Sibilance makes me want to cut my ears off

rocknrollstar

New member
See attached mix.

I've got the guitar to what I think is quite good, but the excessive "s-ing" on this track is annoying me to no end.

I've tried Spit fish (free VST de desser, x2 on this vocal alone) and nothing works.



1.) How can I reduce essing on this track?
2.) Is there a way to record in the first place which prevents this from happening?
 

Attachments

  • BECK SIL QUEST.mp3
    2 MB · Views: 25
I personally think you might just have bright monitors, because in my speakers, it sounds just fine. I'm not hearing too much sibilance on my end.

My monitors are particularly susceptible to sibilance, too. It always jumps out at me on these monitors, so it forces me to keep them in check, but I'm not hearing any sibilance personally.
 
I hear what you're hearing, if its the sibilance at the end of words (that end in 's' obviously). I never had any luck with Spitfish with end-of-word sibilance. Volume automation works, but is tedious.
 
How can I reduce essing on this track?

Since it's already recorded.....just edit out the sibilance.
I do it all the time. Just cut/split the actual "sssss" section of the audio, leaving crossfades between the cuts....then just lower the level on it and/or use some EQ to soften it.
I find it easier than typical volume automation.

Next time you sing, turn on an angle toward the mic, or place it a bit higher than your mouth, and of course, use a pop filter.
 
I personally think you might just have bright monitors, because in my speakers, it sounds just fine. I'm not hearing too much sibilance on my end.

My monitors are particularly susceptible to sibilance, too. It always jumps out at me on these monitors, so it forces me to keep them in check, but I'm not hearing any sibilance personally.



I wish this was true, but it is definitely there. Tried it one a few different speakers/headphones etc - glad it sounds good on something though lol

I hear what you're hearing, if its the sibilance at the end of words (that end in 's' obviously). I never had any luck with Spitfish with end-of-word sibilance. Volume automation works, but is tedious.


Yes that's where I'm hearing it. I tried automation but it still didn't sound too good. I don't mind tedious if it's going to end in a good result. MIDI for some reason is what screws my mind down quickly - I love alot of electronic music/music with electronic elements, but for some reason MIDI just gets me agitated.


Since it's already recorded.....just edit out the sibilance.
I do it all the time. Just cut/split the actual "sssss" section of the audio, leaving crossfades between the cuts....then just lower the level on it and/or use some EQ to soften it.
I find it easier than typical volume automation.

Next time you sing, turn on an angle toward the mic, or place it a bit higher than your mouth, and of course, use a pop filter.

This could be a bit easier. I'll give it a go the now and post any results.

I was sitting straight in front of the mic, but it was pretty much straight on. Next time I'll turn it slightly. You wouldn't know it to listen to the track, but there is actually a pop shield over the mic.


Thanks for the responses folks, it's good having somewhere to talk about this stuff.
 
Try backing off the mic and pointing it towards your chin? That's usually what I do with sibilant voices.
 
I happened to be succesful with multiband compression. Find out the Sibilant band, put that into one of the multicomp bands and fiddle until you have a gain reduction at the 's'. (Of course don't use makeup gain ;-) ) Find the band with an EQ raising the freqs.I did not use the 's'-band as the upper band as this might take away the air from your recording...

And leave your ears where they are :D
 
Try backing off the mic and pointing it towards your chin? That's usually what I do with sibilant voices.


Only disadvantage is I try keep the volume down and record close to avoid picking up the room. Hmmm....


I happened to be succesful with multiband compression. Find out the Sibilant band, put that into one of the multicomp bands and fiddle until you have a gain reduction at the 's'. (Of course don't use makeup gain ;-) ) Find the band with an EQ raising the freqs.I did not use the 's'-band as the upper band as this might take away the air from your recording...

And leave your ears where they are :D


I've used x2 spitfishes, a bit of this method and actual automation (see attached)


In the words of Jack Nicholson, this is as good as it gets.

Next time I'll record it correctly in the first place - the more I plugin the more I take out it would appear. Can't put toothpaste back in the tube eh?


What do you make of attached mp3?


Also, I've decided to keep my ears purely since it'd be a shame to waste these headphones.
 

Attachments

  • Beck Without Silb.mp3
    4.9 MB · Views: 4
I just do a deep V cut with automation... doesn't take long and you work out after a while exactly how deep to cut. You can almost do it "live" whilst the song is running through if you don't have too many s words. Works better than Spitfish for mine...
 
I don't think it's particularly sibilant in that it's not terribly harsh or cutting. I think you may just be self conscious about the performance of the esses. One of the tricks I picked up is to, if you have a boom stand, mount the mic from above at about nose level. This does two things: it gets the mic capsule out of the direct firing line for plosives and it forces a singer to change their air way a bit by keeping their head tilted which seems to help a bit with sibilance and overall vocal projection; tends to keep the jaw low.

In terms of good de-essing plug-ins, I do video editing/voiceovers so I rely on Fabfilter's Pro-DS. Barring that, http://sleepytimedsp.com/software/lisp/ Lisp has given me some great results when I'm stuck working on someone elses computer. Free but Windows only.
 
The best solution is to re-record again without facing the mic upfront but if you need to fix this during mixing and simple de-essers don't work then it's time for some manual automation :)

Tiring? yes. Effective? Hell yes.
 
I don't think it's particularly sibilant in that it's not terribly harsh or cutting. I think you may just be self conscious about the performance of the esses. One of the tricks I picked up is to, if you have a boom stand, mount the mic from above at about nose level. This does two things: it gets the mic capsule out of the direct firing line for plosives and it forces a singer to change their air way a bit by keeping their head tilted which seems to help a bit with sibilance and overall vocal projection; tends to keep the jaw low.

In terms of good de-essing plug-ins, I do video editing/voiceovers so I rely on Fabfilter's Pro-DS. Barring that, Lisp [De-Esser] Lisp has given me some great results when I'm stuck working on someone elses computer. Free but Windows only.


Yeah essentially this recording was an experiment - I got the guitar to a point where I think it works, and now I know how to go about recording the vocals. All this computer editing tires me out - thank you for the response + link though. Hopefully tomorrow I can get some time to myself and have a good read at it.

Sounds odd but would turning the mic on it's side also make a difference or am I just being weird?


The best solution is to re-record again without facing the mic upfront but if you need to fix this during mixing and simple de-essers don't work then it's time for some manual automation :)

Tiring? yes. Effective? Hell yes.



Yeah I'm going to re-record. Tons of computer editing really frustrates me. The less work I do on a DAW the happier I be and probably means it's the best "sounding" as well if that makes sense e.g. caught naturally at source.


I just do a deep V cut with automation... doesn't take long and you work out after a while exactly how deep to cut. You can almost do it "live" whilst the song is running through if you don't have too many s words. Works better than Spitfish for mine...



Spitfish has been unhelpful so far, however willing to be user error could be an attribute to that.
 
I heard a song on the radio yesterday (might have been Dave Matthews) that had a sh*tload of sibilance compared to your recording. Just shows what you can get away with ...
 
Last edited:
Be aware of when it happens and move your head up when the S happens durning the tracking.
Should help a bit.

I noticed many years ago that some recordings the singer never sang the S and the listeners brain just automatically filled it in. Until you notice it then it sticks out.
Give a listen to some Elvis or Beatles songs and you'll see what I mean.
 
I know I'm likely over-hearing it, however it is there.


Out of interest, any Dylan fans here? specially the very first album self entitled.


It sounds all live to me, was just wanting someone to confirm.
 
Be aware of when it happens and move your head up when the S happens durning the tracking.
Should help a bit.

I noticed many years ago that some recordings the singer never sang the S and the listeners brain just automatically filled it in. Until you notice it then it sticks out.
Give a listen to some Elvis or Beatles songs and you'll see what I mean.

I'm always not singing ts and ps and ss in harmonies... sounds weird at the time but you don't noticed it when mixed in.

And whenever I try to turn my head on the S words I end up stuffing something else up... these days I try to work on not making as much sibilant noise in the first place and then doing the v cut thing, which works for me...
 
Cant see it mentioned here, but the pencil trick works pretty well when recording sibilant vocals.
 
Because it diverts the air stream from hitting the mic capsule directly.

I gave it a go, using that, a handkerchief over the mic and a pop shield. It was literally for one line (for a friends electronic song) and we've since scrapped it, but I will try again on next actual recording and post results.

I take it the pencil faces the persons mouth vertically?
 
Back
Top