Reduce Track Volume or Reduce Master Volume?

DarkFriend

New member
So while mixing, I typically start with the drums, bring in the bass, and everything else one at a time. The individual track volumes can get really loud and while the DAW shows the track is clipping, there's no audible distortion.

When the mix is complete, I reduce the master volume to an appropriate level for rendering.

Questions:

1) Should my master volume be set at 0 at all times, and reduce the track volumes to an appropriate level?

or

2) Should my track volumes sit around 0, and reduce the master volume to appropriate levels?

3) Is there any difference in sound quality between the two above?

My recordings always sound good, but I'm wondering if all of the track clipping is really having an effect on the quality.

Thanks
 
When you say "sit around 0", if you mean 0dBFS on the meter, the answer is you should lower your recording levels before anything else.

Then during mixing, adjust your track volumes as necessary, leaving your mix buss faders at unity gain. If you start to or approach clipping on your mix buss, then work on reducing the individual track levels rather than throttling the mix buss.

Assuming you're mixing in-computer and not on an analog mixer, here's a technique I like to use: from the start, set the levels automation path on each track to somewhere around -3dB throughout the entire song before trying your first "faders up"' rough mix. It's nice to have a blank template set up that way on your computer that you can copy and use for each project so speed your set-up time.

Then set up your rough mix (or submix, if you're working on drums first, for example) by setting rough fader levels, never exceeding unity gain on the track faders - i.e., your loudest track will be set for fader unity gain. Once you have a rough mix set, fine tune the mix using your automation path instead of your track faders. Not only will this limit your added track gain to 3dB per track (if you need more gain separation between tracks, you need to drop the other track level), but it will facilitate the mindset of using automation to actually *MIX* your tracks from moment-to-moment instead of just layering flat-gain tracks.

This combination of forcing yourself to mixing downwards as much as you mix upwards and actually dynamically mixing track levels really results in a better-sounding mix, and not only tends to eliminate potential clipping issues on the mix buss, but also tends to result in a mix buss mix that holds together in volume "mastering" better.

G.
 
This has interested me for a while. When mixing 'in the box' and i overload the master, i pull it down and it stops clipping. So in a sense the clipping isn't destructive in the signal flow? Is this because the summed signal of all the tracks is buffered before the master? In an analog signal flow i assume if you overloaded the input of the master then lower the master slider after this you'd still have distortion, like with a guitar overdrive?

For reference i never move the master, it just made sense to me to stop clipping at source and lower individual faders. In some programs it seems to make a big difference to the final fidelity (FL Studio especially for some reason) as opposed to dropping the master.
 
jgharding hit the nail on the head. Yes, the question is if there's a quality difference between reducing the individual track volumes to a level where they don't clip, or leave them where the tracks clip and reduce the master bus so the entire mix doesn't clip.

As jgharding mentioned, in the digital world, there's no audible clipping or distortion like there would be with tape. But we are all after the same thing - getting the best sounding results.
 
2) Should my track volumes sit around 0, and reduce the master volume to appropriate levels?

If you leave all the track faders at zero and just move the master volume, you're not really mixing. Your just turning the overall volume up and down. :p

I jest.

I never really move my master fader from zero. When I mix, I start with every fader at zero. I try my best to set the levels when I'm recording, by recording at an appropriate input level. If I've done a decent job of getting the input levels right, none of the faders really need to go far from 0. Sure, they get moved, but not by a huge amount unless I'm having to compensate for the effect of a plugin or something.
 
Yeah, naturally the tracks don't all sit at zero..

So sounds like the idea is to leave the master at zero and reduce the individual track volumes, perhaps starting at -3db like Glen suggested.

I do have some cases where an effect increases the track volume a lot, so it has to come down considerably.

In addition, I like mixing multiple copies of the same track with different effects on each and doing submixes. This will always create submixed tracks that are louder.
 
jgharding hit the nail on the head. Yes, the question is if there's a quality difference between reducing the individual track volumes to a level where they don't clip, or leave them where the tracks clip and reduce the master bus so the entire mix doesn't clip.

As jgharding mentioned, in the digital world, there's no audible clipping or distortion like there would be with tape. But we are all after the same thing - getting the best sounding results.

When my individual tracks clip I hear it. If your individual tacks are clipping turn them down end of story. If they are clipping when you have the fader at unity then you are probably tracking to hot unless it is being caused by a plugin.
 
So sounds like the idea is to leave the master at zero
This is probably in large part because I grew up on analog mixers and not DAWS, I admit, but I've always kind of looked at it this way; the master faders are great for performing fade-outs and fade-ins, but levels are best set by the upstream gain structure and the mix levels. (the name "faders" does come from somewhere! ;) )

This doesn't mean that one can never use the buss faders for throttling overall mix volume, it's great to have control at as many gain stages as possible, and to have that *option*, but it does kinda mean (to this way of thinking, anyway) that the main purpose of the main faders is not to set levels, and one should not nomally *plan* to use the main faders for this purpose.

G.
 
In the box you often have unlimted bussing, so if you find your master is geting clipped but you have your drums balanced just right - send the drums to a new stereo chanel, then pull that down. That way you don't have to pull down say, five sliders by the same amount, or everything down a bit.

Setting up a couple of groups for groups of tracks you're happy with is a great way of pulling down offending section easily without disturbing other sub-balances. It's also an easy way to pull your mix to zero if you need inspiration or are a bit stuck with it, by dropping all the groups and then starting by bringing up the drums, vocals, then bass and getting them balanced, then dropping in the other instruments and details. It's really easy to set it up quickly in FL Studio and Reaper, and in fact most other sequencers!
 
Interesting (albeit somewhat irrelevant/silly) idea: what if, from tracking on, you started with the master fader at +6db, and mixed that way? You'd be guaranteed to max out at -6 if you made sure your mix never went above 0db on the master right?
 
When you say "sit around 0", if you mean 0dBFS on the meter, the answer is you should lower your recording levels before anything else.

Then during mixing, adjust your track volumes as necessary, leaving your mix buss faders at unity gain. If you start to or approach clipping on your mix buss, then work on reducing the individual track levels rather than throttling the mix buss.

Assuming you're mixing in-computer and not on an analog mixer, here's a technique I like to use: from the start, set the levels automation path on each track to somewhere around -3dB throughout the entire song before trying your first "faders up"' rough mix. It's nice to have a blank template set up that way on your computer that you can copy and use for each project so speed your set-up time.

Then set up your rough mix (or submix, if you're working on drums first, for example) by setting rough fader levels, never exceeding unity gain on the track faders - i.e., your loudest track will be set for fader unity gain. Once you have a rough mix set, fine tune the mix using your automation path instead of your track faders. Not only will this limit your added track gain to 3dB per track (if you need more gain separation between tracks, you need to drop the other track level), but it will facilitate the mindset of using automation to actually *MIX* your tracks from moment-to-moment instead of just layering flat-gain tracks.

This combination of forcing yourself to mixing downwards as much as you mix upwards and actually dynamically mixing track levels really results in a better-sounding mix, and not only tends to eliminate potential clipping issues on the mix buss, but also tends to result in a mix buss mix that holds together in volume "mastering" better.

G.

When you set the master fader to unity gain, do you mean the level indicator should never exceed 0db(i.e. going into yellow/red zone) or the fader itself should appear be positioned at the 0db marking? Do I sound confused, well yes I am.
 
When you set the master fader to unity gain, do you mean the level indicator should never exceed 0db(i.e. going into yellow/red zone) or the fader itself should appear be positioned at the 0db marking? Do I sound confused, well yes I am.

Generally Both
The level should never go above zero DBFS as this will result in harsh digital clipping.
Generally the master fader is left alone at zero (unity) except for fade ins/outs of he entire mix
 
Assuming you're mixing in-computer and not on an analog mixer, here's a technique I like to use: from the start, set the levels automation path on each track to somewhere around -3dB throughout the entire song before trying your first "faders up"' rough mix. It's nice to have a blank template set up that way on your computer that you can copy and use for each project so speed your set-up time.

Then set up your rough mix (or submix, if you're working on drums first, for example) by setting rough fader levels, never exceeding unity gain on the track faders - i.e., your loudest track will be set for fader unity gain. Once you have a rough mix set, fine tune the mix using your automation path instead of your track faders. Not only will this limit your added track gain to 3dB per track (if you need more gain separation between tracks, you need to drop the other track level), but it will facilitate the mindset of using automation to actually *MIX* your tracks from moment-to-moment instead of just layering flat-gain tracks.

This combination of forcing yourself to mixing downwards as much as you mix upwards and actually dynamically mixing track levels really results in a better-sounding mix, and not only tends to eliminate potential clipping issues on the mix buss, but also tends to result in a mix buss mix that holds together in volume "mastering" better.

G.
I love this! Next mix I do, I'm giving this a go. Seems a hell of a lot easier than trying to nudge everything channel down the exact same amount when it is tracked too hot.
 
When you set the master fader to unity gain, do you mean the level indicator should never exceed 0db(i.e. going into yellow/red zone) or the fader itself should appear be positioned at the 0db marking? Do I sound confused, well yes I am.
Bristol Posse got it right, but just to try and make sure I address the probable confusion with the terminology, "unity gain" refers specifically to the physical "0" position on the fader control itself, somewhere around 2/3rds to 3/4ers of the way up the fader. "Unity gain" simply means no signal boost, no signal cut; you're passing the signal through that fader without changing the volume at that point.

When you do that, what the signal reads on the meter at that point will be totally dependent on the strength of the signal going into the fader since the unity gain setting on that fader won't be changing the volume. By setting the master bus to unity gain, this ensures you're looking on the maim meters at the actual mix volume and you can that way monitor your mix levels as you mix.
I love this! Next mix I do, I'm giving this a go. Seems a hell of a lot easier than trying to nudge everything channel down the exact same amount when it is tracked too hot.
Exactly. I found myself often having to pull tracks down during mixing, almost always at least one track.

Pulling down the channel volume (input/trim volume, track fader, etc.) doesn't work as well for me, because that sets a maximum track volume above which one cannot automate. If you want to the go above that, you have to change the track volume and then adjust the entire automation track in order to keep the track at the same relative mix level, which is a pain in the ass.

So I figured that I'd just use the man track volume to set the maximum track level, and then set the automation down a few dB on each track as starting levels, leaving me some "headroom" in the automation tracks to be able to boost the track up to as high as maximum track volume if I wished for certain mix points. Think of the volume automation like if it were a fader in and of itself, and you're starting with the fader set most but not all the way up (kind of like a unity gain setting, in a way).

This allows for plenty of flexibility in actual mixing - i.e. fader jockeying via automation - while keeping the amount of global track re-tweaking to a practical minimum .

But the unexpected beauty to this technique, Chibi, which I didn't plan for, but which works out nicely, is that the mix levels seem to just naturally work themselves out that way. For a typical medium density rock or pop mix, I'll typically wind up with a final mix with an RMS within a couple of dB of my A/D converter calibration levels, pretty much like god (and the AES) intended it ;), without having to take the master levels off of unity gain.

G.
 
If you feel most comfortable maxing out all your tracks and pulling down the master fader until it stops clipping, you probably won't have any issues, but it is bad practice. In the box, everything is 32 or 64 bit, so even if you go over 0dB, it won't clip. If you bring the master fader down, all the information is still there.

But, keep in mind that if you use a plugin that doesn't like you going over 0dB, then that plugin will distort.

So really, in most cases it won't matter, other than the fact that if you got so used to doing it that way, if you ever sit in front of an analog board, you'll find it clipping all over the place.
 
So I figured that I'd just use the man track volume to set the maximum track level, and then set the automation down a few dB on each track as starting levels, leaving me some "headroom" in the automation tracks to be able to boost the track up to as high as maximum track volume if I wished for certain mix points. Think of the volume automation like if it were a fader in and of itself, and you're starting with the fader set most but not all the way up (kind of like a unity gain setting, in a way).

This allows for plenty of flexibility in actual mixing - i.e. fader jockeying via automation - while keeping the amount of global track re-tweaking to a practical minimum .
G.

It seems really logical. But how do you set maximum track level using the main track volume? Do you mean the faders on the vst mixer? I thought they'd be overwritten once the automation was on.

I have just learned about the master fader issue and now I am in the thick of remixing 6 songs, because the master fader was set all below 0db on each song. I found myself going into each track adjusting the automation level and it's so much work! I wonder if there's an easier way...

Thanks much for the response!!
 
A lot of the posts allude to the real point here. That is signal flow and distortion. If you have mixed on an analog console, you will notice that turning down the master fader of a too hot, distorted, or clipping mix simply gives you a softer (or less loud) distorted mix. The distortion is occurring as all of the tracks sum together on their way to the master out. The tricky thing is that many DAWs cheat for you. I know that Pro Tools does this for sure and imagine that most of the others are the same. When you turn down the master, the DAW actually adjusts the output level of each track going to that master buss equivalently therefore getting rid of the distortion of the master summing buss. If you know that your DAW does this, you can get away with this "cheating" fix rather than go to the trouble of lowering every track.

But...

I call this cheating because if you mix this way you are lying to yourself and missing the boat on your bigger distortion, clipping, and signal flow issues. If you are clipping the master, I guarantee that you are clipping between plug-ins. Some posts referred to not thinking that there were issues because there was no audible distortion. How do you define audible distortion? Is it by listening in you poor acoustic environment through your M-Audio interface and Behringer speakers. Guess what, there is a lot of distortion that you are missing (it is a lot more subtle than distortion on a guitar amp). This is when flying by the instruments is a good idea. If you get your signal flow in order so that you aren't close to clipping between your plug-ins and between your tracks and your master, your mix will all of a sudden sound much clearer. It is because all of that "non-audible distortion" adds up to be audible lack of clarity for your mix. Get your signal flow in order, mix by the meters (or fly by the instruments), and keep you master at unity or 0 (meaning that you are neither turning the signal up or down and the meter is showing exactly what you are sending to the master buss). Then your mixes will magically become much better without getting rid of your poor acoustic room, M-Audio interface, and Behringer monitors.

Ben
theDAWstudio.com
 
It seems really logical. But how do you set maximum track level using the main track volume? Do you mean the faders on the vst mixer? I thought they'd be overwritten once the automation was on.
This may vary between different editors, but what I'm describing is setting the initial track "trim" or volume level, usually found in the track strip controls. This can also be done by simply setting a starting level for the track fader. Whether the fader "physically" (i.e. graphically) follows the automation curve depends upon the editor you use and how it's configured in that regard.

But either way, the idea is that you're setting up an initial "rough mix" with those initial fader settings. They will in essence wind up being the "maximum automation levels" for each track, since a 100% automation level will equal that fader level.

Now, when you set up this rough mix on a project template that already has the automation level on each track set to, say, -3dB, this allows you to use the automation band as a seat-of-the-pants mix fader (which is really what it's supposed to be), that pulls everything a bit down to compensate for the overall volume increases when summing tracks together, as well as leave some "headroom" in that automation to be able to ride the track louder if/when need be.

Here's a partial sceenshot from a project I did in Audition a while back, with some captions added in to highlight parts of the organ track, just as an example. Note how the initial track "trim" volume is set to +7dB (it was tracked rather quietly), as shown both on the track control and the track mixer fader. I also started with an automation band set to - 3dB across the whole length of the song, and used that as the "centerline" (so to speak) for the automation. I have arrowed to highlight a couple of the areas in that track where that -3dB level still remains. You can see how I ride that automation level up and down as necessary for the mix.

But if I should decide somewhere along the way that the whole organ tracks needs to go up or down, that the +7dB trim level is too high or to low relative to the rest of the mix, I can pull that up or down on the fader or in the track control area without having to go and make any changes whatsoever to the automation track. the relative automation levels (i.e. the fader jockeying mix plan) remains the same, I'm just boosting or cutting the overall global level of the track:

aa_automation_levels.jpg


The cool part is that if you need to adjust the overall track volume, say, for example, bring the whole track down 2dB relative to the mix, you can just pull down the main track volume/fader at the start by 2dB without having to make any changes to the automation band. Which should address this last question pretty well, if I understand it correctly:
I have just learned about the master fader issue and now I am in the thick of remixing 6 songs, because the master fader was set all below 0db on each song. I found myself going into each track adjusting the automation level and it's so much work! I wonder if there's an easier way...

HTH,

G.
 
This may vary between different editors, but what I'm describing is setting the initial track "trim" or volume level, usually found in the track strip controls. This can also be done by simply setting a starting level for the track fader. Whether the fader "physically" (i.e. graphically) follows the automation curve depends upon the editor you use and how it's configured in that regard.

But either way, the idea is that you're setting up an initial "rough mix" with those initial fader settings. They will in essence wind up being the "maximum automation levels" for each track, since a 100% automation level will equal that fader level.

Now, when you set up this rough mix on a project template that already has the automation level on each track set to, say, -3dB, this allows you to use the automation band as a seat-of-the-pants mix fader (which is really what it's supposed to be), that pulls everything a bit down to compensate for the overall volume increases when summing tracks together, as well as leave some "headroom" in that automation to be able to ride the track louder if/when need be.

Here's a partial sceenshot from a project I did in Audition a while back, with some captions added in to highlight parts of the organ track, just as an example. Note how the initial track "trim" volume is set to +7dB (it was tracked rather quietly), as shown both on the track control and the track mixer fader. I also started with an automation band set to - 3dB across the whole length of the song, and used that as the "centerline" (so to speak) for the automation. I have arrowed to highlight a couple of the areas in that track where that -3dB level still remains. You can see how I ride that automation level up and down as necessary for the mix.

But if I should decide somewhere along the way that the whole organ tracks needs to go up or down, that the +7dB trim level is too high or to low relative to the rest of the mix, I can pull that up or down on the fader or in the track control area without having to go and make any changes whatsoever to the automation track. the relative automation levels (i.e. the fader jockeying mix plan) remains the same, I'm just boosting or cutting the overall global level of the track:

The cool part is that if you need to adjust the overall track volume, say, for example, bring the whole track down 2dB relative to the mix, you can just pull down the main track volume/fader at the start by 2dB without having to make any changes to the automation band. Which should address this last question pretty well, if I understand it correctly:

HTH,

G.

oh wow that's pretty cool. I can see what you mean now. I use Cubase LE and am not sure if it has a cool feature like that. I was a big fan of Cool Edit Pro before I switched. So there might be hope after all, instead of slaving away my time on those automations... A big thank you, Glen!!
 
oh wow that's pretty cool. I can see what you mean now. I use Cubase LE and am not sure if it has a cool feature like that. I was a big fan of Cool Edit Pro before I switched. So there might be hope after all, instead of slaving away my time on those automations... A big thank you, Glen!!
I have never tried LE, but I'd be surprised if it didn't. If you got automation you've got the ability - or at least you should. Think of it this way:

The automation control is nothing more than control of the track fader. The track fader on any mixer is nothing more than an an up/down volume control for *adjusting* the the playback volume of the track from it's input level. But you still gotta be able to set the initial track level, whether it's by a separate trim control or simply by the starting point for the input fader.

Automation is nothing more than a control saying what percentage of the initial fader level you want to have it at at that point in time on the timeline. All the way up in the timeline is 100% of that initial level. But that actual initial level still needs to be set.

Actually, CEP/early Audition as shown is kind of a hybrid setup; it has the seperate locations for setting the initial "trim", but as can be seen from the faders on that screenshot, that trim is really just moving the fader to that volume level. I'd prefer for them to actually be entirely separate controls where the trim level does not affect the fader, just like it is on an analog mixer in the real world. But even if you don't have that, the basic ideas for using automation remain the same.

Have fun with it! Until one fully embraces automation in DAW software, they are not really mixing yet.

G.
 
Back
Top