KVR Audio

BERFAB · Post by **BERFAB** » Mon Apr 30, 2012 4:57 am

I'm recording a series of spoken word recordings (my own voice)which will ultimately be distributed in MP3 or similar format.

I'm generally happy with the overall sound--i.e. no real sibilance issues or pops. However, I do notice the occasional breath noise (drawing breath between phrases) and was wondering what FX and settings I would use to minimize or eliminate these.

Using a few of the presets in Ozone 5 (basic version)for the compressor and EQ, I'm able to come close, but I'm really just punching around in the dark. I'm sure there's an actual science to this and if anyone could explain it to me (or illustrate it with an O5 preset--or any other plug) I'd really appreciate it.

Cheers
-B

AudioGuy720 · Post by **AudioGuy720** » Mon Apr 30, 2012 5:15 am

Use a gate or manually go through a mute the breathes. Or better yet replace them with room tone so there's not a digital silence gap (heard when listening with the volume up high).

Compyfox · Post by **Compyfox** » Mon Apr 30, 2012 9:01 am

I'd go for an gate -> EQ -> Compression route.

The gate could be used to turn down the breath noise with a "depth" or range setting. Therefore if you add further EQ and compression, you have less breathing noise.

I wouldn't completely cut the breathing - it sounds natural and pleasant if handed well in voice only applications. But it can be too prominent if you compress your voice reconding to a squarewave.

hcv242 · Post by **hcv242** » Mon Apr 30, 2012 11:45 am

leave the breathing, its part of speech.

thecontrolcentre · Post by **thecontrolcentre** » Mon Apr 30, 2012 12:13 pm

hcv242 wrote:leave the breathing, its part of speech.

+1

BERFAB · Post by **BERFAB** » Mon Apr 30, 2012 1:48 pm

I agree that generally, breathing is a natural part of speech, and I don't want to eliminate it completely. However, a few weeks ago, I was listening to Adam Carolla's daily podcast and there was some kind of glitch with their studio compressor and every breath was amplified (annoyingly so) but normal speech still sounded the way it should. The following day, the engineer apologized for the mistake and said he made some adjustments. So now, I'm maybe just a little over-concerned.

The gate idea had occurred to me, and I will definitely try incorporating that into the FX chain as suggested by Compyfox.

Are there any other compression or limiter tricks to look out for or be concerned with?

Thanks
-B

geroyannis · Post by **geroyannis** » Mon Apr 30, 2012 4:59 pm

The dynamics section of Ozone should be enough. Just use a small gate ratio (like 1.5:1) and mess around with it's threshold until you find the point where the breathing lowers a bit but the actual speech is not affected.

You don't necessarily have to use compression, and it will only make any breathing problems worse.

Compyfox · Post by **Compyfox** » Mon Apr 30, 2012 5:11 pm

There are no tricks, whatever suits your need best.

I've seen various techniques:
- compression only
- EQ then compression
- just EQ and slight compression
- no EQ whatsoever (since the voice was recorded with EQ already) and then only some stuff added like MaxxBass for more lowend

You can really do a lot with a normal EQ already, and a slight gate of you want to "automate" the breathing. Else, simply mess with the compressor threshold so that the breathing is unaffected but the voice is.

skitchy · Post by **skitchy** » Mon Apr 30, 2012 5:25 pm

If it only happens a few times, and you are 100% happy with the rest of the recording it is silly *not* to do it manually (Audiguys method of pasting over the bits you dont want with room 'noise' is the way to go). If you want to keep some of the breath then do a 50/50 70/30 80/20 mix of the breath+room noise (easier with a multitrack DAW). Or cut/paste the breaths on to a different channel and eq it.
Anything else is bound to intrude on the rest of the recording in some way at some point.

BERFAB · Post by **BERFAB** » Mon Apr 30, 2012 6:02 pm

All very good advice. Thanks guys.

@geroyannis, precisely what I was looking for, i.e. a suggestion of actual ratio settings as a starting point.

geroyannis wrote:The dynamics section of Ozone should be enough. Just use a small gate ratio (like 1.5:1) and mess around with it's threshold until you find the point where the breathing lowers a bit but the actual speech is not affected.

You don't necessarily have to use compression, and it will only make any breathing problems worse.

And, yes, the manual method may not be a bad idea either. Overall, I'm happy with the sound files so far. I have a fairly deep voice that is, I think, generally pleasing to the ear, so I'm really just looking for tweaks to maximize overall clarity.

Cheers
-B

Compyfox · Post by **Compyfox** » Tue May 01, 2012 10:32 am

Since I've been there with editing podcasts, editing out breathing is definitely a PITA.

Automation could work, like "cut noise below a certain level" , but this removes the natural feel of the recording. Breathing is(!) essential.

Again, it depends on what you want to shoot for. Here, audio examples to mess around with would definitely help.

Ja.x · Post by **Ja.x** » Tue May 01, 2012 8:36 pm

thecontrolcentre wrote:
hcv242 wrote:leave the breathing, its part of speech.
+1

+1

What FX settings for spoken word recordings?