Cutting Vocals With Compression

[bdam123]

Are there ideal attack and release settings when cutting with compression. Since I'm using it to jus tame some peaks do I want the attack as quick as possible? Since I'm going to be adding more compression during the mixing stages do I want to avoid making he vocal punchy coming in? Am I shooting for a smoother setting coming in? Thanks.

 

[bouldersoundguy]

Why do you need to "tame some peaks" in tracking? If it's to avoid overs then skip the compressor and just lower your record level. If it's to make the track sound good then you need a good reason to compress in tracking rather than in mixing. A possible good reason is having a particularly good hardware compressor with a sound you like better than any plugin.

 

[bdam123]

i've realized that a compressor during tracking adds a bit of spark. i have an API 527. Are my ears playing tricks on me? For the ones that do compress on the way in, are their ideal attack and release settings for you?

 

[bouldersoundguy]

i've realized that a compressor during tracking adds a bit of spark. i have an API 527. Are my ears playing tricks on me? For the ones that do compress on the way in, are their ideal attack and release settings for you?

Well, that seems to be a piece of hardware that might justify getting used in tracking. As for settings, whatever makes it sound good. Be aware that you might not have the context to know what will sound good down the road at the mix stage, so conservative settings (lower ratios) would be safer.

 

[bdam123]

Well, that seems to be a piece of hardware that might justify getting used in tracking. As for settings, whatever makes it sound good. Be aware that you might not have the context to know what will sound good down the road at the mix stage, so conservative settings (lower ratios) would be safer.

Yea I got the ratio set at about 2.5. It sounds pretty good to me but I'm no veteran so I don't have a great point of reference. My main concerns are the attack and release settings. I know my ears aren't really trained so the differences are super subtle. Generally what should the settings be when using compression on the way in?

 

[bouldersoundguy]

Yea I got the ratio set at about 2.5. It sounds pretty good to me but I'm no veteran so I don't have a great point of reference. My main concerns are the attack and release settings. I know my ears aren't really trained so the differences are super subtle. Generally what should the settings be when using compression on the way in?

There aren't any settings you "should" use except what sounds good. If you haven't trained your ears for compression you might want to skip it in tracking where it can't be undone.

My default start condition is slow attack and fast release. Then I speed up the attack until it starts to catch as much of the transients as I want, then slow down the release until it sounds most natural in the context of the music. I often end up with moderately fast attack and release fast enough to track the decay, for a sound that's a bit aggressive and "in your face" but doesn't sound too obviously compressed. I try to find a threshold point just above the quiet parts.

 

[mixsit]

Yea I got the ratio set at about 2.5. It sounds pretty good to me but I'm no veteran so I don't have a great point of reference. My main concerns are the attack and release settings. I know my ears aren't really trained so the differences are super subtle. Generally what should the settings be when using compression on the way in?

Is the situation one where it's going to cause a problem if you get it a little wrong (or it doesn't turn out as well as it could)? Then go real light, or skip it.
If not, this is the time to dial in something you can hear or want to try.
My attitude (mostly in case one' mode where I don't want to be the one that screws it) is to try to err on the conservative side where ever my ear and best guess takes me- particularly in regard to stepping on transient and life you can't get back.
I've been using a P38 lately (vocals, bass, guitar sometimes ) which just has numbers at each end of the dial. But I would go along with Boulder' and add you're looking in the 70-20ms attack range, 200-50 release, 2:1 if it's 'just a little easy off the top.
Then watch the threshold, and attack, and adjust (and learn) ..in real time if you feel it.
Consider the dbx160xt for example; much faster than above (faster than I like for this kind of vocal tracking), but an auto release -that is also fast, but also tough to beat.
But really, you need to get a handle on what happens (even at low ratio) when attack gets down into where it's seeing the range of the rise times of the track, (i.e., large differences, deeper grab, and consequently farther to release for a given threshold), and the pulled back ('smoother if you like) sound of slower release vs. the more forward presence' sound of fast release.
At this point you have to be extremely general 'cause there's no plan.. neither of us know what the gig is.

 

[bdam123]

Yea, I'm gonna have to play with it a lot more. I hate spending so much $$ on it and not putting it to good use. It does give a nice sparkle to the vocal and I like that a lot. Thanks for all the info guys. I'll def be back on here in the following weeks asking more questions.

 

[dco]

compressor settings should absolutely vary depending on the input source.

If what you're going for is transparent sounding compression, you want to match the envelope of the audio with your compressors attack and release times to avoid distortion.

but honestly, if the harmonic/envelope distortion/specific compressor flavor is what you're looking for, then you really need to find the best sound by trying different settings/combinations.

too fast of an attack will cut off much of the transient/attack/punch, but maybe also try aggressive compression, but using parallel compression too (youtube it)

 

[Todzilla]

What "sparkle" gets imparted by a compressor on tracking that can't be imparted using that same compressor on the already recorded dry track?

I always thought the major cats compressed on tracking just to save time and they totally know what they're doing.

 

[bouldersoundguy]

What "sparkle" gets imparted by a compressor on tracking that can't be imparted using that same compressor on the already recorded dry track?

I always thought the major cats compressed on tracking just to save time and they totally know what they're doing.

The performer may react in real time to hearing themselves compressed which will make for a different performance than recording uncompressed. That may or may not be a good thing, but it will be different. It also avoids a round of D/A and A/D conversions and any additional record offset issues.

 

[dintymoore]

The performer may react in real time to hearing themselves compressed which will make for a different performance than recording uncompressed...

Exactly. I always record with a compressor. I want to hear myself singing through it, i don't want to record with one sound and then "fix it later", I want to hear the final sound as I'm singing it - it makes me sing different.

My Summit compressors do not have numbered dials for attack and release settings, they have 3 position toggles, and I keep both of them at the fastest settings almost all of the time because I want to compress the peak, not after the peak.

 

[Todzilla]

Interesting.

I find it's harder to convince singers to work the mic with compression in the signal chain.

 

[miroslav]

I don't use a compressor...but I do occasionally like to use my analog AD&R Superdynamic F601-RS limiter to control just the most offending peaks on vocals that have lots of dynamics.
This is a broadcast application limiter that has only higher +dB Treshold settings (+4 going out to +14 dB), and is very natural sounding. If you really push it hard with one of the lower dB settings, you can hear it working as the peaks are taken out...but it's still not a rude cut-off.
At the higher threshold settings, when your vocal jumps out real loud, the limiter just gently takes out the peak without any of the "sucking" effect that you can get with a typical compressor set to infinity.

 

[bouldersoundguy]

Interesting.

I find it's harder to convince singers to work the mic with compression in the signal chain.

It's not something I'd automatically do or not do. Some of the people I work with are used to it. If I don't know that a singer is okay with it I'll keep the threshold high, lower the ratio or bypass it.

 

[LeeRosario]

Are there ideal attack and release settings when cutting with compression. Since I'm using it to jus tame some peaks do I want the attack as quick as possible? Since I'm going to be adding more compression during the mixing stages do I want to avoid making he vocal punchy coming in? Am I shooting for a smoother setting coming in? Thanks.

This may or may not have been answered....


A little compression on the way in is very common for a number of reasons. There are 3 common reasons you'd want to do this:

1) You have a limited time with a very special compressor/limiter and you want to get "that sound" for anticipation you'll never get a chance to use it again. All compressors sound different and some are valued for that purpose.

2) To tame peaks, and this is especially true for wild singers. Unfortunately, not all singers know how to back away from the mic when they sing (or even have proper technique for that matter), so it's pretty necessary to control that early on.

and then the last (which I feel is the most important): to get a good average signal to noise ratio out of your track. When you receive "professionally recorded tracks", you're looking at tamed dynamics and a very tamed looking waveform. All with ample headroom.

The idea with tracking is to either be as transparent as possible or if you're experienced enough, get the "right sound" right away. There's no rule to not compressing things on the way in. Actually, I think things tend to sound better during mixing that way.

That being said, on a vocal, if you're unsure, stick with a 2:1 ratio and a moderate attack with a quick release on your tracking. Depending on the compressor, you shouldn't be getting more than say -4 to -6 db in gain reduction. It's very important to gain stage things properly at that point.

 

[Ephi82]

good advice in previous post.

I also consider the goal of the recording and the type of song when deciding how to use compression on a vocal track.

If you have a song with sparse backing (lets say acoustic guitar, slide, bass, and drums) and you want to impart that "you are in the room" realism to the recording, I only compress/limit enough to keep the vocal out of distortion. If you have a singer with good technique, you may not need the compressor at all. The sound of a great singer and a great mic, and a vocal recording that goes from whisper quiet to a more aggressive sound can be heard clearly. (2:1, fast attack, long enough release to sound natural, threshold only enough to tame a transient here and there)

In contrast, lets say the second song has a very busy backing, lets say a rockin tune with layers of distorted guitar tracks, bass, drums, keys and 3 BU vocal. The lead vocal tracked in the manner of the first example wont allow the singer to go from whisper to a SCREAM without distortion (even a singer with great technique) and the resulting track will be very challenging to sit in the mix with enough command and presence. You have to tame the loud transients while at the same time make it possible to hear the subtleties in the voice, and the quieter elements in the vocal delivery. (4:1, fast attack, long enough release to sound natural, threshold depends on how much gain I am driving from the mic, but anything more that 6db reduction starts to sound more like the "compressor" than the vocal)(sometimes this is a sound that is desired, again in the context of the song)

As to why shouldn't you wait until the mix to put in compression?

1) I want the singer to hear a GREAT vocal sound when tracking
2) I believe in building a song and a recording by getting the tracks to connect and compliment with each other while you record. Inspiration and creativity flow from this.

Does a painter lay down a bunch of objects and colors in a painting, without concern that they compliment each other and then, when the canvas is full, start twiddling around to make the elements work? Hell no, it's too late!

 

[mixsit]

...If you have a song with sparse backing (lets say acoustic guitar, slide, bass, and drums) and you want to impart that "you are in the room" realism to the recording, I only compress/limit enough to keep the vocal out of distortion. ..

Just a small point.. Reasons to; almost always for the sound or the effect of the compressor, almost never to keep out of distortion ..a least in anything near a controlled situation. Namely, if the level at the mic is too hot for the pre, comp's too late, if the level out of the pre is good, level to tape' is also in alignment.
A second reason to consider not doing it for protection; often fast enough for protection runs counter to the speed or setup that gives the desirable sound of compression you might have gotten.

 

[Ephi82]

Just a small point.. Reasons to; almost always for the sound or the effect of the compressor, almost never to keep out of distortion ..a least in anything near a controlled situation. Namely, if the level at the mic is too hot for the pre, comp's too late, if the level out of the pre is good, level to tape' is also in alignment.
A second reason to consider not doing it for protection; often fast enough for protection runs counter to the speed or setup that gives the desirable sound of compression you might have gotten.

Good point, but i used the word "compress/limit" , in the sense that I want a hot vocal track, but never driving the input to distortion in the
DAW.

In the case of using tube mics, it is often advantageous (IMO) to drive some gain to get more of the tube "sound" to print. This may result in some peaks that have to be limited. Make sense?

 

[mixsit]

Good point, but i used the word "compress/limit" , in the sense that I want a hot vocal track, but never driving the input to distortion in the
DAW.

In the case of using tube mics, it is often advantageous (IMO) to drive some gain to get more of the tube "sound" to print. This may result in some peaks that have to be limited. Make sense?

Absolutely. I think really the only distinction here is 'volume controls for levels, compressors (limiters) for compressin'ing
Lots of comps limiters have output knobs too.