Jump to content
RICHARD HUTCHINS

Is there a simple way to even out vocal signal

Recommended Posts

Hi

I have a song with lots of vocal bits, each done in small chunks as my voice is pretty pathetic, and I need lots of takes etc for every phrase. I end up with fairly okay vocal takes, but my problem is the relative volume of each take. Some loud, some clipping in the red, some too quiet. I now know how to use automation clip gain, but I still cant quite get this to be consistent through the track. And it takes a long time to draw the nodes and drag them up or down. Is there a quicker way to choose a median signal, say -12db and hook everything onto that? I know those who help me have a far more subtle approach to this, but for me I just need it simple! This might be a dumb question, so apologies in advance if it is. 

Share this post


Link to post
Share on other sites

I have melodyne studio 5 and it has a  "Leveling Macro" feature that can make loud "notes" quieter and another that can make soft "notes" louder. Its pretty simple. I have used it on vocal parts and guitar parts and I like it better than compression for some things. It also has a fade feature that can be used to drop a peak or do a fade on a "note".  

Share this post


Link to post
Share on other sites

Clipping is bad. I would redo those takes, but with the rest of it, you've given the perfect example of what a compressor can be used for.

Share this post


Link to post
Share on other sites

There’s a few ways to do this. One is to go clip by clip and normalize to about -1.0 db. This will take care of takes that we’re recorded at lower levels. 
I prefer to do this in Wave Lab but in a pinch Cakewalk can do this. 
Then for sure you’ll need to hit a compressor to level it out. I use the pro channel PC 76 u   And then push the input until it just starts to move. I will use the BT brick wall set at -1.0. To catch overs 

  • Like 1

Share this post


Link to post
Share on other sites

There are a few different choices.

I always cut vocal phrases into clips, so if necessary, I do what you do, which is clip gain automation. Another technique, which John mentions, and is used by Craig Anderton and Mark McLeod, is to just have Cakewalk normalize them. I believe that this is peak normalization, though, which may not address your issue.

Since you're experiencing tedium editing every clip, a plug-in might be better. There are ones that analyze the signal and "ride" it so that everything comes out near an average level: W.A. Production Outlaw is on sale for $7.90. Waves Vocal Rider and Meldaproduction MAutoVolume are a couple others. These two go for around $50 each (the Melda one can be had for $10 off if you sign up for their newsletter, plus another 20% off if you use a referral code-ask and I'll post mine). Outlaw and Vocal Rider are both 4-knob affairs, MAutoVolume (predictably) has more controls and includes a sidechain that can be used to make sure that the vocal (or whatever) always stays on top of whatever comes in on the sidechain.

 I will say that once I cracked the Cakewalk Reference Guide and learned some of the shortcuts and modifiers for editing automation, my automation work is going much faster and is less tedious.

Share this post


Link to post
Share on other sites
10 minutes ago, Starship Krupa said:

Since you're experiencing tedium editing every clip, a plug-in might be better. There are ones that analyze the signal and "ride" it so that everything comes out near an average level: W.A. Production Outlaw is on sale for $7.90. Waves Vocal Rider and Meldaproduction MAutoVolume are a couple others. These two go for around $50 each (the Melda one can be had for $10 off if you sign up for their newsletter, plus another 20% off if you use a referral code-ask and I'll post mine). Outlaw and Vocal Rider are both 4-knob affairs, MAutoVolume (predictably) has more controls and includes a sidechain that can be used to make sure that the vocal (or whatever) always stays on top of whatever comes in on the sidechain.

There are 2 good plugins from TBProAudio that do this, too:

  • DynaRide (39 EUR) is similar to Waves Vocal Rider, but it has more possibilities and it can also be used for Bass and other instruments!
  • GainRider (59 EUR) is more visual and more academic with a lot of controls.
17 minutes ago, Starship Krupa said:

 I will say that once I cracked the Cakewalk Reference Guide and learned some of the shortcuts and modifiers for editing automation, my automation work is going much faster and is less tedious.

Tell me more!

  • Like 1

Share this post


Link to post
Share on other sites
3 minutes ago, marled said:

Tell me more!

p. 484, "Using the Smart tool on automation" has a diagram that shows the different hotspot areas of a clip and a list of the modifier keys. Since the Smart Tool is what we mostly stick to, this page is the one to tack up over your desk if you, like me, find yourself mystified by what causes the Smart Tool to switch modes.

Then there's pp. 1110-1118, "Creating and editing automation envelopes." That's the meat 'n' p'taters guide.

For the individual tools, the applicable ones have their own pages. p. 488, "Using the Select tool on automation," p. 491, "Using the Move tool on automation," p. 495 "Using the Edit tool on automation," p. 505, "Using the Freehand tool on automation," p. 512, "Using the Line tool on automation,"  p. 513, "Pattern tools," p. 517, "Using the Erase tool on automation."

Memorizing which tools are attached to F5-F9 has helped my Cakewalkin' immensely, especially when I can't get the freakin' Smart Tool to behave. Those keys would be some of the first Cakewalk-specific assignments I would have someone new to Cakewalk learn. I wish I had done so for the years I spent flying up to the Tools module in the Control Bar or calling up the HUD. The Smart Tool does so much different stuff that I can forget that there are dedicated tools that don't switch modes on me depending on where I click.

Then the section on automation, p. 536 "To raise/lower an audio Clip Gain envelope," and another biggie, pp 544-546.

I appreciate your asking, because it made me go through and find all these sections that I had kind of stumbled upon earlier. So now I have a list to refer to. As can be the case with that 1080-page monstrosity that is the Ref. Guide, the information is spread out and duplicated. Guaranteed, if you read those 20 various pages or even just skim them, you will pick up at least one new "hey!" automation editing move.

It's another one of those areas of Cakewalk that you can get really fast and precise and creative with if you put in the time. A problem I have is that it's hard to retain stuff like this if I don't use it very often, and my frequency of doing automation makes the learning process go slowly.

  • Like 1
  • Thanks 1

Share this post


Link to post
Share on other sites

Hi Richard

This really is a performance issue.  My best advice is to work on your vocal performance & mic technique.  'Fix in the mix' is such a tedious process! A great vocal take should only normally need a little levelling and eq plus any effects so try to work on your performance - it'll give much more satisfactory results.    

Yes, automation and normalisation are an option.  Avoid compression at this stage!

To avoid clipping: Try to stay about 8-12" back from the mic to prevent proximity effect, prevent plosives and improve tonal consistency. Also, to stop plosives, use a pop filter. If you don't have one, place your mic just higher or lower than your mouth to let your breath 'overshoot' the diaphragm. Move a little further away in higher energy passages and a little closer if lower energy to even out the gain.

Turn the preamp gain down a few dB's. Digital recording doesn't need the same 'hot' signal as analogue. 

To improve your performance technique, practice before hand and try to gain consistency over levels, mic placement, distance from the mic etc. You should be able to see where you perform better and where consistent adjustments need to be made.

The artist has the right to destroy their work and try again. I would delete the clipped / poor takes: it's quicker to re-record!

Here's a summary:

  • Try to get mic-distance down to 3 options e.g. 
    • Quiet passages: 6-8"
    • Normal: 8-10"
    • Loud: 12-18"
  • Set preamp gain to only peak into the yellow, never red.
  • Control breath and pops with technique, mic placement & popfilter.  
  • Delete what doesn't work and redo it. Be ruthless!
  • Do a few trials and nail it down!

Hope this helps!

G

 

  • Like 2

Share this post


Link to post
Share on other sites
4 hours ago, Starship Krupa said:

There are a few different choices.

I always cut vocal phrases into clips, so if necessary, I do what you do, which is clip gain automation. Another technique, which John mentions, and is used by Craig Anderton and Mark McLeod, is to just have Cakewalk normalize them. I believe that this is peak normalization, though, which may not address your issue.

Since you're experiencing tedium editing every clip, a plug-in might be better. There are ones that analyze the signal and "ride" it so that everything comes out near an average level: W.A. Production Outlaw is on sale for $7.90. Waves Vocal Rider and Meldaproduction MAutoVolume are a couple others. These two go for around $50 each (the Melda one can be had for $10 off if you sign up for their newsletter, plus another 20% off if you use a referral code-ask and I'll post mine). Outlaw and Vocal Rider are both 4-knob affairs, MAutoVolume (predictably) has more controls and includes a sidechain that can be used to make sure that the vocal (or whatever) always stays on top of whatever comes in on the sidechain.

 I will say that once I cracked the Cakewalk Reference Guide and learned some of the shortcuts and modifiers for editing automation, my automation work is going much faster and is less tedious.

To be clear, what Craig and I do is slightly more than normalize them...  what we do is split the clips at each phrase, then normalize or clip-gain adjust them individually.  So each individual phrase is normalized against itself, not the whole vocal track.

Although in saying that, I tend to reach for Vocal Rider or MAutoVolume far more often now, even if it means applying that effect, then doing the tweaking afterwards using clip gain automation, as it saves a lot more time.

  • Like 1

Share this post


Link to post
Share on other sites

There you go. Answers to help you weigh the matter of "do I address this by building my mixing skills or my singing skills?"

I feel kinda sheepish saying this now, but a few years ago, when I decided to really start singing, at first it was an exercise in learning how much I didn't know after winging it for decades. Somehow I was of the belief that I either had singing "talent" or not, so I should just go with what I had and see what I could make of it.

Fortunately, my ability to focus on imperfections got the better of me and I just kept recording the same song over and over and over, like for a period of a couple of weeks every day or so I'd try again to get a usable take. I noticed a bunch of stuff, like on listening to my playbacks, I would do some in one register, then some in another, like I didn't even know what the singing melody was supposed to be even though I had written the song. It turns out that my sense of being in tune was pretty good, I wasn't singing sour notes, they were on key, just kind of whatever my voice wanted to do with the note.

One of the things I find fun to do after getting a lot of takes of the same instrument is to unmute ALL of them just to get a laugh. So I started to do this with my "discarded" takes and I started to turn into this huge men's choir. Kinda sounded like the Moody Blues on a camping trip. I could hear my technique getting better over the weeks, the later takes had things like vibrato and held legato notes.

Well DUH, what I hadn't realized is that the voice, just like piano or guitar or bass or whatever, is an instrument that you make better by practicing. Singing and recording the same song over and over again for a couple of weeks got me able to do some cool stuff like extending my range, singing in (intentional) harmony, pay attention to phrasing, breathing, all this stuff I took for granted having previously only "sung" in the very forgiving environments of church congregations and inadequately mic'd underground rock shows.

TL/DR:

Moral of the story: singing improves with practice just like any other instrument. Mixing skills are necessary to make singers sound their best.

You have takes with irreparable flaws: the clipping. Anything else can be patched up pretty well, levels, plosives, whatever. But clipping is so baked in, even with RX it's hard to get rid of it without losing transients in the highs. You have total access to the singer, you. Studio time is presumably free-to-no-cost.

If it were me, I might do both (and sometimes I do exactly this): Comp together the best vocal track you can from your pieces, then use that as a guide vocal (and/or insurance), and try some new "pressure's off" takes.  Set up the popper stopper like Gary says, get a glass of water, set up a music stand with your lyrics, lower the lights, turn the input knob on the mic pre 1/8 counterclockwise from whatever you think is "safe," put it in comp/loop mode and let it roll. Don't have to even do whole takes, just step back and let it come around again if you blow it. It will be easier with the guide vocal you comped together. Give yourself permission to improve as a singer rather than beating up on your voice for being "pathetic."

And breeeeeeeeathe.

  • Like 1

Share this post


Link to post
Share on other sites


A lot of us old geezers learned how to sing live on stage in front of a dance floor. There were no in ear monitors or even floor wedges most times. Board mixes of those nights were a huge learning experience of how bad some of us sang. 
But you did learn mike techniques and how to breathe. We got the mileage in on our vocal cords. 
My voice definitely improved when I went solo. I could hear myself and every mistake at hi volume. 
I learned to just avoid songs that didn’t get better. There’s lots of songs I can sing, so why fight it. 
I think this is also a mistake with song writing. You get an idea and only to find out later that you can’t even sing your own song. 
I spent my last 3 years working in a Care Facility and got to sing and play guitar for over 12 hours a week.  I learned how to project. 

vocal technique: 

Stand up straight never sit 

Practice yoga breathing 

Learn to find the spot in your body that creates good tone and projection 

Find the right mike  I cannot use a Large Condenser mike  I’d blow it up  

Do not blast your headphones  this will put you out of pitch  Doppelgänger effect . try using one ear off so you hear your self naturally 

Rig up a wireless keyboard so you can get away from your computer and focus. Learn to toggle the now time between stop in place and rewind to last start using ctrl W 
Set up your signal path so there’s absolutely no danger of clipping absolutely!  It’s easy to turn up a track. It’s impossible to fix an over. 
Take note of everything and settings so you can duplicate it later. 
Water makes me burp. Scotch or JD is better. 

 

  • Like 2

Share this post


Link to post
Share on other sites
Posted (edited)

Another thing I realized about singing that seemed obvious after it came to me is that it's not necessary for me to "belt." Some of my favorite singers like Elliott Smith, Tom Petty, Nick Drake, and (Reprise era) Frank Sinatra seldom got their volume very much above normal speaking level (with exceptions). They let the mic (and the LA/2A) do the work, and the result is that it sounds more intimate and less strained.

So I don't have to be able to do Roger Daltrey, Freddie Mercury, or Robert Plant in order to sing rock 'n' roll effectively. Nick Drake and Elliott Smith often got down almost to a whisper, to amazing emotional effect. It can even compel the listener to work harder to hear the vocalist, and therefore pay more attention to the lyrics.

Next time I get on the mic, I'm going to take John's suggestion to pay attention to where in my body my singing voice is coming from.

And yes, using the mic in your collection that most flatters your voice (and try all of them, you might be surprised and find that your SM57 is The Answer).

Edited by Starship Krupa
  • Like 2

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now

×
×
  • Create New...