmusklump]

I couldn't get satisfying results with the ren de-esser either.
You should try the free techivation de-esser.

https://www.kvraudio.com/news/techivati ... g-in-52307

[trmupstage]

A deesser is basically a single band dynamic EQ. If you are trying to take out some harshness, it may be in the 3Khz range. Deessers usually start around 5khz, though you could lower it. I'd try a dynamic EQ to solve the harshness. On a deesser, don't try to reduce more than 3-4dB or you may find your vocals have a lisp. I use Waves F6 and Soothe for getting rid of harshness that is intermittent. I also use the standard Wave deesser for general deessing. I've also used renaissance de_esser but like the regular one better. You can also try to change how it behaves. I think they normally default to doing a hi shelf for the reduction. Change it to a bandpass and you might be able to reduce more without bad artifacts.
Also, watch what microphone you record vocals with. Really bright mics are going to add sibillance that is going to be near impossible to get rid of. Less expensive (and even some more expensive) condenser mics have issues with harshness. I'd try a different mic on future recordings and see how that works out.

[fese]

Lispy esses after using a de-esser is usually a sign that you’re overdoing the deessing, too much reduction. Depending on the signal and how sharp/harsh the sibilance really is, one de-esser might not do the job. Sometimes only a combination of clip gain/volume automation, dynamic equalizers and potentially multiple de-essers (e.g. one before and one after compression) is needed. Sometimes a tape plugin can help.

I’ve actually achieved some of the best results not with a dedicated de-esser plugin but with the free FirComp (Sidechain to “de-ess”, ratio to 10, knee to hard, fast attack and release times). Surprisingly hard to make vocals lispy with that.

[Ploki]

A deesser is basically a single band dynamic EQ. If you are trying to take out some harshness, it may be in the 3Khz range.

Basic de-esser is more like a compressor with a sidechain bandpass on 3-6k

To the op:
Don’t do splitband de-essing on single vocal tracks, its pointless and sounds artificial, and sounds lispy way quicker than broadband de-essing.

Use a compressor with sidechain EQ or a good wideband deesser like Fab Pro-DS

[Funkybot's Evil Twin]

Agree that you’re probably over doing it. Also, different de-essers are just better or worse at the task. I suggest checking out the new Techivation de-esser that’s free and trying that one out. Or checking out other options. But remember: you’re supposed to hear the esses, just make them less harsh.

Other option: do it manually with automation in your DAW. Manually de-essing a vocal with clip gain automation and/or splits gives you the most control.

If you’re doing it with clip automation that’s pre-fx you can also combine with a de-esser working less hard. The idea is you can start with light de-essing but if there’s some really offensive s sounds in there, instead of de-essing everything harder, you do some manual pre-fx adjustments and be less heavy handed with the automatic de-essing from the plugin.

[pascual]

Lispy esses after using a de-esser is usually a sign that you’re overdoing the deessing, too much reduction. Depending on the signal and how sharp/harsh the sibilance really is, one de-esser might not do the job. Sometimes only a combination of clip gain/volume automation, dynamic equalizers and potentially multiple de-essers (e.g. one before and one after compression) is needed. Sometimes a tape plugin can help.

Awesome and super helpful answers. And I like the idea of the tape plugin helping to smooth out the harsh tops. I will try going softer on my settings and probably using a couple plugins instead of just one. I'll give the fab filter deessr a spin as well....

Right now my nice condenser crapped out. I'm looking to spend somewhere under 300$ for a replacement. In the mean time, the last vocal recording I did of my female singer was with the sm58, which always works wonders for her actually...

[Mathematics]

You're over-doing it as we all have done. The clip gain automation that FunkyBot mentioned is actually how I start DeEssing serious projects, then use Soothe or RX. No need to repeat what he said...IMHO, it's the definitive way of doing it. Now, what I mentioned so far is if I was doing this for a paid project. For simple YT videos or family video shorts...I just use Soothe or MSpectralDynamics. It's simple and works. These are rather expensive solutions though....so, here's a free solution:

Any free dynamic EQ with notch filters where you need them. Between 4k and 8k. You may need two notches. Done.

non-Free Remedies - Soothe2 (this is my be-all-end-all solution for now), Izotope RX DeEss post-production spectral repair (this one is the most surgical of any solution mentioned by me...I get the best results from this but it can take time), ProAudio DSM V2, Pro-DS, Waves DeEsser, MSpectralDynamics, Kramer Tape, Ozone 8/9 Spectral Shaper...I can get similar results with any of these. With practice and experience, I learned to get great results.

The list goes on...but who wants to read these days?! (Lol, I actually do)