How to apply volume automation during softer parts of a song?

T

Tesgin

Member
I'm trying to understand the right way to use volume automation together with compression when there are parts of a song that are deliberately quieter.

I "get" the idea of using volume automation (VA) to keep levels steady, and the importance of doing that prior to compression, which will get a more even effect from the compressor throughout the song, and with less work from the compression.

What is the proper way to do it when, let's say, the first verse is very soft, building to a crescendo on the chorus for example?

If I use volume automation – or let's say, instead, gain automation – up front to make the volume constant on the vocals, for example, I lose the effect of the softer part building into the louder part.

So, I'm getting lost in this: for purpose of applying compression evenly on both loud and soft parts of the song, yes, VA will accomplish that so that I don't squash the louder parts and only mildly compress the softer parts of the song.

I've read that a number of people simply recommend leaving the quieter part alone and only compressing the louder parts. However, that doesn't make sense to me: the softer parts can benefit from compression just as much as the louder.

As I'm thinking it through, I'm thinking this would be a way to do it, but I'd like validation (or correction):

I "think" the way to do it would be to use gain adjustment (either by envelope automation, or just cutting the clip up into sections and using the trim tool; I use REAPER, btw) as the first step to make the volume constant for the entire song, prior to compression. So, consistent gain-staging so the clips are consistent around -18 or so. THEN use compression, which would now be applied evenly across all sections of the song. THEN use volume automation, post-fader, to adjust the volume of the softer parts to taste. That seems to me to be the correct way to apply compression equally across soft and loud parts of the song while at the same time preserving the volume changes.

Am I right here? Am I thinking things through clearly? Am I missing something? Is there an easier way to do this?

Am I making any sense at all?

:)

Thanks, all.

Tesgin
 
If you're using volume automation to get the vocal levels more consistent then what are you using the compression for? Also are you talking about just the vocals or all the instruments in your arrangement? It seems to me that the method you outlined would work but im not sure if its more than you need to do or not. I'd say give it a try though and see how it sounds.
 
I would have to hear the track but I would have to say this is going to come down to the arrangement. Say the dynamics of the vocal performance is flatter and quieter during the verse and expands in the chorus. In such a case the vocal can be out front or pulled back depending on the needs of the arrangement. If it needs to be up front then in order to keep from having the level cranked the instrumentation has to be lowered or moved to the side or eq'd to get out of the way, again, depending on the arrangement. Whichever works best, and it should be automated so that it only is happening when needed. A simple separation of the instrumentation and the lead vocal into two buses before the mix bus can be set up with a ducking compressor on the instrumentation bus triggered by the LV bus. But the simplest way is to have the instrumentation more sparse when the vocal is quieter.
 
Gtoboy said:
... If it needs to be up front then in order to keep from having the level cranked the instrumentation has to be lowered or moved to the side or eq'd to get out of the way, again, depending on the arrangement. Whichever works best, and it should be automated so that it only is happening when needed. A simple separation of the instrumentation and the lead vocal into two buses before the mix bus can be set up with a ducking compressor on the instrumentation bus triggered by the LV bus. But the simplest way is to have the instrumentation more sparse when the vocal is quieter.
Click to expand...

I'm not thinking so much in terms of how to make it sit with the other pieces in the arrangement, but more looking at, say, the vocal by itself: I'm reading so much about how important it is to get an even signal prior to compression, by using clip gain, or volume automation before sending to, say, a buss with compression. So just looking at compression on the vocal track, for example, if there's a lot of dynamic change between the louder choruses and softer verses, only applying compression to the track as a whole won't touch the softer parts (unless I over-compress the louder sections). Should one use just a milder compression to touch the louder peaks and not worry about the softer parts? Or is it more appropriate to compress the louder parts and the softer parts equally by using the clip-gain adjustment I described earlier, and then use volume automation post-compression?

So I'm asking just in terms of the proper use of VA with compression on a track, with the next step then being to blend it in with the arrangement. Not sure if I'm explaining myself clearly.

Thanks,
TB
 
Tesgin said:
I'm trying to understand the right way to use volume automation together with compression when there are parts of a song that are deliberately quieter.

I "get" the idea of using volume automation (VA) to keep levels steady, and the importance of doing that prior to compression, which will get a more even effect from the compressor throughout the song, and with less work from the compression.

What is the proper way to do it when, let's say, the first verse is very soft, building to a crescendo on the chorus for example?

If I use volume automation – or let's say, instead, gain automation – up front to make the volume constant on the vocals, for example, I lose the effect of the softer part building into the louder part.

So, I'm getting lost in this: for purpose of applying compression evenly on both loud and soft parts of the song, yes, VA will accomplish that so that I don't squash the louder parts and only mildly compress the softer parts of the song.

I've read that a number of people simply recommend leaving the quieter part alone and only compressing the louder parts. However, that doesn't make sense to me: the softer parts can benefit from compression just as much as the louder.

As I'm thinking it through, I'm thinking this would be a way to do it, but I'd like validation (or correction):

I "think" the way to do it would be to use gain adjustment (either by envelope automation, or just cutting the clip up into sections and using the trim tool; I use REAPER, btw) as the first step to make the volume constant for the entire song, prior to compression. So, consistent gain-staging so the clips are consistent around -18 or so. THEN use compression, which would now be applied evenly across all sections of the song. THEN use volume automation, post-fader, to adjust the volume of the softer parts to taste. That seems to me to be the correct way to apply compression equally across soft and loud parts of the song while at the same time preserving the volume changes.

Am I right here? Am I thinking things through clearly? Am I missing something? Is there an easier way to do this?

Am I making any sense at all?
:)
Thanks, all.
Tesgin
Click to expand...
Sure, and yes.
But let's explore some. :)
Let's break leveling -track automation and compression, into two general camps;
'Most transparent and retaining the most natural and least effected sounding at one end. Volume and gain automation, and/or compressors..

And the more altering and effected sounding -intentionally hopefully- by way of the compressors. ..With huge ranges of variation and overlap between the two via the compressors of course.

So that leads to a few questions.
What is the general intent or goal in wanting the compression on all of the track, and/or in particular on 'the softer sections?

And, what makes soft sections of a vocal 'sound soft?

Obviously when the arrangement is quieter or more sparse, the vocal may come down some too.. but not necessarily a lot.. It's still the lead' and 'center piece..
By simple volume rides, yes. Or how about because it was also sung softer?

Those are two different things -or 'means, and convey two different 'expressions. (Poor choice of words there but I hope the meaning comes through.

Perhaps in the soft parts the singer also works the mic closer? Another variation, but the level is not the bigger part of the 'effect.

Your mention of track gain rides, and pre compression gain rides, and track fader rides post compression is spot on.
I typically have both automation lines -on tracks that call for it, and often choose to use one or the other -and both :>) , depending on the choice and 'effect on where and how it hits the compressor -or not. :>) (Sometimes it doesn't much so it's a wash' either will do.

There is also automate the compressors' threshold, but that's getting into the weeds, and a little ahead of the real questions here. :)
 
Yes to the above. I would say the most common way of mixing a lead vocal would be, first volume rides/automation, then light compression to control peaks, then often another compressor with a higher ratio and/or eq (with or without automation/dymamic eq)and/or a de-esser. Not every element is needed on every track or even every part of a track which is what makes automation critical. But yes first step is almost always going to be volume rides.
 
Gtoboy said:
Yes to the above. I would say the most common way of mixing a lead vocal would be, first volume rides/automation, then light compression to control peaks, then often another compressor with a higher ratio and/or eq (with or without automation/dymamic eq)and/or a de-esser. Not every element is needed on every track or even every part of a track which is what makes automation critical. But yes first step is almost always going to be volume rides.
Click to expand...

Thanks for this, GTB. To clarify for me, when you say "first volume rides..." do you mean chronologically that's what you do first, or do you mean first in the signal flow? I'm understanding you to be referring to signal flow, yes? I guess it could be first, but I'm interpreting you to mean that the first volume automation should feed (precede) the compressor. That would be done either by gain automation or clip trims, or by using volume envelope and then feed that signal to another buss with the compressors on it, yes?
 
Ugh. I must first point out that any given song has it's own needs.

Compression is just a tool to use when needed. It does not need to be on every track to make it work. If a part of a song is less 'on top' then it needs less or none.

Mix with your ears and less with what someone tells you you need to use.

I know it is a vague response, but it is completely correct.

Best. :)
 
Until I had been recording for nearly twenty years I never used compression. I really hated it. Then I started recording music that I didn't like, and discovered that for some genres, compression is very important. You cannot predict when you need it, and in my early years my fingers were my compressor, because I'd be wibbling them up and down like mad - the human compressor.

I'm a bit confused by the notion that you have some kind of rule. My only real 'rule' is that I put light compression on my bass, because without it, it far to easily vanishes in the mix., This is my bass, not a rule for all basses. My usual strategy is to start with a rough set fader mix, then automated the things that need to change in the verses and choruses, then once they're pretty much done, I'll ride the main Vox or instrument fader as required. Then repeated runs tweaking individual channels. For me, that's pretty well it. for the past six months I've been working on a collaborative project and I get rough mixes with no eq or inserts, which I then sort. I still have no plan until they arrive.
 
rob aylestone said:
Until I had been recording for nearly twenty years I never used compression. I really hated it. Then I started recording music that I didn't like, and discovered that for some genres, compression is very important. You cannot predict when you need it, and in my early years my fingers were my compressor, because I'd be wibbling them up and down like mad - the human compressor.

I'm a bit confused by the notion that you have some kind of rule. My only real 'rule' is that I put light compression on my bass, because without it, it far to easily vanishes in the mix., This is my bass, not a rule for all basses. My usual strategy is to start with a rough set fader mix, then automated the things that need to change in the verses and choruses, then once they're pretty much done, I'll ride the main Vox or instrument fader as required. Then repeated runs tweaking individual channels. For me, that's pretty well it. for the past six months I've been working on a collaborative project and I get rough mixes with no eq or inserts, which I then sort. I still have no plan until they arrive.
Click to expand...

And that is the truth. My 'In my opinion' one of my best mixes has a few light stages of compression. Nothing extreme in any way. Acoustic guitars have their own bit, the vocals have three layers of slight compression. The drums have likely 5 separate compressors on the final mix.

But they are used for a purpose for each instrument. And very lightly by the way. Not somethong you would necessarily hear on an individual track. But the end result is the why...
 
[MENTION=131779]Tesgin[/MENTION], I'm not sure what "gain automation" is in that first post, but you have some of the things a little confused from my POV. But, as others have stated, these kinds of things might be different depending on the original track, the arrangement/mix and genre. There's no rules, per se.

So, yes, I use clip gain, for instance, to even out a performance that may be suffering from a couple mic technique problems (or even a lot), or perhaps to level a couple different takes if I'm creating a comp or punched something in later and they don't match up. I don't use gain on the pieces of the track whether clip or "automated" to "level" the overall performance. I might adjust the gain of a complete track if it helps keep my faders centered when I'm working on that initial static mix.

If you've got starkly different levels in some parts of the arrangement that might require compression threshold on one part of a track that is wrong for another part, then you can always automate that effect or even split out that part into a duplicate track with different FX and sum/bus them together for treatment as a single entity, once it's all figured out.
 
Tesgin said:
I'm not thinking so much in terms of how to make it sit with the other pieces in the arrangement, but more looking at, say, the vocal by itself: I'm reading so much about how important it is to get an even signal prior to compression, by using clip gain, or volume automation before sending to, say, a buss with compression. So just looking at compression on the vocal track, for example, if there's a lot of dynamic change between the louder choruses and softer verses, only applying compression to the track as a whole won't touch the softer parts (unless I over-compress the louder sections). Should one use just a milder compression to touch the louder peaks and not worry about the softer parts? Or is it more appropriate to compress the louder parts and the softer parts equally by using the clip-gain adjustment I described earlier, and then use volume automation post-compression?

So I'm asking just in terms of the proper use of VA with compression on a track, with the next step then being to blend it in with the arrangement. Not sure if I'm explaining myself clearly.

Thanks,
TB
Click to expand...

The arrangement is going to dictate what VA and or compression is needed or not needed-if the track needs more dynamics why would you want to get rid of them? I see what you are asking, but I think you may be going about this a bit backward. The vocals are either important or they're not-if not just squash 'em and stick 'em in at the level you want. If they are supposed to be the "lead" and out front then you're generally going to be better off doing only what is necessary to make the performance work in a way that fits the songs intent. If you are just using VA to get an even level to compressor, the compressor is going to be doing little anyway so why bother? My point is, before doing anything you should have an end result in mind and that will dictate exactly what you will be doing to the track and in what order. The order I gave was just a common "for instance" way of working, and yes, I was talking about chronological order, since if VA achieves the result you want, why would you need to add compression except as a "color" or "effect"t that works for the arrangement(planned for)?
 
Gtoboy said:
If you are just using VA to get an even level to compressor, the compressor is going to be doing little anyway so why bother?

... if VA achieves the result you want, why would you need to add compression except as a "color" or "effect"t that works for the arrangement(planned for)?
Click to expand...

Thanks, GTB. It's easy to get caught up in "rules" and forget the final end point in mind, as you said.

I'm not too particularly thinking in terms of cookbook rules, per se, but more understanding tools and techniques. As I'm studying, and reading, watching videos, doing trainings, I am learning that there actually are a number of mixing engineers that strongly extol the virtues of using VA and compression together. E.g., this article in Sound on Sound makes this point:

"Vocal takes that have a very large dynamic range can cause big problems for compressors: the settings that sound right for the quiet parts tend to be far too aggressive when you reach the louder sections. In this case, the answer is often to use automation to even out the level of the vocal prior to its going into a compressor — if your DAW provides post‑fade insert slots, place your compressor into one, or alternatively, bus the automated vocal track to an auxiliary or group and compress that. (It can be useful to have an additional layer of automation after the compressor in this situation!).​

That's what I'm talking about. This article from audioskills makes the same point. Graham from Recording Revolution talks about it as well (click here), and hosts an article from Rob Mayzes, who says that "most engineers use detailed volume fader automation to manually smooth out the vocal" in addition to compression, to get a more natural effect – a vocal that sounds less processed, more real.

I'm just giving some examples cuz some of the posts above are questioning why one would ever do both. As I'm listening and learning, I'm finding it's a commonly used approach, especially on vocals. It's new to me and I'm just trying to understand it as a tool.

So, to clarify, I'm not looking for an easy fix ("one size fits all"), but wanting to see how you guys, all more experienced and knowledgeable than me, do this stuff.

I liked the Mayzes article above: he talks about using gain automation instead of VA, to even out the signal before going into the compressor, cuz if you automate after the compressor, the volume of louder parts could've already been squashed. So, obviously a lot of it has to do with whether one knows how to use a compressor properly, but his point is that there are advantages to taking care of business prior to compression.

So that's where my question came from: I understand what they're saying about the value of evening things out pre-compression, but trying to understand how that wouldn't kill the dynamics of a natural performance. Unfortunately, none of the sources I cited or read talk about that. Hence my thoughts about gain automation prior to compression (per Recording Revolution and Musician on a Mission, above), followed by VA to bring the realism of softer sections, and hence my post.
 
Well if, as the quoted article states you have a "very large" dynamic range than you can certainly reduce it without getting rid of all of it(the dynamics). As for using gain automation, i think for me it would be simpler (I use Pro Tools so YMMV) to duplicate the track then just use the clip gain to reduce/raise the levels as needed to get them close to each other without losing the performance and print or render that track so that i wouldn't have extra automation to slow the session. Then you can do VA for smaller bits, like turning down one syllable here or there and again compress if wanted. I dont see much point in having a bunch of automation just for gain matching. Some helpful info IMO:
YouTube
 
Gtoboy said:
Well if, as the quoted article states you have a "very large" dynamic range than you can certainly reduce it without getting rid of all of it(the dynamics). As for using gain automation, i think for me it would be simpler (I use Pro Tools so YMMV) to duplicate the track then just use the clip gain to reduce/raise the levels as needed to get them close to each other without losing the performance and print or render that track so that i wouldn't have extra automation to slow the session. Then you can do VA for smaller bits, like turning down one syllable here or there and again compress if wanted. I dont see much point in having a bunch of automation just for gain matching. ..
Click to expand...

I haven't watched the vids yet (I may have seen them though :>)
The reason I often go for both gain and track automation is a) the detail I can have -and often want is via automation lines. If I was splitting things into chunks for 'bigger sections of level correction, likewise there would still be the detail work to do anyway. And, the bigger (wider) corrections can be done just as well with flat wide gain changes. (With the smaller detailed stuff placed in later as things progress.
If the whole track (or section for some reason?) is grossly off, I'll do the 'gain lift process (Sonar) the continue with the detailed stuff. That would more apt to be on a live tracked gig.
b) I don't see this kind of detail and options slowing projects down..?
 
mixsit said:
I haven't watched the vids yet (I may have seen them though :>)
The reason I often go for both gain and track automation is a) the detail I can have -and often want is via automation lines. If I was splitting things into chunks for 'bigger sections of level correction, likewise there would still be the detail work to do anyway. And, the bigger (wider) corrections can be done just as well with flat wide gain changes. (With the smaller detailed stuff placed in later as things progress.
If the whole track (or section for some reason?) is grossly off, I'll do the 'gain lift process (Sonar) the continue with the detailed stuff. That would more apt to be on a live tracked gig.
b) I don't see this kind of detail and options slowing projects down..?
Click to expand...

Great points. That makes sense to me. So, if I'm tracking with you (pun here is absolutely intended!) the idea I was describing of automating or balancing the gain, then going to compression, then tweaking with VA would seem very much to fit with your workflow, yes?

I also have not yet had a chance to watch the videos. Started but had to get back to work. :) Will watch this evening.

However, the beginning of the first video fits, I think, with what I'm talking about: he discusses nicely how compression applied to shape the transients can bring a track forward or push it back in the mix. The idea of gain automation prior to compression illustrates that nicely: by making the gain/trim of a vocals track, let's say, more uniform prior to compression allows more consistent shaping of transients, or sound-shaping -- or whatever the effect you're trying to accomplish with compression -- across the entire track. Then VA to bring the softer sections of the song back to where you want them should produce a more consistent, natural, dynamic track.

This is new to me! I'm not the expert here. But thinking this way seems to make sense to me. Again, if I'm up a tree or overthinking, I definitely wanna know. And another "again" -- I'm not talking about mechanically applying a cookbook technique to all situations; I'm just trying to understand different tools and approaches, I know that "if it sounds good it is good," but I also don't wanna be altogether kooky either!

:)

P.S., Great discussion, all. I really appreciate this forum. Learning tons.
 
Whatever works for you is cool. The OP was actually looking at the order of gain , VA and compression. As you pointed out, volume changes are going to be an ongoing thing(VA). But to just get a ballpark gain level to hit a comp evenly with, for me doing a bunch of automation would take longer. As I said earlier though, there is more than one way to do it and to my mind it depends on what you are starting with and where you want to end up.
 
Not trying to derail this, but the best automation is mic technique, closer in the softer parts pull away a bit in the load parts, makes the compression and automation far easier later, it also sounds better.

Alan
 
If the variation is great enough, or if I just want to use different setups. I’ll separate the parts, soft on its own track, loud on another. Apply VA to each, and then set the compression to fit each track. I might even apply different eq if I like.
 
You must log in or register to reply here.
Back
Top