Spoken word: Minimizing "wet mouth" sounds

[DCtoDaylight]

Hey, all - I do a good bit of spoken word recording (for CDs, video soundtracks, narration, etc.), and wonder how the real pros deal with the issue of the reader/speaker's mouth making noise when they move their lips or tongue or whatever. I'm not certain what causes it; maybe an excess of saliva or something, but I've run into some people (including people with otherwise outstanding voices) who have quite a bit of it.

I've tried moving the mic farther away, which helps some, but it starts to lose presence. EQ can also help, but again, with a loss in the overall sound. Using a less-bright mic also seems to help some. But with some people I still seem to end up with enough smacking sounds to be distracting. The problem seems especially pronounced on material that gets played over computer speakers (podcasts, narration, etc.) - there's less bass, so the mids and highs, where the mouth noise is, are more accentuated. It wouldn't be an issue if they were singing, since the music would cover it, but in straight spoken word it can become distracting.

I'm just wondering if I'm missing any tricks here - some way to either dry peoples' mouths out, or to avoid picking up the sounds they make.

Thanks in advance for any tips!

[thethrillfactor]

Originally Posted by DCtoDaylight

.

I'm just wondering if I'm missing any tricks here - some way to either dry peoples' mouths out, or to avoid picking up the sounds they make.

Thanks in advance for any tips!

A glass of apple juice and don't compress/limit so much.

[LeeStoker]

Mouth full of flour, make sure to use a pop filter you don't want flour "spittle" getting in the diaphram. On a serious note one time I had a guy doing a soft acoustic passage and he was a loud breather so I had him where a dust mask so as to kill the noise in the hot mic.

[GregP]

Sometimes the "wet" sound is actually a dry mouth. Pasties.

[zenkonami]

Green Apples have a great reputation for solving the problem (not in everyone.)

Sometimes a swig of water ever page or two will fix most of it.

Definitely different mics make a difference. Women also (as a rule) seem to be noisier than men, probably just due to higher frequencies eminating from them.

That's my experience. To a small degree, you just live with it with some artists. The better artists really learn the tricks. Scott Brick, reader of many, many audiobooks and some narration, uses a wide variety of techniques, and has a very specific plan of attack (every three pages...every six pages...) but it's a bit complicated and I don't recall everything he used off the top of my head. This is a guy who reads for around 300 audiobooks a year, though.

Definitely make sure they've got water in the booth, though. That usually fixes it at least for a few lines, if not more.

Good luck!

-Zen

[RobMacki]

Originally Posted by thethrillfactor

A glass of apple juice and don't compress/limit so much.

That's the secret.
That and the pencil tool.

[nlc201]

Originally Posted by RobMacki

That's the secret.
That and the pencil tool.

I can't count all the hours I've spent "de-spitting" VOs and vocals. The pencil tool is the main tool in the arsenal for this. I've gotten fairly adept at redrawing waveforms. I always wonder what frequency(s) I'm actually drawing in. Often times when the spit is in the clear (like between words), editing out the break is a fast way to deal with those guys in bulk. It should be said though that the scrubber tool is the pencil tool's best friend. You can usually pinpoint a click pretty quick by homing in on it with a scrub.

Another method for the really stubborn clicks is to copy a section of waveform from an area immediately preceding or following the spitty area. You have to be very accurate with how you place the waveform segment(keep the overall waveform consistent by editing in and out at the starts if cycles) and crossfade. This is a very complex process and doesen't work much of the time. I only attempt it as a last resort when I just can't get the things out any other way. Sometimes you can get away with actually editing out the specific waveform section and splicing the two ends to make a slightly shorter word. This works better on long words with short sections of affected waveforms.

Spitty consonants (ffffff sounds in particular) can often be copied from other words and flown in. I also use fades as a length control a whole lot. Often, long spitty syllables can be instantly be made better with a proper fade in/out. This technique is also great for breaths. I hate removing breaths all together because it sounds unnatural to me. But compressed VOs can often have some annoying breaths pumping in and out unnaturally. I set the beginning of the fade up right at the start of the breath and make the end point of the fade (where the fade brings the volume to max) about at the end of the breath. The breath is still there but it's usually much more managable and usually sounds pretty organic. This process takes less than a second in PT (click the front of the breath, hit A, click the end of the breath, hit D, done). That's important when you have to do this hundreds if not thousands of times.


Sometimes is useful to put a "work" EQ setting on the VO while your despitting it. A boost on the top end will bring the spit out a bit more to be easier to hear. You can always change/ditch the EQ later to actually suit the sound. Also during this process I'll typically put an EQ accross my master buss sharply cutting off frequencies below 50 Hz. This is done not for sound purposes but rather as speaker protection. If you ever watch your woofers while you scrub a VO at 1/10th speed you'll see why this is needed. I sure as hell ain't killing my speakers just to edit VOs!


I can truly say that I hate this process. Even getting paid righteous money to do it, it still sucks. But, it can make a hell of a difference if done diligently and properly. I certianly don't hate the results when it comes time to mix!

[juniorhifikit]

Originally Posted by thethrillfactor

A glass of apple juice and don't compress/limit so much.

OK, secret's out.

Though, juice has a lot of extra sugar - I find a bite of green apple works best.
















there is SO much good free advice on this board. I wish it had existed when I started 20 years ago

[Yetti]

Originally Posted by LeeStoker

Mouth full of flour, make sure to use a pop filter you don't want flour "spittle" getting in the diaphram. On a serious note one time I had a guy doing a soft acoustic passage and he was a loud breather so I had him where a dust mask so as to kill the noise in the hot mic.

I also did the dusk mask one time!
I felt like I also needed to put one on in the control room, as watching a dude record with a dusk mask on made we want to laugh!!!!

[The Byre]

The closer you are, the more 'mouth noises' you will get. The farther away, the more of the room you will hear.

We use a good vocal booth and an M149 at about three feet distance at eye-level height. It works.

[RadioMoo]

Originally Posted by zenkonami

Green Apples have a great reputation for solving the problem (not in everyone.)

Sometimes a swig of water ever page or two will fix most of it.

There's one studio where I cut vox occaisionally that keeps a stock of green apples on hand - they help, and so does the water.

But what really helps is if the talent (I'm speaking for myself here, too) just relaxes more and doesn't try so hard to form the words - a lot of the smacking noises come from tension around the mouth, and that's best fixed with diplomacy, not a different mic.

(Just as an aside since this is in So Much Gear, I'm not allowed to just pick up at this studio if I make a mistake - the engineer takes the session back to the point before I screwed up, and I don't know what system they're running, but it makes a sound like tape being rewound, and it's a digital system !)

Originally Posted by nlc201

Another method for the really stubborn clicks is to copy a section of waveform from an area immediately preceding or following the spitty area.

When I'm doing my own stuff at home, I use pieces of room ambience for this - I mean, even a dead room has a "sound," and I just select a small section that includes the click or mouth noise, time it, cut it out, then replace it with a piece of the ambience that's the same length. After the crossfades are applied, you can't tell.

Sometimes, depending on the pace, you can just select the noise - if it's between words or sentences - and delete it.

[DCtoDaylight]

Wow - thanks VERY much to everyone! I can't tell you what a luxury it is to get so much good info on this somewhat arcane topic!

I'll definitely have green apples on hand next time, and experiment with more/less water. And diplomacy!

And thanks, TheByre, for the placement tip - three feet is more than I've tried. I would think that a very dead booth would be very helpful in that situation.

I'll definitely experiment with all this in my upcoming sessions....y'all have given me a lot of good stuff to chew on. On my current project (which prompted me to write this), the noise is pretty well woven in with the words, and re-tracking isn't an option, so I think it'll be an EQ job - I've gotten to a curve that I think keeps the voice sounding OK (if different) while getting out most of the noise.

Also, in the course of his exceedingly helpful post, NLC201 mentioned using a "work EQ" setting - one way I've found of doing that, especially for work that's going out over the Net, is to monitor over either cheap little PC speakers, the speakers in my laptop, or iPod earbuds. They seem to provide a nice "worst case" listening environment that accentuates bad stuff and is in line with how people will be hearing it.

Thanks again!

[rwhitney]

I haven't had much trouble with this since switching to the Coles 4040 mic for a lot of vocals, as opposed to the M149 (a great mic, but sometimes problematic).

[will84]

Originally Posted by rwhitney

I haven't had much trouble with this since switching to the Coles 4040 mic for a lot of vocals, as opposed to the M149 (a great mic, but sometimes problematic).

Browsing old plosive/mouth noise posts and this is one that I have a vested interest in -- being both a fan of spoken word (blame Henry Rollins, though there certainly is better) and an ADR engineer for animation.

One of my main jobs is "cleaning up" dialogue for the mix engineer of the show. I've had numerous times where, even with plenty of water and retakes, the actors have what many of our engineers have dubbed "mouth clickies", "mouth squishies", etc. After a year, I've gotten to where hearing these things on live action audio kills me the same way an over-compressed song kills a mastering engineer

My most recent mix engineer has given me lots of tips on getting my dialogue cleaner before it gets to him. My experience...

1) One trick: compress as much as you can without distorting and over EQ the top end. Not for the final mix, but while you're cleaning. Like training in higher altitudes, what you catch here will sound better once it's EQed/compressed for the final mix. Just try not to kill your ears in the process. HPF up to 80Hz or so depending on the actor (reader) and get rid of some mud, boost above 2k to reveal the snap-crackle-pops.

2) Mic technique makes a difference. Most recently I started micing at about nose-level with the mic at least 6 inches away with a pop filter halfway between, mic at approximately a 15 degree angle aiming back. This cuts down on plosives (as does the pop guard) and helps control sibilance without sounding off-axis if done right. Distance will also make the difference... I made the mistake for a while of having people very close for the sake of levels and paid the price with hours of cleanup.

3) HPF your S's, F's, sh's, many V's... longer ones especially. I use 1k as a starting point, go higher for parts of spittier ones and lower sometimes for male ones that sound thin. You could consider compressing individually them with AudioSuite (or the non-PTs equivalent) to even them out.

4) HPF to 200 for plosives. 24 db/octave for me. Just the first part... and-a-half, so that you can trim back the region separator and cross-fade as is appropriate. Fade-ins can help too.

5) Sometimes little sounds between words can be completely cut. Or, if that sounds unnatural, LPF them with a non-steep filter just enough to make the next step easier...

6) Pencil tool is the end-all. I use it every day at work and, used correctly (which takes practice), it can do 80% of the work. Clicks, pops, ticks... little sawtooth waves on a sample-sized level that plague my life But getting to be a good pencil tool artist can make these easily tamable.

I plan on applying these newly-learned skills to vocals on this new album I'm recording. Even if not immediately noticeable upon listening, getting rid of these unneeded/unwanted frequencies can't hurt. Especially in the vocals ever need to soloed for anything...

[celticrogues]

A swig of water and a bite of green apple usually works here!

Cheers!
-Mike

[Skrazz]

Has anyone had any luck using Noise Reduction plugins such as X-Click, or X-Crackle? I'm editing an audiobook which exemplifies the "mouth sounds vs room noise" battle.

[voiceactor]

Try using a small atomizer with nothing but a little warm water. You'll be amazed.

[mitgong]

Izotope RX will change your life. It won't save you from having to do this type of work, but..