• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

Recording loud and quiet vocal performance

Mytt88

Mytt88

Jan 18, 2011
217
76
#1
Hi all,

Last night I tried to record a vocalist for my band with a track that has quiet and loud parts. It goes from talking volume to full on screaming.

If I brought the levels low to avoid clipping the loud parts the quiet parts are way too low in the mix and when I raise the volume I get the noise levels creeping in. If I set the volume/input correct for the quiet parts it overloads and clips on the loud parts. We tried moving further away from the mic for the loud parts but I can notice the tonality change a bit too much for my tastes when doing that. I had used the minus 20db pad on both mic and audio interface to stop the clipping.

I was using a condenser mic so I'm thinking that was the problem. Would a dynamic mic be the way to go? Anything similar to an sm7b in the cheaper price bracket?
 
#4
Set input gain to -12 to -6 region to cover loudest input. "Normalize" track once recorded. Then use one or more compressors (LA2A is a great on for this) and gently compress the hell out of the track! You'll know by the sound if you've gone too far with it. Joe Meek used to say "if is sounds good, it is good."

If you have the outboard gear, use a compressor going in, and only knock off about 2 db. If you have a 2nd comp/limiter, use both the same way - reducing signal by only 2 db each compressor.

I don't know what you know about compressors, so please don't think I'm speaking down to you. But the idea is to reduce the dynamic range you spoke of. Once recorded, you'll be able to reduce that range further.

The most important thing is to record the vocal without clipping, since that is not a pleasant sound. Then work with it afterwards "in the box," as they say... There are many free or nearly free comp/limiters available to help you, if you don't already have one (or a few).
 
#5
Oh, and yes, a dynamic mic might be better in this situation. A Shure SM57 works well. But my fave dynamic is an EV RE15 (or RE10, or RE20). The 57 would be the least expensive of these mics... But any dynamic would do the job.
 
#6
Thanks for the response. I've used waves bass rider (I'm a bassist and dont have vocal rider) along with cla la3a plugin. I don't have outboard gear at the moment but will probably get a comp.

I'm quite new to this so I'm just experimenting so all advice is welcome.

Is normalising evening out volume? And any plugins good for this? I have heard of audacity having it but would prefer staying in one place rather than exporting and importing back in.

I've got it to a decent take level wise and avoided clipping the only problem now is the noise floor. Not sure how to get rid of it will probably have to re track with a dynamic mic

Thanks again
 
#7
Stumbo said:
How about if you record the soft and loud parts on different tracks?
Click to expand...

Mytt88 said:
I thought about that but it would just mean pulling the volume fader down on one track on the daw and not actually stop clipping the interface on the way in
Click to expand...

Stumbo's implying that you'd record the loud vocal sections to one track, w/ the preamp's gain set to avoid clipping. Then you'd adjust the preamp gain (and move the singer closer to the mic, and maybe use a different mic for different character) before tracking the quieter vocal parts to a separate track.

That gets you around clipping on the loud track and—assuming your vocal preamp is decent—gives you better signal-to-noise on the quiet track. It also makes it easier to use different compressor settings on the quiet and loud sections.

That's good practice when the situation allows. After tracking, you can also befriend your DAW's automation:
  • Automate the fader or automate a trim insert to get the different sections of the vocal track to sit in the mix. A "section" here could be as small as a single syllable.
  • Automate the faders between lines and phrases of the vocal tracks to reduce hiss. Alternatively, you could automate mutes, although that often sounds too obvious if the track is exposed.
  • Automate the compressor settings—that is threshold, ratio, attack/release, and make-up gain—to get different sections of that vocal track to behave. If you're automating a ton of settings, then in terms of workflow, it's usually easier to drag a sections of a vocal performance to separate tracks, grouping sections onto the same track when they need similar processing or FX.
  • Alternatively, put more than one comp (w/ different settings or different characteristics) on a track, and automate their bypasses to determine which comp is acting on different passages or phrases.
Etc.
 
Last edited:
  • Like
Reactions: Stumbo
#8
EastCoastThump said:
Stumbo's implying that you'd record the loud vocal sections to one track, w/ the preamp's gain set to avoid clipping. Then you'd adjust the preamp gain (and move the singer closer to the mic, and maybe use a different mic for different character) before tracking the quiter vocal parts to a separate track.

That gets you around clipping on the loud track and—assuming your vocal preamp is decent—gives you better signal-to-noise on the quiet track. It also makes it easier to use different compressor settings on the quiet and loud sections.

That's good practice when the situation allows. After tracking, you can also befriend your DAW's automation:
  • Automate the fader or automate a trim insert to get the different sections of the vocal track to sit in the mix. A "section" here could be as small as a single syllable.
  • Automate the faders between lines and phrases of the vocal tracks to reduce hiss. Alternatively, you could automate mutes, although that often sounds too obvious if the track is exposed.
  • Automate the compressor settings—that is threshold, ratio, attack/release, and make-up gain—to get different sections of that vocal track to behave. If you're automating a ton of settings, then in terms of workflow, it's usually easier to drag a sections of a vocal performance to separate tracks, grouping sections onto the same track when they need similar processing or FX.
  • Alternatively, put more than one comp (w/ different settings or different characteristics) on a track, and automate their bypasses to determine which comp is acting on different passages or phrases.
Etc.
Click to expand...


Yes I see I tried that and it gave a different tonality to the vocals which I didn't like. It's what I ended up doing but I know for next time it could be better with the correct mic.

The automation sounds a bit too advanced for me at the moment but I will definitely look into it solid advice thank you
 
  • Like
Reactions: Stumbo
#12
Use a compressor and even more important, as JRA wrote: mic techique by the vocalist!!!
Mics have a "Proximity Effect". If you are very close to a mic, the low frequencies are overly enhanced. It feels like someone speaks directly into your ear. It gives a feeling of intimacy.
And the other way round: during loud passages go further away from the mic!
 
#13
Set the recording levels at the loudest part the singer sings, just touch the red a little.
There will be plenty of information at the lower levels that will get recorded where you could boost them later if they're too quiet.

Cardioid mics have that proximity effect, an omni mic doesn't. If the vocal is being done by itself, try an omni mic
 

Latest posts

Top Bottom
Back