I would take it out in the waveform, but I spend a lot of time taking sounds out of voiceover stuff, so it comes as a second nature to me!

Feverdream wrote: I would take it out in the waveform, but I spend a lot of time taking sounds out of voiceover stuff, so it comes as a second nature to me!

Would that work though on a live mix? Wouldn't that risk the integrity of the rest of the audio. I don't know as I've never tried to surgically remove errors like this.

You could try a very VERY tight bit of volume enveloping (or limiting) applied to only that part of the audio. As fast an attack and release as possible, and go easy on the reduction.

Sometimes I just zoom in massively on the problem on screen and then automate a radical, split second volume change - suddenly drop the volume just for the offending sound and then bring it right back afterwards. If there are other things going on in the background and you do it fast enough, deep enough (but not too deep), and in just the right spot, it seems to work.

The method outlined by apstrong works well for me too - on any unruly peak that isn't clipped, as a matter of fact. Snare, vocal, whatever.

And he's right about being careful. You have to get it right (depth and width in particular) for it to sound natural.

It's not something I'm quick to do, but if you've got a random peak or two that's problematic, this is a great way to deal w/ the situation and avoid issues w/ having your compressors work the way you want.

Another option (the above explanation just led me to this) would be to use a hard limiter w/ a threshold set high enough that it ONLY touches the peaks on the "t" sounds. While not as precise as drawing on the waveform, it's quicker.

And then there's the question of whether you should use this kind of surgical digital editing on a LIVE recording.
Beyond saying that if it isn't presented as live, it doesn't matter - I won't touch that.

It's a bit of a PITA to do it a lot, but sometime I get tired of fighting to find the right level of compression that gets rid of the peaks without mushing the dynamics. Compression is hard. Unless it's soft. Either way, I find it difficult to get right. Or left.

Hueseph's concern is a good point - this only works if you have good separation on that track. If there's lots of bleed you'll hear the change in more than just the vocal, or whatever it is that you're adjusting.

Thanks guys I'm gonna give enveloping a go I think, I'll let you know how I get on.

I'm not using any compression really only a light one if any, obviously with these acoustic recordings you start messing too much and it stops sounding "live" and starts to sound over processed.

audioangel wrote: I have recently recorded a christmas cd for a school choir. During it I recorded a track with a soloist. I used an AKG 414 on the vocalist so I could bring her up in the mix (I did "live" mixes) and on this track she completely overpronounces a "t" try as I might I haven't been able to eq it out - does anyone have any ideas what I might try? I'm reluctant to start playing about with the waveform.

Hi,
Maybe you could also try use an editing program with a spectral editing
feature. It will show you the offending "k" and you can just select and edit that specific frequency ! Try out a trial of such a program.

Dirk Zuber

spectral editing sounds good - any recommendations for mac?
I'm using Logic Studio 8.

i'm not having a lot of luck with enveloping the "t" is quite long and you can hear any sort of change no matter the slope! - i wouldn't fuss so much but it sounds so unruly and loud! There's one thing I might try... on the previous take she doesn't sound a "t" at all in that position, I might try mixing it into the track assuming their timing is close enough

Try Soundtrack Pro 2, it came with Logic Studio, try that. I've had success with cleaning up video audio and it should provide all the tools you need.

Check out the videos: http://www.apple.com/finalcutstudio/soundtrackpro/