What are the Best vocal settings for Multiband Compreesion

I have vocals that sounds very good and even.

Plain narrative male vocals. No music.

I just don’t want the vocals to sound raw.

I want to make the vocals sound a little smoother, but not too much compression.

I have been trying c3 multiband compressor.

What are the best compression settings for the vocals without making them sound muddy.

I have been trying to compress the high band only. About 2800 hertz upward.

Seems to help a little without sounding muddy.

Sounds like you are de-essing , which can be done using the upper band of a multi-band compressor, or via a de-essing plug-in like Spitfish or LISP.

2800Hz is unusually low for de-essing, (i.e. may remove too much high-frequency causing dull muddy result).
In my experience the lowest threshold is 4KHz , and typically 6-8KHz.
It depends on the performer/mic/ etc, we’d have to hear a short unprocessed sample which included sibilance to be more specific.

You might be right about the de-essing.

Here is a audio sample:

IMO the main problems are …
a hint of overload distortion (mic a bit too close to mouth)
the occasional click, (e.g. on “T” of “the basis”).
Noise level could be reduced a bit.

There isn’t excessive sibilance on your sample, quite the reverse : a big notch in the spectrum at 5KHz has reduced the high frequency content. De-essers only apply notches like that momentarily when sibilance gets above a threshold, whereas applying such a notch with an equalizer removes high frequency from the entire track, which can make everything apart from sibilance sound muddy.

Okay. How do I fix the effects of the notch filter?

To undo a notch apply equalization with peak where the notch is ( @ 4500Hz , not 5000Hz as I said in my previous post :blush: ) …
equalization to (partially) correct the notch @4500Hz.gif
Notch @ Audacity frequency analysis showing notch @ 4500Hz partially corrected on ''after''.gif