What's the typical voulume to mix/master to?

[squibble94]

I just recorded a new song, but I don't know if it's too loud or too quiet. It's mixed to about 0 db. Is this good?

P.S. The song's not finished, so I won't post it anywhere until I'm finished. I should be done this week.

 

[mattr]

By loudness war standards, even having not heard your song I can confidently say it probably won't be classed as too loud! Only exception would be if your song was simply a single square-wave synth line of the maximum possible amplitude (and even then some deaf person would probably still suggest another pass through the L2)!

Assuming we're talking about digital meters, your song can peak at a maximum of 0dbfs, but it mustn't go anywhere over. Ideally you should be allowing a bit of headroom at the top (which by today's standards usually means about -0.1db!), and if things are pushing right on 0db then you might also have to watch out for inter-sample peaks.

But peak values say little about the overall perceived 'loudnes' of the program material. You could have one song which is really dynamic and only has a few peaks nearing 0db, whilst another one is mixed with heavy amounts of compression then squashed to death with a limiter so it is always pushing on 0db. Both peak at 0db, but one would be 'louder' than the other. Loudness is a very difficult thing to define though. I've found I sometimes perceive heavily limited songs to be quieter - it is the contrast and dynamics that make things seem 'loud'... your brain doesn't register something as being loud unless it has something quiet to make a comparison to.

I've waffled on a bit

Basically, post a clip once its done and I'm sure you'll get some pointers. I wouldn't want to recommend ruining your music by trying to compete with commercial 'loudness standards', but lets say we may find you are leaving unnecessary amounts of headroom and the track could be boosted a bit to fill this up without affecting the dynamics. We don't know until we hear it...

 

[Waltz Mastering]

I just recorded a new song, but I don't know if it's too loud or too quiet. It's mixed to about 0 db. Is this good?

If your song is mixed to where your peaks are hitting close to 0 dbfs, that is not ideal.

For mixing in a 24 bit system, your mix can peak any where around -12 dbfs and be fine.

But when you say mix/master it leads me to believe your putting limiting and other processing on your master channel as your mixing, in which case all bets are off.

 

[Massive Master]

I'd go with at least a few dB of headroom (as TW mentioned, 12, even 20 is fine also).

THAT SAID - the peak value of the mix has very little to do with the apparent volume. You can have an extremely quiet mix that peaks at -0.1dBFS and you can have an extremely loud mix that peaks at -3dBFS.

 

[NYMorningstar]

If your song is mixed to where your peaks are hitting close to 0 dbfs, that is not ideal.

For mixing in a 24 bit system, your mix can peak any where around -12 dbfs and be fine.

I'd go with at least a few dB of headroom (as TW mentioned, 12, even 20 is fine also).

To the average Joe(like me) this seems to say to mix to -12dbfs and up to and "even 20 is fine".

The OP is talking about mixing/mastering and you guys switched to tracking I'm pretty sure . It's like Tom started talking about mixing and ended up talking about tracking and John started talking about tracking and ended up talking about mixing but anyway, I'm sorry it's confusing and I need another coffee

 

[dintymoore]

I just recorded a new song, but I don't know if it's too loud or too quiet. It's mixed to about 0 db. Is this good?

P.S. The song's not finished, so I won't post it anywhere until I'm finished. I should be done this week.

How about find the best recorded CD in the genre you're in and compare?

There's peaks and then there's average volume, there's presence, balls... you can have 2 songs that look the same volume and one pops way more.

 

[SouthSIDE Glen]

I just recorded a new song, but I don't know if it's too loud or too quiet. It's mixed to about 0 db. Is this good?

P.S. The song's not finished, so I won't post it anywhere until I'm finished. I should be done this week.

Personally, I think that if the song isn't finished yet, any serious talk of specific mixing volume is somewhat premature.

That said, though, if the song is not finished yet, and you're already peaking out to 0dBFS (I assume that's what you mean by "dB"), that you're running too hot somewhere along the line, because you have no room left to finish the song.

But there's a dozen other questions that are most relevant to the O.P., including what levels are you recording at (that's the BIG one), what kind of setup and technique as far as fader jockeying you're using for the mixing, whether your peaks are the result of final master bus adjustments, and whether you are talking mixing levels or mastering levels (considering you mention both in the thread title, it's kind of confusing.)

G.

 

[masteringhouse]

There are really two levels that you should be concerned with, the peak level and the average or RMS level. It's the difference between these two (look up crest factor) that are the main concern. A crest factor of anywhere between 10 to 20 dBFS is what I would personally shoot for in a mix (depending on genre) with a "cushion" on the peak anwhere from -6 dBFS or lower to help ensure that clipping did not occur and to leave room for mastering. As mentioned peak level has less to do with overall "loudness" than average level. One could smash the daylights out a mix and still only peak at -12.

Don't try to make your mixes as loud as mastered CDs. The decision on final levels should be made once you have completed all of the songs and can listen to them in the context of the album.

 

[Massive Master]

To the average Joe(like me) this seems to say to mix to -12dbfs and up to and "even 20 is fine".

The OP is talking about mixing/mastering and you guys switched to tracking I'm pretty sure . It's like Tom started talking about mixing and ended up talking about tracking and John started talking about tracking and ended up talking about mixing but anyway, I'm sorry it's confusing and I need another coffee

Not at all. I get mixes that peak at -12dBFS (or even -20dBFS) in here all the time.*** Nothing wrong with that. It's 24-bit audio... I'd much rather have a project with "waaaaayyyy too much" headroom than one with not enough.



*** They're usually far superior sounding mixes, done by engineers who are obsessed with ridiculous amounts of headroom at every possible stage.

 

[Waltz Mastering]

To the average Joe(like me) this seems to say to mix to -12dbfs and up to and "even 20 is fine".

The OP is talking about mixing/mastering and you guys switched to tracking I'm pretty sure . It's like Tom started talking about mixing and ended up talking about tracking and John started talking about tracking and ended up talking about mixing but anyway, I'm sorry it's confusing and I need another coffee

I mentioned mixing levels which could also be associated with tracking levels as being ideal with peaking "anywhere" around -12 dbfs and average "anywhere" around -22 dbfs.

John mentioned the numbers 12 and 20 and probably meant to put a minus sign in front of the number (so it would be -12dbfs and -20 db full scale).

The bottom line is that you have tons of headroom in the digital domain below 0 dbfs, so when mixing there is no advantage to try to get your mix as hot as possible. You can save that for mastering (separately)

In the analog domain you sometimes have headroom above 0 dbvu in excess of 20 db.

Generally, if you have your converters calibrated so that -20 DBfs = 0 DBvu
this is a comparison chart of how that would look:

 

[SouthSIDE Glen]

To the average Joe(like me) this seems to say to mix to -12dbfs and up to and "even 20 is fine".

The OP is talking about mixing/mastering and you guys switched to tracking I'm pretty sure . It's like Tom started talking about mixing and ended up talking about tracking and John started talking about tracking and ended up talking about mixing but anyway, I'm sorry it's confusing and I need another coffee

It is all confusing at the start, which is why I kinda feel sorry for the O.P., because all these responses probably seem varied and overwhelming, yet they are mostly all pretty much correct.

NYM, not everybody does it this way, but for me there's typically not a lot of difference between individual tracking volumes and final mix volume. The neat thing is that all this stuff is designed where if you follow good and intended gain structure practices in your mixing technique, when all is said and done, the final mix - before mastering and with a master bus set to unity gain - the final mix levels don't really change all that much.

My MOTU converters are calibrated to 0VU = -18dBFS, and my tracks typically lay down RMSing somewhere near there - a couple of dB higher or lower depending on the actual nature and density of the track - and my typical rock/pop mix will come out to a pre-master mix RMS typically in the -16 to -20dBFS range.

The interesting thing was that 10+ years ago when I was recording to ADAT (calibrated to -15dBFS), I still usually wound up pulling back the tracks a couple of dB extra each when I brought them into the computer to mix, both the push the digital noise floor down and to give myself comfortable headroom for mixing, and the final mix levels still pretty much came out the same as above within a dB or two or so.

This is why it can seem confusing and why I kind of feel for the O.P. because the best answers to his question revolve around not only understanding the difference between RMS and peak, but relating RMS to good gain structure and mixing technique. It's not had at all when you get used to it, but the learning ad getting used to it seems the be the real hard part for most folks.

G.

 

[masteringhouse]

I think that the advent of peak only digital meters in DAWs is one of the reasons for confusion regarding levels. The trick is to find a good digital meter that measures both RMS and peak.

 

[Massive Master]

John mentioned the numbers 12 and 20 and probably meant to put a minus sign in front of the number (so it would be -12dbfs and -20 db full scale).

I was mentioning relative dB of headroom...

 

[masteringhouse]

"Headroom" is another one of those ambiguous terms. Does it mean the margin between the optimal level and the level where distortion is reached in a device, the difference between the peak level and 0dBFS, or the average level of a mix and it's peak level. For mixes I think that we should be talking about the latter.

If you take a hypercompressed mix/master and lower it by 12 dBFS does it sound any better?

 

[squibble94]

Okay, thanks guys. I pretty much got all I need right here. I think the levels could be knocked down a tad. I may have been a little vague about what I needed, sorry. The song isn't finished, but it has every part down except the lead. I'll post it sometime this week.

 

[NYMorningstar]

I guess what confused me was my own habits. In trying to get an optimal signal, I track peaking at -18dbfs(24bit). When I mix digitally, the tracks sum together and I usually set my goal to about -6dbfs for exporting a stereo file to master.
To have my mix hit around the numbers you guys are saying what's optimal? Should I track at lower volumes or keep tracking like I am and pull the individual faders back and/or pull the master fader when mixing? I usually keep the master at unity.

 

[BrentDomann]

Do the answers above apply if you're tracking in 16-bit, or should you push the tracking levels a bit more? It's my understanding that 16-bit was designed to really give the educated user all the headroom he or she would ever need, but we've moved into a 24- to 32- bit tracking world (in the least).

 

[Waltz Mastering]

I was mentioning relative dB of headroom...

I see that now. my bad.

I guess what confused me was my own habits. In trying to get an optimal signal, I track peaking at -18dbfs(24bit). When I mix digitally, the tracks sum together and I usually set my goal to about -6dbfs for exporting a stereo file to master.
To have my mix hit around the numbers you guys are saying what's optimal? Should I track at lower volumes or keep tracking like I am and pull the individual faders back and/or pull the master fader when mixing? I usually keep the master at unity.

I think the numbers you are tracking and mixing are good. There is a lot of leeway and the thing to remember is that there is no need to record "hot" in 24 bit digital system.

It's sort of left over from the 16 bit days when people thought you had to use all the bits.

I guess, how I look at is that most analog gear is designed to work optimally at or around +4 dbu.

So if +4 dbu = -16 dbfs, then your not pushing the analog gear (pre-amp) to it's limits by trying to get close to 0 dbfs as many mistakenly think you have to do in the 24 bit digital world.

If you were shooting to track at levels close to 0 dbfs, your analog gear (pre-amp- compressor) is being stressed at the top of it's range and then in return your daw is also being stressed at the top of it's range once you add more dbs of eq. compression. and summing etc.

As Tom mentioned, understanding the various meanings of the word "headroom" is important to know as well as how it applies to the analog and digital world.

 

[SouthSIDE Glen]

Do the answers above apply if you're tracking in 16-bit, or should you push the tracking levels a bit more? It's my understanding that 16-bit was designed to really give the educated user all the headroom he or she would ever need, but we've moved into a 24- to 32- bit tracking world (in the least).

For quick calculation you can figure you get 6dB of range per bit of word length. That is, roughly, 16 bits gives you 96dB of "digital canvas" on which to paint your music, and 24 bits gives you 144dB of canvas range. (technically, the last bit will probably be noisy, making the useful range 6dB less in both cases, but it's a good rough view to start with.)

On the surface, because 16-bit is a smaller canvas, you'd want to push your levels higher to give yourself maximum range. But when you consider the reality that the analog signal coming in has a total dynamic range between the peak and the analog noise level of maybe 60-65dB if you're quite lucky, the need to push up the levels to "use all the bits" really isn't the critical move it may seem at first blush.

Perhaps if you're feeding something truly highly dynamic and very low noise, like a digital synth or clean, optimized top-shelf samples, you may find an advantage to "pushing the bits", but otherwise, IMHO, I wouldn't lose too much sleep over it.

G.

 

[masteringhouse]

In totally agreement with Tom (this could get confusing).

Operate gear at its nominal level for the best signal to noise ratio. Recording at a peak of -18dBFS seems a bit low to me, but it's only part of the picture anyway. One could be distorting a preamp and just lowering the input to the track so that the peak sits lower. It doesn't mean that it was recorded cleanly. I would much rather have a clean signal from the preamp recorded at -1 dBFS (unless you are going for distortion).

Keep an eye on the meters but don't let them rule your life in either extreme (too low or too high). Make your main concern the quality of the perfomance and sound of the track.