Beginner question on loudness during mastering

[H-evolve]

Hi,

I've never quite understood how to precisely get the loudness when mastering a song. Let me explain my process and you'll probably get what I do wrong, miss, or misunderstand:

- I mix my songs so that everything sounds as good as I can, with the "relative" levels between instruments being OK. When looking at the Master Fadder in Reaper I'm usually in the range of -10 to -5dB, it really depends.

- On some songs, maybe the guitars will be recorded louder or something, so when adjusting my mix, I don't really try to achieve the same "Master Level Fadder" on every song. I just try to make it work and sound OK. I don't really look at the Master Fadder all that much.

- Then, when at the mastering step, I usually start with some EQ'ing and Multiband compression, and then I use 3 bus compressors in series. For example, my first one will only tame the big peaks, and the other ones will just slightly even out everything. I usually boost the output of the last 2 bus compressors, to get some volume/loudness.

- At that stage I'm usually still not quite at 0dB on the master fadder, and this is the part where I'm confused. Even if I crank my bus compressors enough to be flirting with 0dB... when I compare to reference songs, they are still louder..... So, what am I suppose to do? At the moment I simply use a L1 Limiter, I set the limit to -0.1dB and I lower the threshold until I find that the volume is loud enough. But this seems wrong. And to top it all, this makes matching the loudness between each song quite complicated. It's really an iterative process that isn't really precise... :S

I'm sure most of the stuff I do isn't alll that legit. I'm no pro after all, but that last step with the limiter is the one where I just don't know how to do it. I just lack education in that department. Hence my kind request for help.

Thanks in advance.

 

[newamerikangospel]

I am far from a mastering expert, but I found my mixes sound better the less mastering stuff I do. My master has satson bus on it, and "The Glue", that I mix into. When I am happy with the mix, I put Ozone on it, and split mid/side. On mid, I do a small bump where the kick sits (generally around 65-75hz) and then a small cut at its octave (130 if I bump 65, for example). On side, I find where the lowend blurs the bottom end, then do a shelf until it clears (generally 2-3db, but entirely song dependent). I do a stereo widening on everything above the point where I did the cut, and then mildly narrow everything below the cut (130, following the example above). I put the maximizer on, set it very relaxed and do a volume push until it starts to hit 0db, but not limiting yet. Then I put kclip on, find i where the point where it clips but still sounds good but saturates. My rms is generally around 10 to 8, so still hot but not winning any loudness wars.

 

[newamerikangospel]

I generally only make changes to guitar levels or snare levels once I have the mastering chain up. Those seem to be the two things that either overpower or get buried.

 

[KillFrenzy]

It's almost impossible to achieve that "commercial" loudness without a limiter/clipper. What you do with your L1 is pretty much the standard procedure. I match the loudness of the songs by ear. Don't be afraid to do that!

 

[Machinated]

Loudness very much comes from the production and mix - over the years i’ve found it much easier to get a loud final product just from getting a better frequency balance and dynamic control of things. I’ll rarely add more than about 2-4dB i’d gain from a mastering limiter at the end and even without it, things will be at a decent level.

I’d say you’re better off trying to get the dynamics as close to perfect as you can while mixing as it’ll be very hard to do a lot of work at the mastering phase. Volume is determined by a number of factors so it makes sense to hear towards it from the outset....

 

[H-evolve]

Thanks all for your answers.

I think, one concept that I can't grasp yet, is why, when the master signal level is already close to 0 during the mix... why isn't it loud enough? For instance, on the limiter, when "limiting" everything to -0.1 dB... why does reducing the threshold keeps increasing volume, even though the master signal is limited to the same value of -0.1 dB?

As a poor attempt at understanding this, I'm thinking it comes from the fact that as I reduce the threshold value, the limiter takes "everything" above that threshold and boosts it to -0.1dB. The lower the threshold, the more stuff is clogged at -0.1dB, hence making it sound louder. Is my understanding correct?

But then, who decided that "this volume" is the correct "commercial loudness". I mean, the only way I know my stuff is loud enough is by using a commercial reference. However, is there a numerical value somewhere of loudness that is universally accepted as "the standard required loudness for commercial music"?

I mean, you can hear fluctuations from song to song, when listening to say, Spotify. But they are usually pretty close from one another, making me believe at some point, somebody decided that "This is the standard loudness volume required"! XD

 

[newamerikangospel]

The same way you can squash a drum room and it sound louder even if it is 6db quieter than the uncompressed mic. It seems like the human ear is not designed to recognoze loud only as spl, but also as how loud something would be to make that sound. Like how a compressed snare drum makes the drummer sound like he is hitting harder.

 

[KillFrenzy]

As a poor attempt at understanding this, I'm thinking it comes from the fact that as I reduce the threshold value, the limiter takes "everything" above that threshold and boosts it to -0.1dB. The lower the threshold, the more stuff is clogged at -0.1dB, hence making it sound louder. Is my understanding correct?

It's pretty much that. Some limiters do a better job that others on increasing loudness without destroying transients and audible clipping.

That and what Machinated said: a loud-but-not-ruined master comes from a loud mix - some well done compression and a great frequency balance in a way that you fill the spectrum

But then, who decided that "this volume" is the correct "commercial loudness". I mean, the only way I know my stuff is loud enough is by using a commercial reference. However, is there a numerical value somewhere of loudness that is universally accepted as "the standard required loudness for commercial music"?

You can't put a definitive number on that simply because every music is different. A loud and busy chorus is louder than a smaller verse.
Another thing is that perceived loudness is affected by frequency balance. If you take 1kHz sine wave and a white noise with the same RMS level, the noise will sound louder.

 

[Megin]

Loudness very much comes from the production and mix - over the years i’ve found it much easier to get a loud final product just from getting a better frequency balance and dynamic control of things. I’ll rarely add more than about 2-4dB i’d gain from a mastering limiter at the end and even without it, things will be at a decent level.

I’d say you’re better off trying to get the dynamics as close to perfect as you can while mixing as it’ll be very hard to do a lot of work at the mastering phase. Volume is determined by a number of factors so it makes sense to hear towards it from the outset....

I second that

 

[Trodden]

I've never quite understood how to precisely get the loudness when mastering a song

you need more cowbell

 

[Kaza]

In your mix session, make sure your mic/instrument tracks do NOT go past -6.
And do not put ANY plugins on your master fader. Watch your RMS too.

Bounce to stereo WAV and start a new session and put that WAV file on a new track.

Add some very soft compression, a bit of reverb just to give it more air and then add your multiband processors, minor EQ, limiting and stereo enhancement.