Can my mix levels be too quiet for Mastering?
Old 10th January 2011
  #1
Gear interested
 

Thread Starter
Can my mix levels be too quiet for Mastering?

Hi there Slutz!

Engineer from Ireland here, question I couldn't find anywhere else on the forum for you guys, sorry if it is but it doesn't seem to be.

I'm about to send an album off to be mastered. Most of the tracks have peaks around -6dbfs and an RMS of -12

But some have peaks of only -10 dbfs and RMS of -24 or even lower. Are these lower levels too low for mastering? I am happy with the sound of the mixes. I may just have been too cautious with my gain staging.

PRO ME's, do I need to take action? If so, what do you think I should do?

Thank you in advance!
Old 10th January 2011
  #2
Gear Head
 

That will not be a problem for your mastering engineer of choice. He/she can easily bring the peak level up themselves either digitally or analog. If you like the way the mixes sound now, leave them alone and send them in.
Old 10th January 2011
  #3
Gear Head
 
brethes's Avatar
 

Verified Member
Not a problem is the mixes are 24 bit.
Old 10th January 2011
  #4
Gear maniac
 
Mastering7's Avatar
 

Quote:
Originally Posted by nightsearcher View Post
Hi there Slutz!

But some have peaks of only -10 dbfs and RMS of -24 or even lower. Are these lower levels too low for mastering?
Absolutely not if the sign/noise ratio is OK.
Old 10th January 2011
  #5
Gear interested
 

Not a problem in 24 bit. If you are working in 16 bit you could find some of the low level material to be a little lifeless.
Old 10th January 2011
  #6
Lives for gear
 
Mark D.'s Avatar
 

At 24 bit, a -6db peak level for a mix is actually great. It's also useful to be sure you are well
below -0db at earlier stages (hitting the bus, tracks and hitting each plug-in). You will hear a
difference. What threw me off and maybe others, is you'd said it's -6db peak and -12db RMS.
-12db RMS is loud for an unmastered mix. At -6db peak you will be left with just 6db in range.

A mix at a more ideal level (ie. -6db, to -3db, peak), if dynamic, should be more like -20db to
-14db RMS. -6db peak, at -12db RMS is basically -6db RMS, just turned down 6 db. -6db is a
very loud master, even for the loud masters we hear now. Turned down to -6db peak (-12db
RMS) it would still sound squashed. What is on your output bus (limiting, compression) there?
Old 10th January 2011
  #7
Lives for gear
 
dcollins's Avatar
 

Verified Member
Quote:
Originally Posted by nightsearcher View Post
Hi there Slutz!

Engineer from Ireland here, question I couldn't find anywhere else on the forum for you guys, sorry if it is but it doesn't seem to be.

I'm about to send an album off to be mastered. Most of the tracks have peaks around -6dbfs and an RMS of -12

But some have peaks of only -10 dbfs and RMS of -24 or even lower. Are these lower levels too low for mastering? I am happy with the sound of the mixes. I may just have been too cautious with my gain staging.
The level is fine, but there is no point in giving RMS readings (is there ever?) just the peak.


DC
Old 10th January 2011
  #8
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
Unless you mixed to analog tape you are probably fine.

A mastering project that peaked to -30 was what tipped me off to the fact that it is lots lots better to record digital audio too low than too high.
Old 10th January 2011
  #9
Banned
 

Not for nothin', but, the level that a mix comes in at doesn't really tell you much. All of the tracking could have been done too low, with a lousy signal to noise ratio, and then mixed so the 2-mix is right at -3dBfs. Or the tracking could all have been done too high, with a bunch of digital hash from overs.

The crucial point for snr is at the initial tracking. You only get one chance to do the initial conversion. Too high is no good, and too low is no good.

Too low can be just as bad as too high.

Maybe we're at the point where we can start recommending that people make a reasonable attempt to understand and record at optimum levels. That's good engineering, analog or digital, I think.
Old 11th January 2011
  #10
Gear Head
 

Without question when the mix is up par within peak Db's there is no doubt a mastering engineer can lift it. Seems okee to me especcialy when you post the values given. Nothing wrong with the values. If these songs need to be soft go for it. If not there is really something wrong with the way you hear your mixing levels. This should be a setpoint value to your choosing and something you are comfortable with. Getting to know the volume of a mix is an important thing in mixing. Really make this a way of life. Not that you should listen everything at that volume when you just want to relax with some music. But for mixing it's a very good starting point. Also if you are mixing really well you need two days give or take depending on who it is and how often he or she does it just to really start with a mix. A focussed mix is a state of mind!! Set a point where your ears are really starting to think ... oww this is loud!!... Mark that point so whenever you start mixing you set it that point. Always the same point no matter what. You will feel it when you get accustomed to that. Start a mix with the highest dynamic range possible Vocals for example. Make it so that it doesn't sound too loud on that set loud level. Wait with bass until the rest sits in the mix. Low freq's will cloud judgement.. Not perse.. There are no rules just guides. But once the mix balances out without that it will fit right in. And it's only my vision on how to get thing going.

Good luck with that. But i try to help.

Misja.
Old 11th January 2011
  #11
Lives for gear
 
Mark D.'s Avatar
 

Quote:
Originally Posted by dcollins View Post
but there is no point in giving RMS readings (is there ever?)
There certainly is if it's -6db peak at -12db RMS. I'm surprised I'm the only
one who'd noticed that. I'd still like to hear from the OP on those numbers.
Old 11th January 2011
  #12
Lives for gear
 
Jerrick's Avatar
 

Quote:
Originally Posted by Mark D. View Post
There certainly is if it's -6db peak at -12db RMS. I'm surprised I'm the only
one who'd noticed that. I'd still like to hear from the OP on those numbers.

That was the first thing that stood out for me as well. A lot of modern masters out right now I have seen are sitting around -12.
Old 12th January 2011
  #13
Lives for gear
 
dcollins's Avatar
 

Verified Member
Quote:
Originally Posted by Mark D. View Post
There certainly is if it's -6db peak at -12db RMS. I'm surprised I'm the only
one who'd noticed that. I'd still like to hear from the OP on those numbers.
What is weird about that? So there is roughly 6dB from average to peak, it's normal.

Maybe the -6 peak is one hit. There go those numbers again......................


DC
Old 12th January 2011
  #14
Banned
 

Old 12th January 2011
  #15
Lives for gear
 
bgrotto's Avatar
Quote:
Originally Posted by dcollins View Post
The level is fine, but there is no point in giving RMS readings (is there ever?) just the peak.


DC
But there *is* a relationship to a master's "loudness potential" (I think I borrowed that term from another GSer) and the mix's RMS. A louder mix will yield a louder master with less processing, and thus, more transparently.

So, for the mixer concerned with the possibility of a mastering job affecting mix balances (which probably includes pretty much all of us...), having realistic expectations relative to the loudness of the unmastered mix is very important. A louder mix yields a louder master with fewer artifacts.

At least, that's been my experience.
Old 12th January 2011
  #16
Lives for gear
 

Quote:
Originally Posted by bgrotto View Post
A louder mix yields a louder master with fewer artifacts.
When you write it like that you make it sound like fewer total artifacts...but it is really just fewer mastering artifacts. Let me write out a simplified and silly example where all artifacts are equal and you can somehow number them:

Unsmashed mix -> hot master creates 20 artifacts for a total of 20 artifacts.

Smashed mix creates 15 artifacts -> hot master creates 5 artifacts for a total of 20 artifacts.

There is a very real upside that you get to mix "through" the artifacts taking them into account and compensating for the nastiness while you still have a great amount of control with the multi-track mix. But there is a very real downside that now your mix is ingrained with that crap and when the world wakes up and starts asking for clean remasters of music from "the lost decade", there will be nothing undamaged to remaster from.
Old 12th January 2011
  #17
Lives for gear
 
bgrotto's Avatar
Quote:
Originally Posted by Cheebs Goat View Post
When you write it like that you make it sound like fewer total artifacts...but it is really just fewer mastering artifacts. Let me write out a simplified and silly example where all artifacts are equal and you can somehow number them:

Unsmashed mix -> hot master creates 20 artifacts for a total of 20 artifacts.

Smashed mix creates 15 artifacts -> hot master creates 5 artifacts for a total of 20 artifacts.

There is a very real upside that you get to mix "through" the artifacts taking them into account and compensating for the nastiness while you still have a great amount of control with the multi-track mix. But there is a very real downside that now your mix is ingrained with that crap and when the world wakes up and starts asking for clean remasters of music from "the lost decade", there will be nothing undamaged to master from.

I'm sorry, I wasn't clear in what I meant. By "artifacts", I meant it in terms of a change to the mix's balance that is contrary to the mixer's intent. A common example might be the snare getting smashed down to where it's too quiet and no longer correctly drives the song rhythmically. Or a change in the way the bottom end relates to the vocal. Or even something as subtle as the release time on a compressor affecting the groove. Or what ever.

I shouldn't have used the word "artifact"; I chose that one poorly. My bad!
Old 12th January 2011
  #18
Gear interested
 

Quote:
Originally Posted by bgrotto View Post
I'm sorry, I wasn't clear in what I meant. By "artifacts", I meant it in terms of a change to the mix's balance that is contrary to the mixer's intent. A common example might be the snare getting smashed down to where it's too quiet and no longer correctly drives the song rhythmically. Or a change in the way the bottom end relates to the vocal. Or even something as subtle as the release time on a compressor affecting the groove. Or what ever.

I shouldn't have used the word "artifact"; I chose that one poorly. My bad!
That should not be the case, as long as it is only raised level-wise without any compression/limiting involved

regards
Old 12th January 2011
  #19
Lives for gear
 
Mark D.'s Avatar
 

Quote:
Originally Posted by dcollins View Post
about that? So there is roughly 6dB from average to peak, it's normal. Maybe the -6 peak is one hit.
Valid point but if that were so, it would indicate an even smaller dynamic range. An unmastered mix, at -12db
RMS, with the peak anywhere above (ie. -6db or -0db) would be considered loud and likely compressed. If it's
only peaking at -6db, then it has to be heavily peak limited. Go take any song, any genre, ever. Set the peak
limiter to a -6db ceiling and raise the input gain it until you get near -12db RMS. It's obliterated. That's why I
think he misspoke when he said -6db peak & -12db RMS. Perhaps he meant to type -22db RMS (much better).
Old 13th January 2011
  #20
Lives for gear
 
bgrotto's Avatar
Quote:
Originally Posted by Chris Riot View Post
That should not be the case, as long as it is only raised level-wise without any compression/limiting involved

regards
Well....yeah...I sorta figured it goes without saying that a flat transfer would result in fewer artifacts (regardless of your definition of "artifacts").

Or are you talking about an ME using only EQ? Even then, there's potential for inadvertent e-balancing.

I'm confused.
Old 13th January 2011
  #21
Gear interested
 

Quote:
Originally Posted by bgrotto View Post
Well....yeah...I sorta figured it goes without saying that a flat transfer would result in fewer artifacts (regardless of your definition of "artifacts").

Or are you talking about an ME using only EQ? Even then, there's potential for inadvertent e-balancing.

I'm confused.
What I wanted to say was, that a mix which is brought up in his level by pure (digital) gain should not loose his mix balance, as long as there is no compression or limiting (or clipping) on the master bus

for example, if you highest peak in the mix is at -12dbfS, if you bring up the level in the digital domain around 9db, it should still have the same mix balance at -3dbfS

EDIT: I posted that example, because the OP asked about his Mix beeing "too quiet" for mastering. It should only show that there's no problem with giving out a -12dbFS (or even lower) peak mix to mastering,because a higher peaking mix has no actual benefit for having fewer "artifacts" when being mastered. (talking about a 24bit file)
Old 21st May 2011
  #22
Gear interested
 

Thread Starter
Just realized I forgot to thank everyone for their help here!
The mixes turned out great despite their low level
Old 22nd May 2011
  #23
Lives for gear
 
MASSIVE Master's Avatar
 

Verified Member
Quote:
Originally Posted by nightsearcher View Post
The mixes turned out great despite their low level
But that's not a low level -- That's a perfectly reasonable and normal level.

In the past 15 or 20 years, I have never -- Not once -- Ever -- had a project come in that was truly "too low" to work with.

Projects with F'd up SNR's? Sure. Projects that were exceptionally low for no particular reason? Sure. Often, as a matter of fact.

The closest ever was a project that came in (I still do most projects for this particular engineer) that exported for some (to this day as yet unknown) reason peaking 48dB lower than the engineer had it.

Opened up the files and we thought they were "black" for some reason. But I could hear something going on and the meters were bouncing a bit -- Hit "N" (which "normalizes" the object) and 50-some dB later, there they were. And they sounded absolutely fine.

Sure -- I sent him back and he exported them again and if you A/B'd them, you could hear a rather subtle difference between the two (for the record, we didn't actually decide that one was "better" than the other -- There was just "a difference" of sorts).

Point is -- It's really, REALLY hard to have levels that are "too low" in 24-bit digital.
Quote:
A mastering project that peaked to -30 was what tipped me off to the fact that it is lots lots better to record digital audio too low than too high.
Amen, brother. Nolo Contendre.
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Kscope / So much gear, so little time!
11
DCtoDaylight / So much gear, so little time!
21
illacov / Mastering forum
0
MichaelAngelo / Mastering forum
47

Forum Jump
 
Register FAQ Search Today's Posts Mark Forums Read

SEO by vBSEO ©2011, Crawlability, Inc.