multiband compression at 1k-3k Hz range

 

You are right! Four band parallel compressor/expander and saturator. Not just the highs but the highs are usually the most desired band(s).

 

Last edited: May 20, 2019

Some cool techniques posted here!! I pretty much de-ess then eq the air back in then repeat if required. Also use the c4 as well (getting into fabfilter mb though).

 

Buy Weiss De-ess.. Easily the best de-esser or get Soothe. Another good one is Dynamic Spectrum Mapper v2. There are quite a few de-ess presets in the DSPv2 that are great.

 

No idea, send it over I'll do it, and if you like it, I'll tell you how I've done it.

 

You want a de-esser that compresses ALL frequencies instead of compressing selected frequencies

 

That's what a de-esser does.
A multiband compressor on the other hand compresses just certain bands.

Usually the shh/esses are alone (except for the occasional Z sound which is voiced, unlike the S sound).

 

Disregarding de-essing for a moment, you might consider using a non-directional (= omnidirectional) condenser microphone. Keep minimum distance of about 10". In effect this will greatly reduce proximity effect/sibilance, resulting in a more "flat" response, thus less emphasis on (lows and) esses. A quiet and "dry" surrounding (that is, a room isolated from external noise and dampened against reflections) can however prove essential depending on your situation and how "open" you intend you arrangement/mix to be. I normally use a classic tube mic, switchable between unidirectional <-> omni, with very good results, in omni mode, towards that non-essy clarity. Since increased omni-ambience can come in as a trade-off, post-tracking I sometimes apply a compander to quieten the silence between phrases. Where you choose to de-ess, lastly, you will want to experiment with various settings on a variety of de-essers (frequency, width, attenuation), to accommodate for your specific microphone choice, its placement and application; especially with unidirectional microphones a singer's "technique", or lack of it in a way sometimes, can diminish or increase sibilance on the pure, unprocessed, microphone response. Generally, I would suggest to only de-ess post-tracking, before compression and EQ - although no fixed rule will apply here.

>

 

Last edited: May 22, 2019

This is the PROPER order:
1) De-Esser. Try several of them and find the one that fits your needs. I would recommend one that lets you sweep the freq band to tune it as good as possible to the particular "s" and "sh" components of THAT voice. And YES... you should recalibrate for any new voice. Make a preset for each voice you want to process.
2) Compress
3) Add "air"/EQ and/or harmonic exciter on top to compensate for that sensation of "lost highs".
4) Limit

2) and 3) may be exchanged. Just a matter of taste.

Tips:
- A De-Esser acts taking the level of the processed freq band in to account. If you manipulate the levels with a compressor or limiter BEFORE the De-Esser it won't work optimally. In fact, ANY processing that relies on levels must be done BEFORE a compressor, just because a compressor alters the levels! Pure logic.
- Also, if you are gonna use a gate (or, even better, a downward expander) on a voice, ALWAYS put it FIRST in the chain. ANY processor that "cleans" the audio must be as close to the beginning of the chain as possible. In short: First clean, then process.

 

I've been one of his patrons for a few years now. Chris creates incredibly useful audio tool gadgets. Five bucks a month has been super doable to support his useful work. It's a no brainer.