How do I remove breaths from long recordings to save time?

[Girlagogo]

I'm using Adobe Audition CC (2018) to record a 5000 word corporate Explainer voiceover.  I normally just remove breaths individually using either the insert 'silence' or delete method.  However it will take far too long to do this on what will be an approximate 30 minute recording.  Is there a way that I can remove or reduce breaths in one go? Also, secondly  would this also work for mouth noises/ clicks? I'm quite new to using Adobe audition and voice recording and although I know how to use the Noise Reduction capture and process functions, fast breath removal is a new one on me. Any tips/ pointers would be appreciated.  Please note that I am not in any way a sound or music engineer so simple non-tech language works with me

[SteveG_AudioMasters_]

You can't really remove breaths automatically - and by and large you shouldn't be, anyway - especially in something that claims to be explaining something.

The reason for this is gloriously simple; it's during the breaths that people assimilate what was in the previous sentence, and breaths are an important part of this. Put simply, removing breaths will reduce assimilation by your audience - probably the opposite of what you intended...

On top of that, breathing is a natural thing to do - everybody does it... so we are accustomed, when we listen to people in real life, to hearing them breathing. Remove that, and it sounds plain unnatural. If you want somebody to explain something, then let them do it naturally - it always works better.

As far as voice clicks are concerned, only remove them if they're really annoying - sometimes they are, especially if there are a lot of them. But once again, there's no way to automate this, I'm afraid - they are all unique.

[ryclark]

Possibly use an Expander on your vocal track to reduce the level of the breaths rather than removing them all together. Then the natural gaps remain the same but the breathing isn't so loud.

[Girlagogo]

Thanks for the tip  ryclark. How do I use the expander i.e what is the correct function or menu location on Audition?

[Girlagogo]

So it's ok to leave breaths in on an Explainer voiceover? I know that for alot of commercial work it's expected that breaths are removed.  I'll need to work on my breathing technique if that's the case.

[SteveG_AudioMasters_]

Girlagogo  wrote

So it's ok to leave breaths in on an Explainer voiceover? I know that for alot of commercial work it's expected that breaths are removed.  I'll need to work on my breathing technique if that's the case.

It makes no difference what the intended purpose is - it's all a matter of making dialogue easier for the listener to understand. The more you sound like a human being, and not a machine, the easier that is. Breaths also act as a bit of a 'signpost'; intuitively the listener knows that there will be a short pause. Obviously you don't want to make a lot of noise about it, and that's where the expander tip may come in useful. As for using it, you'll get nearly all the information you need from this:

The control I've put a ring around is the one you may need to adjust. The threshold needs to be at a level between speech sounds and breathing sounds, and you will almost certainly have to experiment with that to get it right for you.

[SteveG_AudioMasters_]

Of course the other thing I should mention in conjunction with this is that if you use any form of compression on your voice to get the overall level up, it will inevitably make breathing sounds much more obvious. The solution to this (which works for pretty much any properly recorded piece) is not to use compression with its 'pumping' artifacts, but simply to limit the peaks, and that's what the lower half of the expander is there for. The threshold for that is set much higher, so that it only limits the peaks, and leaves the rest of the voice alone. If you use that, you will have to normalize your file afterwards, as there's no gain make-up.

With a bit of practice, it's possible to get pretty good results from this with most voices. There is also another tool in Audition which will do the same thing (only with rather more detailed control) but that's harder to set up, and for most purposes the Dynamics tool is fine.

[yvesg35131798]

Dear Steve,

May I TOTALLY disagree with you on this?  I am recording audio books and I do need to replace all and every breath sounds from all my files in order to produce commercial saleable audio books. 

I might be wrong, but even Audible is requesting from their suppliers to have -60db as a noise floor level.

But anyway, the need is there : to replace all breath sounds with silence.  Right now I am doing it manually by selecting every breath sound at any given lenght and replacing it with a silence using a keyboard shortcut that I have created.

If there were a function in Audition, like the one we have in Microsoft Word : search and replace, I could save hours of dull work!  Is this possible to replace in a faster way than mine those breath sounds (that I do need to replace)?

Thanks! 

[SteveG_AudioMasters_]

Of course you and anybody else can disagree - that's your perogative. Whether you actually need this approach - well that's not actually debatable, as you clearly don't. I mean, are you seriously trying to tell me that people learn better with all the little signposts and natural sounds of a person talking removed? And do you have any measurable evidence for that?

That may be the best joke I've heard all week!

[yvesg35131798]

Hi Steve!

I am not trying to persuade anyone about the learning easiness associated with removing the mouth sounds.  

I am just saying that for audiobooks I need to identify and replace all mouth sounds (specially breathing sounds) from a file in a more automatic way than with just a manual procedure. 

Listeners to audiobooks prefer not to hear those breathing sounds and that is why I need to replace them with silence.   And I do not argue with customers' preferences!

Anything to suggest about this to do it quickly in a more efficient way?

Thanks!

[SteveG_AudioMasters_]

Sorry, but I don't believe you. If you do a search relating to whether people prefer natural-sounding speech or having it processed for no breathing, I can't find a single reference to anybody prefering the lack of breath sounds at all, and a lot of references to people saying that you shouldn't do that to speech. Also there's some academic research that indicates an actual preference for breath sounds, although the authors want to do some more research on that. But hey, let's take what ACX have to say about it:

'Each uploaded file must be free of extraneous sounds such as plosives, mic pops, mouse clicks, excessive mouth noise, and outtakes.'

Nowhere does it mention removing natural breathing sounds. Plosives and mic pops - yes, they shouldn't be there. Excessive mouth noise shouldn't be there either, and that's an issue for your talent to deal with - but that's not breathing. I don't know why they bothered mentioning outtakes - who'd submit those? Breathing, though, is a part of the natural delivery of the artist, and is an essential component of a good read.

The only 'legitimate' use of what's actually called de-breathing is in commercials, apparently to stop listeners being even slightly distracted from a short commercial message, but that's it - nowhere else is it appropriate.

[yvesg35131798]

Steve,

you are still arguing with me about the non necessity of removing breath noises and this is meaningless to me.

My  request is simple : are you able  to give me a technical procedure to do so (or not) in Adobe Audition?

If you cannot, please say so and please stop arguing with me about what for me is a need even if you think there is no need to do that.

It is possible to do so in Audacity.  Is it also  possible in Audition?

Thanks!

[SteveG_AudioMasters_]

If somebody else feels that they are able to tell you, then they will, but I doubt it. As a matter of principle, I'm not perpetuating bad practice, and I would have thought that was pretty obvious from what I've already said. Is that clear enough?

[yvesg35131798]

Steve,

I have visited different publishers of audio books web sites, and many of their audio books do not have any breathing sound or mouth sound in them.  It is obvious from you what say that, according to you, this is bad practice, right?  Ok.  No problem.  You are clear on that. 

So, if I want to quicly find the «silent or breathing» spaces between different parts of a sentence and replace them with a silence, what you say is that there is NO way in audition, according to you, to do so, right?

Then, I do still have to manually select each and every of those spaces and replace them with a silence, right?

And Steve, I might be annoying to you, or look dumb or hard headed or whatever you want, but if there is a way in Audition to accelerate what I am doing manually, I am more than willing to learn it because it would save me hours of dull work.  So any help in this regard would be GREATLY appreciated.  Thank you!

[yvesg35131798]

FINALLY!

I HAVE FOUND THIS!  

And this is EXACTLY what I am looking for!  ...with no arguing over the need!

OK I think I found the answer! There was a very good video on YouTube here: Adobe Audition CC 2018 - New Noise Gate Effect - YouTube​

Hi Ken! Thanks for sharing my YouTube tutorial on this and glad it worked for you.

A summary of the process:

1. Window > Amplitude Statistics.

2. Highlight the noise floor (background audio with no speech).

3. Click Scan Selection and make a note of the Peak Amplitude.

4. Effects > Amplitude and Compression > Dynamics... tick AutoGate and type a Threshold 1dB or so higher than the number from step 3 and you're done!

If you find your speech is getting cut off at the start and end just increase the Hold setting a little.

[yvesg35131798]

Dear Steve and dear Audition users,

I have found this tutorial on youtube explaining how to meet ACX requirements with Audacity.

https://www.youtube.com/watch?v=wnutKoBzmpA

I have followed the instructions :

1. I installed Audacity, and the two plug-ins : check ACX and RMS-Normalise,

2. Followed the three simple steps explained in the clip (The instructions are on https://wiki.audacityteam.org/wiki/Audiobook_mastering)

Effect > Filter curve... > Manage > Factory Presets > : Low roll-off for speech > OK.

Effect > RMS Normalize: Target RMS Level -20dB > OK.

Effect > Limiter: Soft Limit, 0.00, 0.00, -3.50dB, 10.00, No > OK.

3. run the plug-in : Analyze > ACX-Check.

Boom!  Diagnostic : My file meet the requirements!

DONE! PERIOD!

[jstead1]

Thank you for sticking with it and coming up with the answer. What a great help! This is necessary for TTS (Text to Speech) recordings.

[Daniel Gimness]

Why would it be necessary for TTS recordings?

[defaultk5c3h1h0t33j]

OMG! I FREAKIN' LOVE YOU!!! 

I had this exact question (great question, by the way). I tried exactly what you said here, and it worked PERFECTLY! You just made my life so much easier ❤️ Thank you!

[jfrank at CCR]

I've always used a gate on my studio mikes to cut off room noise, and I didn't really think about Audition doing it post-recording. I'll try it next time I cut a spot. Thanks for persisting with your inquiries.

[Jordan Scribbles]

You are a lifesaver. Anyone who thinks it's unnecessary to find and reduce/remove all breaths has never worked with several hour+ long recordings where the speaker has an involuntary whistle in their sinuses. This saved me sooooo much time!

[ffggghhhhh]

iZotope advanced has breath plugin...

[dougwylde]

Steve you are wrong..... Adobe Audition 3.0 has feature to delete silence unders cetain predefined conditions  EX: peak -33db for X duration.  

I used this feature to to delete BREATHS in longform audio.....

[SteveG_AudioMasters_]

Just because a feature is present doesn't make it correct to use it - as you have demonstrated.

[dougwylde]

Wow... so called experts are a little hostile... I used a feature in audition 3.0 to delete silence under "designated decibels" worked great.  I used it on longform audio to take breaths out... YES,  STEVE. BREATHS!