Nasality: How to minimize effect.

J

j1clark@ucsd.edu

Veteran
The setup: The Wife has been wanting to do some tutorial type videos. We've been talking about this for years... and so I was able to setup for about an hour to do some tests.

The Wife is hard of hearing, and has had 20 years worth of speech therapy during her early years...

Anyway, for this sort of speech issue, I'd like to know how people deal with nasality. Things that occur to me... (and replacing the speaker is not one of them...).

1) nasal spray... not an option... the Wife hates 'artifical/chemical' things...
2) DSP in post processing. How much can be compensated with what methods.
3) Microphone placement. While some of the material we can have a mic in the shot, for others it would be better not to have a visible mic. At the moment I have an Oktava MK-012 with cardioid and hyper capsules, and a couple of 'dynamic' mics.
4) Type of mic. Are some types better suited to minimize such an effect.

There are other speech problems, but I'm not up for paying for major speech therapy at the moment...

And oddly, I don't really 'notice' these sorts of things during a live presentation. And at seminars in the past, few attendees really noted any strange 'speech' patterns. But listening to the recordings the 'problems' really sort of popped out.
 
I wouldn't go with an SM57 or SM58, if those are the dynamics you have available. They have a mid-range "Presence Peak" that will probably be just-the-thing to accentuate the nasal quality you are trying to avoid. Likewise, be careful with mike placement; the "proximity effect" happens with some microphones the closer the speaker is to the mike; it boosts the bass and low mids, which might also accentuate some "nasality."

You can probably deal with most of what you're worried about with a software graphic EQ.

GJ
 
Agreed. Mid-range is where the problem lies. Mic choice and mic placement can help, minimally, but your better bet may be to compensate in post with a good, parametric EQ.

One of the studio shows I used to do had a host who was very nasally in her speech. It wasn't because of hearing issues or anything like that... she just had a not-quite-Fran-Drescher nasal tone to her voice. We had Sanken COS11 mics on everyone. This was a multi-cam show switched live-to-tape. The only thing I could do about it was EQ in the mixing console... and I had to take tons of upper mids out of her just to make it palpable.
 
Can you post a sample?

that term (nasality)could possibly mean several things. Is it a real term?

if its just a very narrow frequency range you are having trouble with you might be able to use a de-esser plug-in to try and control that particular frequency range, but if you are trying to remove a fundamental characteristic of someone's voice, I really don't think you will be able to.

Id like to be able to sound like Dan Lafontaine, but sadly there isnt a plug-in or box in the world that can do that.
 
I have a slightly nasally voice myself, and find that my Electro Voice RE20 makes me sound better than the other microphones I've tried (not that I have any ambitions of being a performer). However, mine is a mild case, not a severe one, and the RE20 is not a microphone you would be able to use on-set in very many scenarios -- it's big and heavy, and needs to be used very close to the speaker. It is far better for radio or off-camera narration recording. The Heil PR40 has a somewhat similar sound at a lower price, but doesn't control proximity effect as well.

- Greg
 
I still think a de-esser that has bandwidth capabilities and a broad frequency range will be your best solution as it will dynamically compress or De-emphasize a certain frequency range with out severing the signal too badly and fixed like an eq would.

Having said that, while searching for proper frequency to de-ess certain people in the past I noticed you can make the effect act so severe that it can make some one sound like if they have a severe lisp, so Im certain if you can find that specific frequency range that your wife exibits you can dynamicaly limit the level of it and still keep it sounding fairly natural.

Good luck
 
>>>>while searching for proper frequency to de-ess certain people in the past I noticed you can make the effect act so severe that it can make some one sound like if they have a severe lisp<<<<

This is why I think those Steely Dan records from the 70's make Donald Fagen sound so, so-- "I crawel lyke a vih-pah, thru theesch such-bur-ban schtreets..." Over-use of de-essing.

GJ
 
recordingwhiz said:
I still think a de-esser that has bandwidth capabilities and a broad frequency range will be your best solution as it will dynamically compress or De-emphasize a certain frequency range with out severing the signal too badly and fixed like an eq would.

Having said that, while searching for proper frequency to de-ess certain people in the past I noticed you can make the effect act so severe that it can make some one sound like if they have a severe lisp, so Im certain if you can find that specific frequency range that your wife exibits you can dynamicaly limit the level of it and still keep it sounding fairly natural.

Good luck
Click to expand...

In my searching I've run across mention of the Electro Voice RE-20, mentioned Greg, for a 'mic', and also for DSP some mention of a comb filter... More simpler suggestions seem to be to use EQ and test which frequency band 'improves things'...

Hopefully once I get some improvement... things wont change...

As for 'de-essing', there's a bit of that, but more noticeable was the substitution of 'sh' for 'ch' in 'chopper'... sounded like 'shopper'... What I'm going to do with this first round is transcribe the audio material then set up again, and have her do a much reduced in length version, with perhaps some note that's on the need for clarity on certain words... if I survive that process... I think the resulting video will be pretty good.

The artifact I'm working with is not a 'dialect' like feature, such as Fran Drescher's Shtick, so I think it can be improved and be less noticeable.

I'll see if I can setup something to post if I have time.
 
Sounds good... Yeah, actually, "comb-filtering" is an artifact that you generally don't want.
Has to do with the thinning-out or hollowness of a sound due to phase issues. If you're recording
voice with one microphone at a time, and you're not doing any fancy doubling in your computer
audio workstation, you shouldn't have to worry about it at all.

GJ

PS-- Provided you are recording your vocal track outdoors, or in a room that doesn't have a lot of nasty
"echo-like" reflections...
 
A good de-esser with full parameters is a VERY powerful tool, that also requires a bit care in using, and is not limited to just the sh or ch sounds some can be adjusted all over the frequency spectrum. and basically acts like a frequency specific compressor that reacts only to a certain frequency range and bandwidth leaving every thing else untouched as well as reacting to that frequency range set dynamically by the parameters set giving you the best chance of it sounding natural.
but anyways. good luck on your project, please let us know how it works out and possibly post it so we can hear how it went.

Cheers.
 
j1clark@ucsd.edu said:
As for 'de-essing', there's a bit of that, but more noticeable was the substitution of 'sh' for 'ch' in 'chopper'... sounded like 'shopper'...
Click to expand...

The general "fix" for this (I hate to say it) is to edit so it sounds like Chopper again. You can maybe find a "ch" sound and cut that in or maybe just cut the lead in "shh" so it's more "ch" like. I know that was just an example but that kind of surgery is what a dialog editor would do to the track to "fix" the problem.

It's a LOT of work if there is much to fix.
 
Noiz2 said:
...that kind of surgery is what a dialog editor would do to the track to "fix" the problem.

It's a LOT of work if there is much to fix.
Click to expand...

Good thing I bought that Purcell book at the last NAB show... I'm waiting for the picture editor to give me the locked film...

But yeah, for the exercise I'll see if I can put the ch back in chopper...
 
Is her nasal sound caused by a blocked nose and does it occur at certain seasons during the year when the pollen is in the air?

As a kid I got a broken nose in a football game and have a deviated septum.

So yrs ago I suffered from this during my stint on the breakfast prog. on 2UE here. Gargling with salt and warm water night and early morning helped but the best remedy was Manuka honey from New Zealand.

It's sold here in different grades, plain up to +25, which is very expensive. I get 'plain' from our local store here. 1/2 table spoon 3 times a day fixes everything.

Hospitals are putting it in bandages / try a Google. HTH.

Cheers.
 
Here's a short clip of the audio with the video. The mic was an Octava MK-012, Cardioid capsule, about 2.5 feet in front of the Wife, just outside the frame top right corner.

 
Some people are born to sing...some better never sing...

Some people are born to direct....some better don't

Well and some people are born to be THE VO for every docu, story, radio...and some...Well...don't have the voice for it..
 
And another thing, but it's not about the nasality, but anyway... If people do like "aah ?? What ?? uuh ???" It's because they didn't get the answer the first time.

Like here (if it's you of course) you ask her not to talk... Then she does like "uuh ?" (like "What did you say ??"), but there you think she likes to talk and you don't know anymore
what to do because you don't want to make her angry so you say "aaah, just do what you want" ?
 
I tried something and if you cut the low's out (EQ) it's better... I'm able to understand her and as not being native english...
 
dre83 said:
I tried something and if you cut the low's out (EQ) it's better... I'm able to understand her and as not being native english...
Click to expand...

Well, this was a test so, my 'direction' was pretty much in the direction of 'say anything'... The entire clip is about 15 minutes of monolog about the typewriter... sort of like getting a stone rolling down a mountain side... once it starts... forget about stopping it...
 
You must log in or register to reply here.
Back
Top