I am working with a female vocalist on original songs, and I keep bumping into this problem that I don't yet know how to deal with. It seems her mic has a definite "sssss" sound to it, but otherwise records pretty well ( I think). I have tried at least one de-esser plugin and i saw no difference. So I am looking for help and to educate myself on this matter. I am going to youtube to see what I can find. Maybe I need to crack some books for the manuals for these plugins I told her that her mic is proned to this, and that she might want to look at using a different mic. In the meantime I do want to educate myself so I can at least address this. Thanks in advance.

 

Well the first youtube video I found (de-essing in audacity), has me worried. I watched as he applied the plugin and I was not impressed at all with his results. It still was quite pronounced AFTER he was done. Reduced, I guess, but not at all fixed. Hope someone has gotten better results than what i saw in this first one.

 

You can address this problem by using a plugin.

free ones includes spitfish (a dee-sing plugin) and paid ones including fabfilter ds (both of these are professional sounding) or you can also use a stock multiband compressor that comes with your daw and compress the problematic frequency (normally in the 7khz to 14 khz area of the spectrum) or you can go all manual and get an audio editor (not a daw but an audio editor) and manually fix the problem areas this will be more transparent but will be more laborious.

Hope this one helps.

 

Have you had good luck with this? Obviously I am aware there are plugins, because I watched a video on youtube, and tried one myself. I suppose what i am really asking for is the HOW. In your response you mention a specific frequency range. That is most helpful. I am a big ozone isotope user, and they have a good parametric EQ on there. I will try it on the vocal to target the range you mention. You also mention specific plugins which is also quite helpful. I suppose I just need to dig in. The info you gave is good and should help me on my target. This is the type of info I am seeking. The more general info is less helpful. I will be putting the mixes in players on the same page so comparrison is immediate. If you are interested in listening in, send me a private message. Maybe you can tell me what I am not doing correctly (although hopefully my ears will point to this). Thanks in advance.

 

Take one Fabfilter DS and call me in the morning. Always if you can, try an array of different microphones and pick the one that satisfies both the singer and yourself. Believe me you, it isn't always the bestest mostest expensive microphone that works best for each singer in each room for each song. Go with your ear and not with your preconceptions. Also, if the singer is sibilant most compression that you use will make a bad situation even worse. Even essssy singers can benefit from working on the their technique to lessen this common problem. Practicing sliding the tongue lower in the heir he mouth away from the their upper palette can have a pronounced positive effect. Practisssse makesssss perfect.

 

This is actually MY first response. I am keenly aware of mic technique myself. One of our songs the beginning is lower and quieter, and I had the distinct impression she was purposely close to the mic, (which is a common mistake). She assured me she was not real close to the mic, and had a 'pop screen" in place. I have actually been wanting to experiment with her on "distance" from mic, and I might get to do that. So I am onboard with this approach, but this is in a perfect world that her technique is optimum. I do intend to work this aspect. She works onstage quite a bit in Germany, so maybe I can get her to try an sm58. If I can get her to shop for a new mic I am thinking of recommending a RODE. The thing is I feel the need to educate myself on the GOOD use of these plugins. Knowing that frequency range is good. I want to zero in on specific freqs and leave others alone so as to not remove good stuff. I know there is nothing like just digging in, and reading etc, but I also have learned so much from people and they have saved me so much aggrivation, that I thought it a good idea to inquire here. Maybe save myself time and aggrivation. For the time being we have just lived with the sibilance. Something that really sticks in my craw. Thanks for the response. Even if it is already what I have been thinking it serves to reinforce those thoughts. Much abliged!

 

Really it would be cool if I found a plugin with a preset that helped. If I did, then i could actually study what the preset did, and learn.

 

I was going to suggest this. If de-essers, multiband comps, or compressors which are sidechain filtered to the high frequencies are not working for you, try a dynamic EQ. If your ears are not trained enough to pick out the problem frequencies, watch the spectrum analyzer and set a dynamic node for the frequencies that are noticeably higher than others when you hear sibilance. Worst case scenario, you can automate the node to follo around the problem frequencies if you have to. Sometimes I think a dynamic EQ is a bit more elegant of a solution than the various forms of compression.

 

I am thinking I can grab a sliver frequency (shaping it to be very thin) in the parametric EQ and lower it dramatically and slide this sliver up the freq range I should be able to come to the problem frequencies with my ear right? I can do this realtime as the track plays in loop mode. Is it as simple as this?

 

btw. de-esser, MB comp or comp in SC on filtered freq is actually "primitive" dynamic EQ (with some phase issue, depending on tools that you are using but nevertheless...)

 

Hello.
First, get the girl to control her breath and air column with the diaphragm muscle, and not to lean on consonants and sybillants. Training method - get an sm58 without the mesh grille, let her sing a few hours at 1-2cm away from the capsule, and try to do it without pops and sybillants. Keep the volume down and duck If necessary... If the girl sings mostly in german, it favors the sybillants - each language has its specific "sound".
Technically, the Ozone is a very good de-essing tool. If the entire track is hissy, activate the dynamics module, make it a 4 band, and use multiband compression. You can play with the green and the blue range cross-over points, in order to de-ess progressively. Check the Options/Dynamics tab, and see what tweaks work best - I use digital crossover, 2048 buffer, 6.1 Q. The female voice sybs are around 7khz, extending to 12 for "s". If only the 's' consonants give you headaches, lower with ~3dB at 11.8k, with Q of 8-12. Don't forget about masking - lifting the bass with 1-2dB at ~370-430hz or the middle at~900-1000hz can attenuate the sybs. And take care of the 3-3.3khz, If too much, the voice sounds brittle. Easy on the 6 to 7khz range too. My advice - eq the voice for maximum clarity, and manually de-ess where required. In this case, use the Ozone eq module, with digital, surgical, linear phase settings. Locate each problem on the waveform in the dae, and apply in/ out crossfades with linear progression ("X") when de-essing. Let the waveform guide you.
Many mics sound good but ess a lot, so it's best to track the voice dry (If the singer lacks the fx in the headphones, make a second 'wet' channel, and route it to the phones).
Another microphone - there are mics specifically tailored for female singers, She should try those first. But the idea is to have an all around best sounding mic for her, not a mic that avoids essing, but sounds worse.
Sm58 is not the best idea If the voice is too bright. 58s sound "glossy", that is, they lack definition. Rode mics sound clear, maybe too clear for a sybillant voice; Nt2000 sounds more velvety. I would go for an Akg Tube, or a C4000 for recording (you can also try Ev RE20, it can be used with good results for female voices), and an Sm57 for stage use. At least judging from what you said.
Well... good luck with the essing, and the de-essing!

 

Last edited: Jan 19, 2017

All the above will work.

Though with some voices, and especially female vocalists that like to get close to the microphone, you will have to EQ some rather ungraceful notches between 5KCPS - 8KCPS. The problem is that some microphones also accentuate these frequencies and mixers also tend to want to boost these as well.

Automation as well as a De-Esser or D-EQ post compression may or may not help to alleviate the issues. It is all source dependent. If you push it too hard, those ESSES will turn into ZEES (which what Jazz singers used to themselves to avoid the issue in the old days)

ZUBIDOU ZHAAAAA.

 

Here, http://www.emusician.com/how-to/1334/how-to-sibilance-smackdown/61500 came in my mailbox today and is apropos to this thread. The Metric Halo Multiband Compressor is my favorite multiband comp and might help you out of a jam. Depending on the singer sibilance will span anywhere from 7kHz to 11 or so. Do the listening thing of what you are lassoing in and if you still hear sssss widen your net. You wouldn't be the first to go crazy trying to get rid of nastiness like this.

 

Just the other day I was using my RE20 through the Daking Mic Pre One, pushing the pre with about 63 or so dB of juice, and although the Daking wasn't noisy (a very nice piece of kit this, especially for guitars and vocals) my voice was super sibilant.I switched over to the Warm Audio Tonebeast and no more esssssss. I guess that the pairing can be very important.

 

Sure, it might be better to automate it or set a threshold with the dynamic EQ though, so it's not always taking a chunk out of your high frequencies. If it's a very thin band the effects might be more good than bad, but there's still some collateral damage.

Yeah, I'm finding the Audient preamp is gentler on sibilance than my ART Voice Channel which, by all other metrics, is probably a better piece of gear for that purpose. Lately I've just been recording through the interface.

 

I didn't really know/think of it like that, I just think the overall concept of compression is often not as suited to the task of de-essing. Multiband comp is great if you're recording vocals over an instrumental and you want to tame some frequency band to make room for the vocal while keeping the balance within that band intact. Or if you need to tame the low end of a track without taking too much punch out of it. But if you just want to remove a very narrow, very select bit of frequency content from a vocal without making things sound unnatural, a carefully placed dynamic EQ node makes way more sense to me.

Especially with wide band compression--okay so you're compressing the whole channel when the highs get too loud. So what? The balance between the whole of the vocal's HF content and that one unruly resonance is practically unchanged. Wide band compressors for de-essing just seems like kind of a makeshift solution to me.

 

I was just trying to shed some light on tehnical aspect of this topic, what will you use or anybody else it's matter of personal taste.
I love dEQ as it is, but simimlar or same result can be achieved with other things and that's it.

 

Go through your track in your DAW and manually reduce the clip gain of all those dense footballs in the track. I had to do this with some very sibilant "ch" and some "s" sounds in a recent track I had to mix.

 

Most of what I'd recommend has already been mentioned. I have exactly this same problem with the female singer in our group. No matter which plugin I tried, it never could remove enough to satisfy me. What I did to resolve it was running a de-esser before any eq/compression, and running another after compression. When that didn't get it, simple level automation has to be used.
I use the pop filter as a sort of barrier to prevent proximity effect. I had to work really hard to establish the rule, "DO NOT TOUCH THE POP FILTER OR THE MIC UNLESS YOU ARE THE RECORDING ENGINEER!!" But seriously, all it took was a few sessions showing how bad the resultant sound was when she "ate" the mic and how open and clear it was when she maintained the distance to the pop filter. It's unbelievably hard to break some learned habits unless you can show the singer the "why" of what your asking them to do. I have had to reiterate that recording vocals is not the same as live performance. It is not nearly as forgiving.
Also, this problem worsens if the angle or height of the mic is either too high or too low. I don't remember which direction made it worse, but your ears will tell you. Something about the way the air leaves the mouth when forming these sounds.
Another thing which can cause some siblance issues is room reflections. I've actually had luck moving the mic stand half a foot or so or hanging a thick blanket either behind the mic or to either side. Especially untreated rooms. I'd assume that this has to do with distance/frequency of the reflections.
I really like the idea of clip gain edits, but haven't ever done it. Really helpful to show the "football" shape in the vid!

 

Last edited: Jan 18, 2017

I have never had any problem with de-essers doing their job. I like the Butch Vig vocals plug in. I also like that it looks steam punk-ish, for no particular reason. Have you tried another mic\ and varying the placement, etc.?