Better to de-ess before or after other compression/limiting?

I de-ess first: makes sense to me.

I do it when it’s an issue....on me , post compression.

Cheers

Wiz

Thanks guys. Is it usually OK to roll off a couple 2db, around 12Khz, to reduce the degree of de-essing? My main "offending frequency", is 6.5 KHz, secondary is 12KHz AKA "the whistle".

Even on an Oktava 219/319, I'm ready to reach out and de-ess! BTW a good buddy of mine,
wants to occasionally record me, so hopefully I'll create some pro sounding clips, within a month or so.
Chris

P.S. Your comment was of value too Wiz, but my voice is unusually bright, by nature.

First fix, second shape.

I de-ess using brainworx dyn eq, then compress. If compressor tames tone means I went too far compressing. May be a little broad stroke eq after but not corrective

I usually de-ess first - but sometimes a second de-esser after compression, if needed - with both instances being used sparingly.

I'm often de-essing early in the vocal chain.

Then, if the vocal goes to a group (like I often have them do) I'll de-ess near the end of the group as well.

I find different amounts in different parts of the signal chain do a better job with my particular mix methods than just doing it all in one instance.

I only use a de-esser if it’s an unruly vocal throughout. Seldom use one on my own recordings because I try very hard to get good takes to print. If I do need to hit a few spots here or there, I’ll cut the wave file on either side of the offensive spot and pull the volume down about 3 dB .

I’ve yet to find one better than the Eloise in split mode for what it’s worth if heavy de-essing is needed.

So to answer your direct question, de-essing after everything as a last resort if their is no other option is my theory.

I know. Although my primary sibilance issue is 6.5KHz... I was amazed to realize around 12KHz, could also be one on a male singer-namely me! Chris

Since you asked what I do....sibilance is a failure of the tradking engineer. Thus I “fix it” at the same time i remove plosives and rumbles and extraneous noises from tracks. BEFORE I mix.

But, the thing is...so often that failure is poor vocal compressor settings....so, if YOU are tracking engineer who failed with whatever compressor during tracking, theres no need to fix it becuase youll likely fail again with the mix context compression. Put it last on the vocal chain so you can keep playing whack a mole as you keep manipulating it.

Nothing wrong with that....the only way to get better is to admit "man, I F'd that up.....how do I do it better?"

....if people will stop printing compression who think deEssers are mandatory/standard issue for vocal production....work with the ITB compressors--and don't start with the typical 76/LAxa....they are huge offenders if you need a certain compression level. Once you wrap your ear around how to adjust a full control VCA style compressor....you can start to use the "old school vibe" ones for the vibe and HEAR when they start to pooch things...one of my compressors on a vocal is nearly always a clean VCA style (like comes with your DAW). The other usually isn't. Whether I'm doing the slow and low leveling with an analog Aphex or a FabFilter ProC2 or Cubase or Logic's built in comps (in Logic I do the Focusrite Red looking variation)... less relevant....whether the vibey limiting from from my old La3a, a BlueStripe 76, or Fairchild 660 plug in....that's all contextual. Aphex661 leveling on the way in and the 660 is kind my fave....but, I also can't deny that the La3a limiting on the way in (it's a slow limiter fwiw) and a slow and low ProC2 makes for a delicious sound, too.

You can also bump the highs in the detector of most modern comps....so, using it like Clearmountian does his SSL compressors....or like the La3a does right out of the box--tilting the sidechain toward higher frequencies. Which is far more natural sounding--but, it's not a deEsser--meaning it doesn't CURE a problem that exists--it just helps the compressor not make that problem worse or cause a new one.

Usually first in line, if needed.

By the way, I want to encourage y'all to try the Hornet Sybilla deesser. I think it's incredible! It costs less than 8 bucks, I would gladly pay 50:
www.hornetplugins.com/plugins/hornet-sybilla/

I'm also a big fan of Sonnox Dynamic EQ. That thing is a magical problem solver.