Using Waves Vocal Bundle plugs to zap mouth clicks

[mixaudio]

Hey David, I just finished an audiobook project where I had the same problem. I edited out the clicks on the first chapter manualy, and then I just realized it was gonna be impossible to do the entire book the same way. Then, I worked with a combination of Digi's Gate, and Waves X-Noise, X-Click and X-Crackle, "cascaded" in that order. Because of the huge amount of plug-in delay induced (and still no compensation in PT) I decided to bounce each chapter separately with this processing printed to the audio, and then take them to another session for adding SFX and doing final mix. In that session I also added Renaissance EQ, DeEsser and Compressor (in that cascading order).

The X-Click and X-Crackle did a prety good job on removing the most obvious clicks and even the majority of the more subtle ones. It took some time and sampling from different places to get a good setting that wouldn't mess with the natural transients and crispiness of the talent's voice. I'd say these two plug-ins reduced the clicks up to a very comfortable listening experience. I'd say the X-Click took care of the more obvious ones, while the X-Crackle worked the more subtle ones. Based on that, I decided to place the X-Click first in the chain. The X-Noise was targeted more to remove some air conditioning noise in the background, so I wouldn't say it helped with the clicks, but it was placed before the X-Click and X-Crackle to have a "cleaner" and "air conditioning reduced" track before removing the clicks. Again, I'd say the most important thing to be careful with X-Click and X-Crackle is to find a setting that won't mess a lot with natural transients and crispiness. My settings wouldn't necessarily work right for your voice track, but I'll put them here: X-Click: Tresh: 43.0, Shape: 64. Then, X-Crackle: Treshold: 37.0, Reduction 62.0. One thing that helped a lot on finding these settings was to monitor the audio using the "Difference" button on both the X-Click and X-Crackle plug-ins. This will allow you to evaluate how much of the original sound is being affected and to isolate and listen only to the actual clicks being removed.

You can hear some of this from the book's website at The Ole Degree book website . Check both the intro and the 0 degrees first chapter. The Warning to the Listener has the voice of the author, which wasn't processed with the same setup described above.

I do agree with the fact that talent should be aware of these things while listening to themselves in the headphones, but the truth is that sometimes you get to do the job with people that has the talent to read well, a great sounding voice... but no experience at all, and you don't the chance to fight your producer or director about it. The project I just finished would be a good example. This girl did a great job, considering it was her first time in front of a microphone. She learned a few tricks and became aware of a few things during the process, so, by about half of the read, she started doing a more concious and professional job.

One trick I've read about is having the talent to chew some apple pieces. I'm not quite sure about whether green or red apples would work better, but apparently, the acidity of the apple helps somehow to prevent the saliva from becoming the problem.

David, if you have any questions or comments about this setup, you can also email me. Hope this helps!

Rodolfo F.

[mixaudio]

I wanted to share a little more about the use of the Gate. Besides setting the Gate plug-in to not close completely, but only to about -14dB, I also did the following: created another mono audio track. Duplicated all my edited audio onto that track. Then shifted it back only on the duplicated track to about -50 milliseconds. Set the output of that track to an internal bus (say bus 9). Then, set the gate's key input to that same bus. This would allow me to set the gate with a slightly slower attack, without sacrificing or getting the beggining of sentences truncated, as the duplicate track would produce the opening of the gate ahead in time. On that same "keying" track I also added a compressor and eq. The compressor had drastic and exaggerated settings to make the slightly lower phrases loud enough to open the gate and make the "keying" work better. The eq had a curve emphazised on boosting the low-mid thru mid-high freqs, and filtering out the low freqs (under 90Hz).

Peace for everyone.

Rodolfo F.

[mixaudio]

David, did you have any luck with your track cleanup?

Rodolfo F.

[David MacNeill]

Not really. The minor reduction I get with the Rchannel preset Waves Support sent me is not strong enough. If I make it stronger, the voice sounds weird. I guess I need to buy X-Click and X-Crackle so I can use some of your excellent suggestions. We're already planning to employ the low-tech "apple solution" to our next book recording, so I'll report on its effectiveness when I know.

I really appreciate everyone's input on this troublesome issue. I'll do almost anything to avoid editing this stuff manually!

[mixaudio]

David, if it is of any help, I could email you a few mp3 samples of "before" and "after", using the processing scheme I described on my previous post. You could use this to compare if the clicks I fixed happen in a similar way to yours, and then decide if it's worth getting the X-Click and X-Crackle plug-ins. You can email me at [email protected] and I'll reply.

Rodolfo F.