Manual De-essing?

Nola

Well-known member
Does anyone do this?
I've been experimenting with it lately and like it a lot better than a de-esser. The De-esser doesn't seem to catch all plosives and lispy lines. You can also pull the clip down various DBs instead of having the de-esser at a fixed db reduction. I really like it other than it's labor intensive. Curious to hear some of the pros' opinions on doing this manually! Thanks
 
Yeah, doing it manually certainly isn't unheard of and, a lot of the time, it's the better approach.
Sure, sometimes just a little tickle is all that's needed, but sometimes getting in there and precisely dipping what you want, and only what you want, gives the better outcome.

I tend to do manual edits for sibilance, if necessary, because I don't like that idea that something might be acting when it's not necessary.

Like a noise gate for gaps and pauses. Sure, you could use one and make sure that it's only acting when you want, or you could manually cut out the silences. I generally do the latter.

If there's a particularly troublesome section or passage I might use a de-esser, but then I'll probably automate the threshold or bypass so it's only on for that passage.

Edit. Oh sorry, you wanted pro opinions.....<steps aside> :p
 
I think in the digital world, a lot of what use to be necessary isn't any longer. Like the one post for fader riding as you mix down tracks. For someone coming from an analog world, that kind of control is hard to imagine or just ins't the way one thinks. Same way with compression, noise gates etc. Lot of ways to correct the problem now verses then.

I am trying out Waves vocal rider, technically it is kind of a compressor, but it isn't. What is nice about the VST, you can record the what it does, make your corrections, and really, not have any compression on the vocals. I am not saying the VST is good or bad, just to the point, in the digital world, there are more precise options. Like removing sound in the wave verses using a gate.

Good idea Nola
 
Not a pro. Normally de-essing is unnoticeable. Volume riding is also normal. Ride before de-ess. Don't de-ess in solo.
 
Always do. You can spot them very easily. Very dense blobs on the waveform.

Yep, exactly, and more things are blobs than you'd think. It's not just ess or shh but many other sounds. I think a de-esser might miss some of those.
Saves money too if someone is considering buying a 3rd party de-esser.
 
I use a de-esser but with moderately aggressive settings, then for extreme sibilance I'll bring that down the rest of the way as necessary, manually. In other words... I like the results using the de-esser most of the time, and I only need to manually adjust a few times over an entire track. Saves time. I used to do it manually, but not anymore.
 
Does anyone do this?
I've been experimenting with it lately and like it a lot better than a de-esser. The De-esser doesn't seem to catch all plosives and lispy lines. You can also pull the clip down various DBs instead of having the de-esser at a fixed db reduction. I really like it other than it's labor intensive. Curious to hear some of the pros' opinions on doing this manually! Thanks

Sure. Good question. Here's my pro opinion (lol) on how this is best done manually. A spectral editor lets you attenuate certain frequency ranges and plosives will almost always occur below 400 hz. The one in this first example lets say that circled region is a plosive from a gush of wind into a lapel microphone on an outdoor bridge occurs from 13.00.5 to 13.01.6. To manually edit this out, you draw a loop around it with a lasso tool, telling the editor to only affect that particular frequency at that particular time. There's a de-plosive preset which can attenuate it, or you can use the spectral repair tool. What this does it looks at what is above, below, before, and after, and re-writes the wave based on what is surrounding it. Very similar to the 'wand tool' in photoshop. The third image is over exaggerated for illustration, but it shows what the new wave form will look like when the plosive is drowned out an re-constructed based on surrounding data.
 

Attachments

  • 1.png
    1.png
    1.6 MB · Views: 8
  • 2.png
    2.png
    1.5 MB · Views: 5
  • 3.png
    3.png
    1.8 MB · Views: 10
Back
Top