De essing question

M

Mr. Cachi

Member
Hey guys, I have a vocal track that needs some deessing, and the vsts for that are not very effective. I was thinking, would it be convenient to compress the track like I normally would, so its nice and even, and then using automation to lower the volume of the esses? or should I just go ahead and separate the esssing parts to a different track and process them differently?
 
You could manually reduce the sibilance with automation. If you've got a compressor with filters on its detector circuit you can make your own de-esser. Set the high and low pass filters to isolate the worst of the sibilance.
 
That sounds like a reasonable solution. I've never had luck with de-esser plugins, I've always had better results making my own.
 
Maybe it's just the performers I record or the genres in which I work, but I have found that de-essing rarely really works and is actually rarely necessary.

First of all, I think that very often when we get to the point of trying to de-ess a track, it's because we've lost perspective. We've become hyperfocused on that one aspect of the mix and get to the point where we almost can't hear anything but esses. That's a great time to just walk away for a while. Maybe check some reference mixes...

A lot of times when the esses are taking your head off, it's because the vocal is actually just too loud. Other times it indicates you've gone too far on some compressor somewhere.

But really, there's nothing in most voices that hits those ess frequencies except for esses, and static EQ usually actually works as well or better than any dynamic fuckery you might try.
 
ashcat_lt said:
Maybe it's just the performers I record or the genres in which I work, but I have found that de-essing rarely really works and is actually rarely necessary.

First of all, I think that very often when we get to the point of trying to de-ess a track, it's because we've lost perspective. We've become hyperfocused on that one aspect of the mix and get to the point where we almost can't hear anything but esses. That's a great time to just walk away for a while. Maybe check some reference mixes...

A lot of times when the esses are taking your head off, it's because the vocal is actually just too loud. Other times it indicates you've gone too far on some compressor somewhere.
Click to expand...

I agree with that for the most part.

ashcat_lt said:
But really, there's nothing in most voices that hits those ess frequencies except for esses, and static EQ usually actually works as well or better than any dynamic fuckery you might try.
Click to expand...

But not that. Meaningful voice harmonics can go above sibilance frequencies. But while they can overlap in frequency they generally don't overlap in time so dynamic processing is a viable approach.

Of course the best solution is better vocal technique.
 
I usually slice the sibilance and then lower the volume of the slice. Usually it doesn't require a crossfade. I don't think it's good engineering or technically correct, but I find this works better than de-essers, which don't work at all on my voice.
 
I always opt for manual de-essing...most times with a parametric EQ using a narrow Q and level adjustment.
It's not a one preset works for everything approach...but if yuo take the time to fine tune the EQ point, Q and level...you hit the most annoying point of the ess, while still allowing the general ess sound to come through, so you don't end up with lisping.

The other times I can just slice the vocal the ess spot, and some simple level adjustment is all it takes to tame it without removing it....and again, each ess sound is different, even when it's the same vocal track...so there isn't a single preset you can create, IMO.

That said, I do have like 3 "ballpark" presets, and I can usually pick one of the three that gets me close, and then I make a final adjustment to it.
I am not a fan of applying any one plug/preset across the entire track.
 
Nola said:
Oh sweet, I didn't know that was a valid method.
Click to expand...

I don't know how your DAW does track cuts/rangers/splits....but all I do is cut the vocal where the ess starts and also where it ends.
I'm not moving any of that around or anything, I've simply isolated that section of the track...and then I just lower the level for that spot.

Sometimes just 1-2 dB level drop, and the annoying ESSSSSS becomes a normal ess sound....because you want some of it to come through. I mean, you can't just cut it all out, because it will sound weird.

The thing with the parametric EQ is the same process, but the EQ adds a bit more focused control on specifc frequencies as opposed to just lowering the level on the entier ess sound you marked off.
So you mark/cut the vocal at the start and end points of the ess sound, same as above...and then with the parametric EQ only added to that small "slice", you apply a tight Q and dial in the frequency and level at the parametric that removes just enough to tame it...done.

:)

It's surgical...but it's also manual, and some people get all giddy about the thought of having to actually do some manual work ;) ...so instead they apply processing to the entire track and try to find a sweet spot that works on all the ess sounds...and you sometimes also end up affecting other shit too.
If you learn to do it manually...after awhile, you can slice your way through a track in like no time at all. :cool:

Hey...whatever works...everyone has their preferences.
 
RFR said:
Quit writing lyrics with 's' in them, problem solved! :D
Click to expand...

When I'm doing my own vocals...I've learned how to avoid it at the mic to a great degree...so much less time needed dicking around with it in the DAW. ;)
 
miroslav said:
I don't know how your DAW does track cuts/rangers/splits....but all I do is cut the vocal where the ess starts and also where it ends.
I'm not moving any of that around or anything, I've simply isolated that section of the track...and then I just lower the level for that spot.
Click to expand...

Yeah that's exactly what I do. It works so much better than a de-esser, imo.
 
bouldersoundguy said:
I agree wiBut not that. Meaningful voice harmonics can go above sibilance frequencies. But while they can overlap in frequency they generally don't overlap in time so dynamic processing is a viable approach.
Click to expand...
Above yes. I'm not sure the overlap matters most times. Note that many popular vocal mics are boosting these frequencies to begin with. Changing mics just might fix the whole thing, but it's too late for that now. I'm not saying it always works, but maybe more often than some would expect.

Of course the best solution is better vocal technique.
Click to expand...
Duh. ;)
 
I guess it depends a lot on the specifics. If it's one of those whistle-like sounds a narrow cut could be used without too much effect on the overall HF. But some singers (especially those who are used to performing without amplification) really push their esses. Those can cover a broader range that eq can't cover without taking out a lot of important tone.

Which reminds me of one of my earliest released recordings. The singer tended to be on the verge of lisping. He discovered that if he deliberately whistled his esses a little it fixed the problem. He trained himself to move his tongue back away from his teeth. It worked beautifully.
 
You must log in or register to reply here.
Back
Top