I'm doing some mixes for a very sibilant singer. I knew it was going to be a problem when recording her, but I still didn't managed to adjust well enough during tracking, so now I'm "fixing it in the mix."

Anyway, I'm short on cash (and I'll be making less than $100 on the project). If someone knows of a very effective, low-cost plugin, I would be very appreciative. I have Spitfish (part of the Fish Fillet collection), but it's not doing much to the track. Because I'm trying to follow good digital gain structure protocol, I'm trying to mix at about -12 average level. At that level, when I max Spitfish, it just barely starts to work on the track. Also, I find that its focus is too narrow (and there doesn't appear to be a way to adjust that). I need to hit everything above 3500 Hz.

Anything come to mind? By the way, I just have to comment that the de-hisser in Samplitude 7 Pro is shockingly lousy. Most of their default effects are at least decent, and I love the convolution reverb (really, really love it), so it's a bit puzzling.

I've tried every software de-esser under the sun that's available for PC, and it's been my experience that Spitfish kicks the crap out of all of them.

I don't see why you need to keep the vocal track at -12. I'd bet if you just normalize it, Spitfish would do fine for you.

If not, there's always the tedious old manual method.

Chris

What Bandini said...Spitfish is cool and works well. The sidechain monitor is very helpful and the price (free) ain't too bad either.

-john

I didn't really like spitfish, but YMMV. HEre's another free one I found on KVR's website that you can try. Look down at the VST effects and it's in there. You have to download the bundle it looks like though. They're free. Anyway, worth a shot:

http://www.mda-vst.com/

Good luck. FWIW I use the waves de-esser which I like pretty good, but it'll be a little out of your cost/benefit for this particular project though.

Thank you for the responses. I will be trying everyone's suggestions this evening.

Chris, I'm not committed to the "-12 as digital zero" concept at all costs, but I do like to work with that gain structure. When I switched over to digital recording, all of my initial recordings sounded terrible. I couldn't understand what had gone wrong. Based on someone's suggestion, I stopped trying to push things as close to 0 as possible, and it has been one of the few things that has immediately improved my abilities in the digital domain (I still have a ways to go).

Thanks again, all.

Hello James. I am also a Sam user. Don't try to use the de-hisser for de-essing though! Try the multyband compressor. This works GREAT and is very flexible. Use the solo function to dial in just what you need.

Good luck
Jeff

quote:

Originally posted by Jeff Sers:
Hello James. I am also a Sam user. Don't try to use the de-hisser for de-essing though! Try the multyband compressor. This works GREAT and is very flexible. Use the solo function to dial in just what you need.

I was just thinking the exact same thing.... I'm also a Samp user and the dehisser is used for removing "hiss" sounds, not for removing siblance, which would definitely explain why you're not having much luck with it. The two are very different funtions.

Like Jeff mentioned try using the multiband comp. Try using three bands to split up the high frequency ranges you want to de-ess and it should make the de-essing more seemless.

You could also try turning the volume up and use the spitfish. That is a great de-essing plugin.

-tkr

quote:

Anything come to mind?

Put on a pot of coffee and manually 'draw' down the "S"es using the automation in your DAW app. Or manually snip all the 'S'es with your scissors tool or equivalent ('snap to zero crossing turned on!), carefully select all of them and scoot them down to a separate track (most apps will have some sort of function key to hold down while doing this so the selection can be moved vertically to another track but not horizontally in time), and then you can automate and/or process your 'sibilant/consonants' track separately from the 'main/vowels' track.

Tedious, but perversely fun, and no $$ investment other than your time.

What chris just described is exactly what I've been doing lately. Tedious as hell, but it never takes as long as you think it will.

I always try Spitfish first, though.

Chris

With Samplitude's object editing you don't have to move it to another track. Just cut on the objectionable consonants and reduce them.

...still tedious, though...

Thanks for the suggestions. I normally do manual editing of sibilance (with my own music). I'm a fairly sibilant singer, so I'm very familiar with the problem. Unfortunately, I just don't have the time with this project. By the time I would finish all of the editing, I would be out of budget to do the rest of the mix. Of course, I've never treated my engineering as a business (probably need to change that if I ever want it to be my day job), but there is also a deadline on this one, and unless I decide that continuing my trend on not getting sufficient sleep this week is the right thing to do, I'm out of time.

However, I still have a few songs to mix for her later, and I'll definitely give both of those approaches a try. I've used the multiband compressor for some basic, untrained mastering, and I like the way it works, especially if you use the less-is-more approach. I'll have to try it on the sibilance.

As for these particular mixes, I finally wrestled with levels and Spitfish and made some improvements, but I'll certainly be paying strict attention to mic placement if I ever need to record this artist in the future. She is very sibilant.

Again, thanks to all. Great suggestions.