De-esser plugin - DeBess alternative

Hi guys!!
I’m looking for a De-Esser plugin. Could you recommend one?
Altough, I admire Chris plugins, I’ve tried with no success Airwindows’ DeBess, actually I think its not working (I’ve used the VST and the LV2 from Airwindows Consolidated).
TAP De-esser is kind of outdated but it could work.
What do you think? Is there another De-Esser out there?

Well first… Must be FOSS? Must be free? Must be LInux native (or is a Windows Plugin with Wine/yabridge acceptable?)

1 Like

If you elected to install the Harrison XT plugins during Ardour’s installation, you can use the XT-DS De-Esser plugin by right-clicking on it and selecting “Edit with generic controls…” If you want to make use of the plugin’s GUI, you will need to purchase a license from Harrison.


Hi @GMaq thanks for your reply!! I 've completely forgotten to include this details. FOSS, free and linux native, please. LV2 is my go to format for plugins but VST would be OK.

Hi @GuntherT what a great comment!! I’ll definitely give it a try. NO GUI, I definitely can live with that XD

Airwindows has another de-esser plugin that is called … :drumroll: … deess.
However this is another one which sometimes works great and sometimes not so.
Since most de-essers are essentially specialized and fine tuned compressors one could use a compressor plugin. Sadly, the ACE compressor can’t be tuned that way (other than with some hackish routing). It lacks an high pass filter for the internal sidechain. Fircomp2 is a linux native but closed source (last time i checked it was free-as-in-beer for linux) plugin which is perfectly suited for that kind of task.
Then there is ZAM DynamicEQ which i use as one stage of de-essing (before compression) practically every time.

1 Like

I definitely could listen to the drumroll… sadly didn’t work as expected but I should give it another shot.
I’d be glad to know more about the ZAM EQ set up.

Unfa did a whole video on DeBess… if you haven’t seen it, maybe it could help you correct your DeBess issues (or confirm them):

I use LSP (Multiband Compressor) for de-essing. Bit of a learning curve, but I love the LSP stuff now. I even got it work on Windows so I can go back and forth using the same plugins (necessary because sadly my RME Digiface doesn’t work with Linux).

1 Like

I’ll second the LSP multiband. Does the thing nicely.

1 Like

Hi @ardourwlk. In fact, that video acknowledge me about DeBess. I love how Unfa use x42 simple scope to visualize the effect of the De-Esser but I couldn’t. Ake it work nor hear the change. Perhaps is my mistake and I screwed up things somewhere.

@werner.back @eighty could you give me more details about it?

Homepage LSP

You can also try with a dynamic EQ like ZLequalizer
ZLequalizer homepage

Great plugin !

1 Like

I just tried the DeBess plugin and a neat thing about that is the SenMon fader at the bottom.
If you set it to 1 you’ll hear what’s being removed, so if you do that, set Intensity to 0.3, Sharpness to 0.3 and Depth and Filter to zero you should be hearing mostly just the esses.

The AirWindows video on that plugin suggests that it works best with 44.1 or 48k material, so if you’re working with 88k or above it may not work that well.

Indeed!! Thank you Peder… I’ve tested with Unfa’s De-esser stress test (or something like that) and I could clearly listen to the nasty “sss” being removed.

Some alternatives so far:
Airwindows-> DeBess
Harrison → XT-DS De-Esser

LSP → Multiband Compressor
Zam → Dynamic EQ
ZL-Audio → ZL Equalizer

Any other thoughts are welcome :grinning:

Do you really need a de-esser though? Is that the best approach for you (in the long term)?

  1. If you record with a good, clean microphone, you will not need a de-esser.
  2. With an SM58 (not my choice, not my recording) I have found using static EQ to correct for its frequency response curve gave acceptable (almost pleasant :slight_smile: ) results.
  3. With other microphones, that were really prone to essing, I was unable to achieve a good result with EQ or with de-essing.

But my expertise and time was limited. So I now try to avoid having to de-ess by sticking to recording with good microphones.

Hi @HuBandiT
That’s a fair question… I agree with you about being more careful when recording in order to avoid having to de-ess. Unfortunately, re-record it is not an option here so :-1: Nevertheless, I must assure that a De-Esser is a valuable and powerful tool to use when needed.

Well, yes and no. Getting the best out of the recording during … the actual recording is the path you should go, sure (GIRATS, as Joe Gilder would say).
For my own productions I can say: as soon as I put a compressor on the vocals, the sibilants usually become “amplified” and therefore I need a de-esser (still using the Calf one) to “work around” it. Just statically cutting down the highs with an EQ usually won’t work for me, because then the vocals would lose some clarity/brightness…

PS / fun fact: for recording the voice of my band singer, the SM57 just beats all other microphones that we have by an order of magnitude (SM58, OD505, MXL990, CM25 MkIII).

I agree about statically cutting down with an EQ @slash. Maybe a dynamic eq or a multiband compressor is a better approach but every music production is different so it’s better to have multiple tools.
Just to clarify, GIRATS means Get It Right At The Source and I think it’s the ideal situation to pursue.
As you already know, there is a problem with CALF plugins and their GUI (something related to GTK version). Have you consider another FOSS alternative?

Yup, heard of it. :wink:
And no, so far I’m sitting it out, until KDE neon (the distribution I use) eventually does a rebase to Ubuntu 24.04, before I’m deciding for myself, what alternative to use…

Curious, did you pick up the Plasma 6 update? If so how did that go for you? Specifically are Ardour and plugins aside from Calf (especially LSP) OK?

I’m on Ubuntu Studio for my daily driver but one of my other machines is running Neon … I’ve been reluctant to take the latest updates to Plasma 6, at least until 6.1 is released.