How do you generally work with Plosives and Sibilance in SAW?

[DavidandMary]

Hey,

First off, this is not a Pro or commercial session. I just want to learn and help my sister in law with some music for her own website she will be building. Doing some female Vox recording in SAC/SAW for her. So, that is the background...

Now, how do you generally work with plosives and sibilance in SAW? Let's say you have a female singer. She has to sing the word "Basic" about 50 times. Her "B's" stick out with a pop and her "S's" are harsh. If I EQ, I cut a whole in that frequency during the clean parts too. I think that I would rather squish it with Bob's transparent Compressor at just those frequencies than to cut it. But, I do not really know how to do that yet. What would you suggest? Thanks in advance for your patience!

[tomasino]

Hey,

First off, this is not a Pro or commercial session. I just want to learn and help my sister in law with some music for her own website she will be building. Doing some female Vox recording in SAC/SAW for her. So, that is the background...

Now, how do you generally work with plosives and sibilance in SAW? Let's say you have a female singer. She has to sing the word "Basic" about 50 times. Her "B's" stick out with a pop and her "S's" are harsh. If I EQ, I cut a whole in that frequency during the clean parts too. I think that I would rather squish it with Bob's transparent Compressor at just those frequencies than to cut it. But, I do not really know how to do that yet. What would you suggest? Thanks in advance for your patience!

Avoid 'em all together.

Use a pop screen and turn the mic off axis ever so slightly..don't have the mic face her straight-on..

It's not really just a pop screen..it also controls the distance between her and the mic.

A slight rotation of the mic..one direction or another..will help both conditions a lot.

If there's still some sibilance..ask her to replace the begining of the 'S' with a 'Z' .. so it's more like 'Zzssss'.

After that I'd say the Sonoris MultiBand Compressor.
But you should be able avoid 'em altogether.

[DavidandMary]

Excellent Michael! I have a pop filter, but did not have the mic off axis. Giving that trick a go tonight...:)

[HapHazzard]

Digitalfishphones can help with the de-essing (free) (http://digitalfishphones.com/main.php?item=2&subItem=5)
http://digitalfishphones.com/images/screenshots/spitfish.jpg

The B's you'll need to check into a Multiband comp to suppress the pop or use a band pass filter to roll off the low end.

There is are many ways to use everything in SAW to do the same but I'm not that good yet. (but getting better)

Hap

[DavidandMary]

Good tips Hap! Glad to hear...

[DavidandMary]

One other thing that I just tried with a bit of success on the "B's"... I split the region at the plosive. Backed up the cursor a bit. Hit "X". The soft edges helped those "B's"...

[Cary B. Cornett]

For vocals, I almost always put the mic at about forehead level and just off to one side, with the mic aimed directly at the mouth. The talent faces straight ahead. If need be, I give them another mic, usually a dynamic, right in front to use as a "target" (second mic is not connected). Done properly, this completely eliminates the need for a pop filter.

The mic is at least several inches away from the mouth, and I have seen this done with the mic as much as a foot or so away. Sibilance should be less of a problem at that distance.

This method works best in a decent room. IMO the average bedroom or small booth does not qualify. I have heard professional-sounding results from in the (carpeted) living/dining area of a moderate-sized house.

[DavidandMary]

Thanks Cary... I will also try that one tonight as well...:)

[AcousticGlue]

Just do not ever park the plosives next to a building and run away! No seriously these were all good remedies.

[DavidandMary]

LOL AcousticGlue!:D

[Dave Labrecque]

One other thing that I just tried with a bit of success on the "B's"... I split the region at the plosive. Backed up the cursor a bit. Hit "X". The soft edges helped those "B's"...

First: do the mic technique stuff outlined above.

Second: If you gotta live with what's recorded...

1) SS's channel strip compressor set up as a de-esser is amazing. Much better Spitfish (or any other de-esser I've tried, other than the Levelizer, which is identical to the channel strip comp) in my experience.

2) For plosives create an automation gallery entry that consists of turning SS's channel strip EQ on and off with a 300-Hz-or-so hi-pass filter, positioning the on and off points on either side of the big, low-freq waveform (often seen as modulating higher frequency content), and at zero crossings. Either automation entry may occasionally require a bit of placement tweaking to work smoothly.

I've been using both of these exclusively and extensively on current VO projects and they solve my problems completely 95% of the time or more.

[DavidandMary]

:)Thanks Dave... Will do!

[DavidandMary]

First: do the mic technique stuff outlined above.

Second: If you gotta live with what's recorded...

1) SS's channel strip compressor set up as a de-esser is amazing.

2) For plosives create an automation gallery entry that consists of turning SS's channel strip EQ on and off with a 300-Hz-or-so hi-pass filter, positioning the on and off points on either side of the big, low-freq waveform (often seen as modulating higher frequency content), and at zero crossings. Either automation entry may occasionally require a bit of placement tweaking to work smoothly.

I've been using both of these exclusively and extensively on current VO projects and they solve my problems completely 95% of the time or more.

Hey Dave, I tried these and you are right. Very cool! I just found the EQ high and low that you have with the Levelizer. This is a whole new ballgame there. The new recording techniques can be done tonight so these tests may not be necessary. I think this is fun!:D

[Dave Labrecque]

Hey Dave, I tried these and you are right. Very cool! I just found the EQ high and low that you have with the Levelizer. This is a whole new ballgame there. The new recording techniques can be done tonight so these tests may not be necessary. I think this is fun!:D

You mean the high and low pass settings for the key input to the channel strip comp? Yes, just like the Levelizer, only no need to plug that baby in. :)

[DavidandMary]

You mean the high and low pass settings for the key input to the channel strip comp? Yes, just like the Levelizer, only no need to plug that baby in. :)

Very Cool! That is a great de-esser. Plus, it is clean which I really like...:)

[DavidandMary]

If you have a plosive, minor or major, and it's on a take you want, try this.
Do a 'K' cut right before the plosive and put a ramp up fade on the plosive.
With some practice, you'll find the right area to mark, cut, and fade up.
It takes care of most small plosives. The nuclear ones? Get another take.
I've tried 'soft edge' also. It works for the light ones.
If there's some plosive on the word leading into the cut, put a ramp or soft edge on that, too.
Light ramps and fades will cure most small plosive hits.

Thanks Stashu,

I just tried that. I had a bit of difficulty figuring the length. Kept shrinking and then undoing until it seemed natural. There is that whole in the sound from the beginning that seemed most unnatural. I will keep working on that technique though. I thank you...:)

[Dave Labrecque]

Or just do what a de-plosive compressor setting would do, only manually: mark the area of the plosive and drop the level by 5 or 8 or 10 or more dB.

Can do the same thing with sibilance -- manual de-essing with simple volume drops of the offensive ess sounds.

Ramping of the in/out automation can smooth these treatments, and make them even more like what an auto-de-esser/auto-de-plosiver would do. But such ramping may not even be necessary.

All that said, I think my prior suggestions are still usually better. These are just more options for your toolbox.

[Dave Labrecque]

Very Cool! That is a great de-esser. Plus, it is clean which I really like...:)

Other de-essers I've tried seem to affect the audio outside of the sibilance area (i.e., on the "good" side of the threshold setting). I've never understood why that is.

[Carl G.]

Digitalfishphones can help with the de-essing (free) (http://digitalfishphones.com/main.php?item=2&subItem=5)
http://digitalfishphones.com/images/screenshots/spitfish.jpg

The B's you'll need to check into a Multiband comp to suppress the pop or use a band pass filter to roll off the low end.

There is are many ways to use everything in SAW to do the same but I'm not that good yet. (but getting better)

Hap
I second both of Hap's excellent suggestions.
And as Dave mentioned - the inline compressor for dessing works well.
Plus I use this in my group of De-essing tools when the going gets tough:
http://www.db-audioware.com/dB-S-de-esser-more.htm

Choose from 3 algorithms for removing the sibilance: wideband (like a traditional studio de-esser), lowpass (attenuate the high frequencies of your choice) or narrowband (attenuate a tuned band of frequencies)

[DavidandMary]

Thanks Dave... The automation way you mentioned is definitely a skill that I need to evolve. Love the Ctrl-Tab to get to the next Automation point...

[DavidandMary]

I second both of Hap's excellent suggestions.
And as Dave mentioned - the inline compressor for dessing works well.
Plus I use this in my group of De-essing tools when the going gets tough:
http://www.db-audioware.com/dB-S-de-esser-more.htm

Choose from 3 algorithms for removing the sibilance: wideband (like a traditional studio de-esser), lowpass (attenuate the high frequencies of your choice) or narrowband (attenuate a tuned band of frequencies)
Appreciate it Carl. Good tips as usual...:)

[UpTilDawn]

2) For plosives create an automation gallery entry that consists of turning SS's channel strip EQ on and off with a 300-Hz-or-so hi-pass filter, positioning the on and off points on either side of the big, low-freq waveform (often seen as modulating higher frequency content), and at zero crossings. Either automation entry may occasionally require a bit of placement tweaking to work smoothly...

What works better in your experience?... automating the EQ strip On/Off, or the filter On/Off?

I often have difficulty with slight pops after this kind of automation and, as you pointed out, it requires further tweaking to get rid of the pops.... sometimes I have to give it up and go with another method altogether, depending on the material being edited.

DanT

[Bob L]

The easiest method for pops and sssssssss's is to simply mark the area of the pop or ssss and just pull the fader down xdb.... simple and clean and it keeps the original punctuation sound... just keeps it from expoding out in your face.

Bob L

[Carl G.]

What works better in your experience?... automating the EQ strip On/Off, or the filter On/Off?

I often have difficulty with slight pops after this kind of automation and, as you pointed out, it requires further tweaking to get rid of the pops.... sometimes I have to give it up and go with another method altogether, depending on the material being edited.

DanT

Dan... I addressed this at length some time ago - also supplied a variety of automation moves for de-popping.

The real cool solution would be if Bob could impliment a "Slope" for Filter on/off (Just like Mute).

I tried just using the filter the other week... and it worked fine with no pops.
It may have just been the audio - or did Bob fix that *already*?!

[AcousticGlue]

Dan... I addressed this at length some time ago - also supplied a variety of automation moves for de-popping.

The real cool solution would be if Bob could impliment a "Slope" for Filter on/off (Just like Mute).

I tried just using the filter the other week... and it worked fine with no pops.
It may have just been the audio - or did Bob fix that *already*?!

Ask Dave, his wall is covered with Bob-Do-s:D

[Ian Alexander]

What works better in your experience?... automating the EQ strip On/Off, or the filter On/Off?

I often have difficulty with slight pops after this kind of automation and, as you pointed out, it requires further tweaking to get rid of the pops.... sometimes I have to give it up and go with another method altogether, depending on the material being edited.

DanT
Doesn't seem to matter here with VO p-pops. If you have any other EQ'd sections on the track, or might have, it's probably a better practice to automate the filter.

I usually turn on the filter before the P starts and shut it off in the white-ish noise before the vowel starts. Of course, the shutoff has to be after the low frequency bump you can usually see. I rarely hear a click or pop from the on/off.

Another method that works, but takes a bit longer, is to simply replace the popped P with another P from elsewhere in the track. It usually sounds better if it's before the same vowel sound.

[DavidandMary]

The easiest method for pops and sssssssss's is to simply mark the area of the pop or ssss and just pull the fader down xdb.... simple and clean and it keeps the original punctuation sound... just keeps it from expoding out in your face.

Bob L

Thanks Bob! I like simple and clean!:D

[bcorkery]

I'm with Ian, if there's a really troublesome one, it might be easier to have a bank of 'em set aside. But prevention is the best.

Before I do ISDN sessions with Chicago Public Radio, they always have the guest say "Power Point" 3 times to check for plosives. Good test of mic placement.

[DavidandMary]

Thanks Bill... Good to know!:)

[UpTilDawn]

Thanks everybody for the replies and suggestions to my question.
I do already put many of them to work (such as the set of automation gallery presets that Carl so graciously supplied) and find that sometimes one method works better than another.

I've also come up with a number of variations to this particular problem that I will try from time to time depending on how well, or not, other methods work. Sometimes automating volume changes works wonders, sometimes it doesn't seem natural... sometimes EQ changes work better than slicing and softedging... I can never have enough options when it comes to cleaning up trouble areas.

I was kinda hoping that Dave would let me know which of those two automations he mentioned worked most often for him too.
I also tend to think that being able to apply varying slopes to at least the EQ switches (on/off/hi/lo) like we can with mutes would be really helpful in those instances where pops from the automation negate the attempt to fix the plosives or sibilance.

Of course, this only happens when I get tracks from other people...... never anything I record myself.:rolleyes:

DanT

[Bob L]

When automating EQ switches, you might try moving the automation entries to zero cross points.

Also... sometimes you are better off not automating the eq on and off... but slicing the small region of audio out and dropping it to a new track with the eq already on and preset.

Bob L

[Dave Labrecque]

What works better in your experience?... automating the EQ strip On/Off, or the filter On/Off?

I often have difficulty with slight pops after this kind of automation and, as you pointed out, it requires further tweaking to get rid of the pops.... sometimes I have to give it up and go with another method altogether, depending on the material being edited.

DanT

Sorry so long to get back to you, Dan. It's been crazy around here, trying to get out of town on Saturday. But I divest...

I've never thought it matters which switch you throw, since they're all doing the same thing: applying the EQ. To be honest, I automate both the EQ on/off and the filter in/out together. So, what do I know?

The ramp idea is cool, but in the past Bob hasn't thought it made sense, since it's a switch and not a volume-type control. He even seemed to bristle a little that he'd allowed it for mutes, IIRC. :p

I think when you're switching in or out a process like this, the energy of the signal is important: the lower the energy, the less likely to hear a click, the higher, the higher. Like editing on zero-cross points. So, I try to keep these automation entries 1) on zero-cross points, and 2) in low-average-energy areas, where possible. I still get the occasional click, though, so I just move the offending entry to the left or right until it doesn't click. Or, use Bob's idea of a second track for the offensive chunk of audio with a full-time EQ setting. Or try another method. The dropping the volume for the offensive area is a good one, too. I should simplify and use that more often.

But, as you say... no one way works for everything. It's good to have lots of ways to skin that cat! :cool:

[UpTilDawn]

...I've never thought it matters which switch you throw, since they're all doing the same thing: applying the EQ. To be honest, I automate both the EQ on/off and the filter in/out together. So, what do I know?

The ramp idea is cool, but in the past Bob hasn't thought it made sense, since it's a switch and not a volume-type control. He even seemed to bristle a little that he'd allowed it for mutes, IIRC. :p

I think when you're switching in or out a process like this, the energy of the signal is important: the lower the energy, the less likely to hear a click, the higher, the higher. Like editing on zero-cross points. So, I try to keep these automation entries 1) on zero-cross points, and 2) in low-average-energy areas, where possible. I still get the occasional click, though, so I just move the offending entry to the left or right until it doesn't click. Or, use Bob's idea of a second track for the offensive chunk of audio with a full-time EQ setting. Or try another method. The dropping the volume for the offensive area is a good one, too. I should simplify and use that more often...


Thanks for the thoughts Dave..... I was hoping you'd discovered a way to avoid the pops..... even zero-cross points don't always do the trick.

There's been times when I feel like I'm wasting a lot of energy on avoiding pops, bumps, what have you... There's been a few concert-length recordings I've edited lately where the MC makes the mic sound like it's broken... that gets real tedious.

I like Bob's idea of moving the stuff to a new track with the EQ already on! Don't know why that one hadn't occurred to me before. That could go a long way to solving that particular situation where nearly every sentence has one or more offenses.

Anything to improve the bag of tricks works for me.:)

DanT

[Dave Labrecque]

I like Bob's idea of moving the stuff to a new track with the EQ already on! Don't know why that one hadn't occurred to me before. That could go a long way to solving that particular situation where nearly every sentence has one or more offenses.

Yeah, in a situation like that, this could be a real Bob-send. :rolleyes: Just cut the problem area out and drag it down to the anti-plosive-EQ track you've created alongside it. What's not to like? I actually should've done this for all the audiobook editing I did in the last week. Would've saved some time fer sher.

[bcorkery]

Dave,

How did the spread-iting work out? I hope you enjoyed getting out of town.

[Dave Labrecque]

Dave,

How did the spread-iting work out? I hope you enjoyed getting out of town.

Thanks, Bill. I'm posting an answer to you over on the original thread...