Vocal comping

maartenl945

maartenl945

Member
Hi,

I just released a video today on Vocal Comping on my YouTube channel, and was wondering if you guys (and girls) over here perhaps have some additional tips, tricks or techniques that you use when comping a vocal. I know the exact tools are dependant on which DAW you are using but still, there might be things to learn even from other DAWs.

Regards,
Maarten
 
Last edited by a moderator:
I've never really done the "do 3 takes and then comp" way of vocals. I've always been a "get it done in one go" kind of person and if a line or word goes awry, do it again. So I won't move to a bridge, chorus refrain or whatever, until the first verse is nailed.
It's probably just me, but I have felt for a very long time that too much "magic" is ascribed to recording vocals. Just go in there, know the song or the lines you're doing, bang them out and be done with it ! It shouldn't take all night. I regard it the same way I regard washing the plates. Be sure the plates are well washed with no bits left on them. With vocals, be in tune. Each singer should be fairly unique but each word sung doesn't have to be.
 
I wish. Even my opera singers embrace comping when they discover it. They get tired, start to make mistakes, so often go with less than perfect entire takes. Give one half a dozen amps tracks and they will go home, and produce a list - perhaps the vocal score, with the best takes - or one who colour coded every one from red - dreadful to green good. being able to replace words and phrases they really like. The wrong slight emphasis on a syllable, or a tiny pitch issue here or timing there. The end product is absolutely better. Same with MIDI piano editing - a great piano VSTi and you can work magic. They love it.
 
I've done it a couple of times on guitar solos, where I just came up with a better riff or something, and on vocals where I screw up words, or just hit bad notes in the middle. I didn't use the "comping technique" per se, which is also available in Reaper. I just did a simple cut/paste to another track. In the end, it gets to the same point. I prefer to do that rather than do a punch in/punch out. I never seem to get a seamless edit when I try to do punches.

One thing that I learned is that you need to lock the tracks so they won't move forward or backward on you. It saves you a bunch of headaches!

You can find the info in Kenny's Reaper Mania videos.
 
The singer I work with a lot prefers singing through the whole song several times and not work on sections individually and do punch ins. So what we usually do is that he records around 3 to 4 tracks from start to finish. Comp from that during the recording session already and then we do punch ins where still necessary. Usually that’s still in one or two places then.
And yes I also use the same comping technique for other recorded parts as well.
 
I almost always comp my vocals. Usually my best overall take is #3 on any particular day, but there will be sections where the phrasing was better on a different take (for example), etc. Bridges often get separate takes.
The singer-songwriter (friend) who I am working with right now, prefers getting it all in one take - and thinks I am deleting previous takes! - but I seldom need to comp anything except she discovered on one song recently that going back and doing the bridge section separately was much easier.
I don't 'cut and join' comps, though, I leave them on their separate tracks (I hate manipulating any more than 2 takes in Reaper) and automate them together.
 
Anybody compress only a particular band of their vocal?

One processor I liked was the Dolby 362 with a Vocal Stretch mod done to the CAT22 cart. It was a mild compression 2:1 or so applied to particular bands.

None of my rack stuff does that. Recently, I watched a video on using a crossover box to isolate a band. Then compress only that band with an external compressor. It seems people take it to the next level with higher ratios than the 362's 2:1.

The Dolby vocal stretch or Lennon mod only applied the effect to higher bands.

It has a desirable sound for voice.
 
LazerBeakShiek said:
Anybody compress only a particular band of their vocal?

One processor I liked was the Dolby 362 with a Vocal Stretch mod done to the CAT22 cart. It was a mild compression 2:1 or so applied to particular bands.

None of my rack stuff does that. Recently, I watched a video on using a crossover box to isolate a band. Then compress only that band with an external compressor. It seems people take it to the next level with higher ratios than the 362's 2:1.

The Dolby vocal stretch or Lennon mod only applied the effect to higher bands.

It has a desirable sound for voice.
Click to expand...
Yes that’s more part of vocal mixing to me. I usually have a multi-band on just the high frequencies as an additional de-essing step. It’s in another one of my videos :).
Vocal mixing
 
Farview is correct.

It was designed to be a Dolby A compander noise reduction system. People started turning off the high frequency band in the decode stage, just like people used to encode cassettes with Dolby and then play them back without the decode system. It boosted the high frequencies. It became a "trick".
 
You're missing the bottom part of the system, where the compression is subtracted from the compressed signal, which is supposed to restore the original response, minus 10-15dB of noise.Dolby A.jpg
 
Yes, but calling it a multi-band compressor is kind of like calling a freight train an electric motor... technically correct, but misses the point.
 
I'd always prefer that the singer re-do whole sections, rather than a word or two, but if they want to do it like that, then we do it. If it's me singing, I always re-do the whole section, to keep it sounding natural. That's just me though. As far as de-essing, well, I think everyone is familiar now with the Attack Of The De-essers on a lot of modern music. I use automation along with the Eiosis 2 de-esser, part of the Slate family of plugs. It works, but you need to tweak the De-esser. It goes on the vocal track itself, as opposed to, say, the bus. So, quite a bit of the de-esseing is done with automation. It is tedious, but worth doing for sure. It's a real bummer when the singer rejects the take (or part of the take) that you spent some time on rolling off the esses and plosives. For the record, multi-band compression frightens me. It seems like overkill, or maybe a solution looking for a problem.
 
TimOD said:
I'd always prefer that the singer re-do whole sections, rather than a word or two, but if they want to do it like that, then we do it. If it's me singing, I always re-do the whole section, to keep it sounding natural. That's just me though. As far as de-essing, well, I think everyone is familiar now with the Attack Of The De-essers on a lot of modern music. I use automation along with the Eiosis 2 de-esser, part of the Slate family of plugs. It works, but you need to tweak the De-esser. It goes on the vocal track itself, as opposed to, say, the bus. So, quite a bit of the de-esseing is done with automation. It is tedious, but worth doing for sure. It's a real bummer when the singer rejects the take (or part of the take) that you spent some time on rolling off the esses and plosives. For the record, multi-band compression frightens me. It seems like overkill, or maybe a solution looking for a problem.
Click to expand...
Yes we usually don't record/redo individual words or sentences with this singer either but we do full takes of the entire song or sections of it. But we then combine these full takes by comping and choosing the best parts of all takes. So it happens quite often that we just use a few words from a certain take, and then get back to the main take. If the sound, intent and delivery of the vocal is the same, I feel it works quite well.
 
Know what you're going to do before you do it. Construct your vocal phrasing anticipating that you're going to be recording this one line at a time, if neccesary.
 
You must log in or register to reply here.
Back
Top