Sibilance reduction?

[bssistmurf]

I hope I'm in the right area with this question. I am mixing a spoken lecture cd for a friend and her s, t, snd sh sounds are seemingly untameable short of manually addressing each one individually.

I have reduced the offending sses some with eq, and deess plugin, but they are still pretty agressive. I have tried applying multiband compression to try to hold back the band of offensive frequencies, but I cannot get a natural sounding result.

It seems that her s, t, and sh sounds are just full of 5k to 12k. When I analyze the wave I see the same thing I hear, a hot band from 5k-12k.

Surely someone here knows what I am missing. It must be a simple oversight or hole in my knowledge. Can anyone shed some light on this for me?

Thanks,
Murph

[CosmicDolphin]

Well if you de-essed it already...??  Not sure what more you can try other than set up your own de-esser by using a sidechain compressor keyed to the problem frequency....but that's not really anything different ...might be able to zone in on the problem area a little more than just a de-esser plug.

Why not post a clip ?  Then we can all have a try at it ?

CD

[detune]

You already indicated you probably don't want to do this, but I manually reduce mine in a wav editor.  I've learned what the waveform of the sibilance looks like, so I just zip through and highlight them and do a gain reduction for each one.  It's really not that bad. 

I can't get de-essers to do anything at all, unless I set them high enough to affect ALL the sound. 

[detune]

Oops, just saw that it's a lecture.  Guess my technique's not going to work!

[bssistmurf]

Thanks for the replies. I'll post a short track CD, thanks for taking a look at it. I might take a shot at a sidechain key. I realized that her voice seems to have allot of those frequencies in it all the time so the s doesn't seem to really spike a wave, it just stands out. I guess that's why I can't get deess or comp to hit the esses without affecting the whole thing.

I can get good results going manual similarly to what you suggest detune. Instead of reducing the level of the highlighted passage I invert and mix it back in. This gives great results, but a 45 minute cd has allot of esses!

Thanks for the help.

Murph

[meekofnature]

Deesser should get it if it's a decent one that is set properly.  The trick is finding out what frequency her sibilance is at.  I'd find a really bad section or two and just loop playback that one word while playing with the desser settings (frequency and threshold are the two biggies) and see if you can make the siblance quieter.  Once you get it knocked out hit playback and adjust the threshold until only the esses are being ducked.

If your daw allows this you can get into manually creating a deesser.  Basically you duplicate the track and put a lowpass and highpass filter on, and basically try to get it to where the only frequency you hear is the awful one...you can also play with boosting the problem frequency.  Don't worry, this is just a tool and you won't ever hear this track.  Then you go back to your original vocal track and put a compressor on with a key input(if the compressor doesn't have this, then this technique will not work).  Then you have to make sure the output of your sibilant track is routed to the input of the key.  The procedure to do this changes based on which DAW you use.  It takes sometime to set this up and it can be pretty confusing at first, but the idea is that the compressor only kicks in when it hears that sibilant sound. 

Deesser's have to be used with caution.  There is usually some sort of price to pay with using them.  The worse the sibilance, the harder it is to eliminate the trouble frequency without removing wanted high-end whenever the deesser kicks in.  If this one track is really that bad it's going to be tough to get it to work.  I tried downloading the file but it came up silent for me.

Good luck, and let me know if I can help.
Marc

[NickT]

Yeah...what he said!

I like this free de-esser

http://www.digitalfishphones.com/main.php?item=2&subItem=5

It's the spitfish.

and she really does not sound bad.

Nick

[bssistmurf]

Thanks guys for chiming in. I tried setting up a track cutting all but the most offensive frequencies to use as the trigger but I wasn't getting a good result. I'm getting used to CoolEdit and some plugins there that I have never used before. I may have to take the files down to my PT machine and load them in there to do it. I'm sure I can get a much better result with the software I am more familiar with.

I'm going to try digitalfish first.

I don't know why the track came up empty. I just downloaded it and it played fine in software other than that used to edit and save the file.

Thanks,
Murph

[NickT]

Played fine for me. 

[TallPaul]

I've found a pretty nice way of reducing sibilance and plosive sounds BEFORE they get to the mic. I used to just use a screen, but I also like to sing close to the mic, as it gives vocals the best depth. But obviously singing close causes additional problems. What I tried (and it works great) is to stick a feather duster between the mic and the screen! The feather duster diffuses sibilance/pop without any real detriment to the vocal quality. Try it and tell me what you think!
Paul

[CosmicDolphin]

I didn't think the sibilance was that bad on the audio clip, I was more bothered by the background noise....something whining ...probably the PC ?

CD

[bssistmurf]

I didn't think the sibilance was that bad on the audio clip, I was more bothered by the background noise....something whining ...probably the PC ?

CD

Yeah, these tracks were given to me the way you heard it. I have been able to weed out some of the other noise, but when I do the sibilance seems to get worse. I'll post my latest processed copy of the same file. Maybe I am just getting hyper-sensitive to it. Let me know what you think. I haven't had my gear fired up for a couple of years so it almost seems like I'm learning all over again. Feel free to point out any mistakes you hear in the work I've done here.

Thanks for all your input!

In the short attached clip the first half is my cleaning. The second half is the extra sibilance removal I did with phase cancellation. Maybe this will give you more insight into where I'm going astray.


I'll have to try the feather duster trick. When I can I like to get right on the mic (or have the vocalist I'm tracking right on the mic). Another trick I learned years back that worked pretty well on stick type mics (like a 58 as opposed to something like a Neumann 103) is to point the diaphram down at a 45 degree angle and have the vocalist have the tip of his nose at the screen. That way the wind goes past the element instead of into it.

Thanks for all your help!

[TallPaul]

Another trick I learned years back that worked pretty well on stick type mics (like a 58 as opposed to something like a Neumann 103) is to point the diaphram down at a 45 degree angle and have the vocalist have the tip of his nose at the screen. That way the wind goes past the element instead of into it.

I've tried that myself, and it works. John Clark told me to hang the microphone like that and point it away from the vocalist. The only problem I have with that is it just doesn't get the same 'tone' for me. If it's a rock tune, maybe no big deal. But on soft jazzier stuff, the 'breathiness' is important to me.

Paul