-7 db rms without clipping

[KillFrenzy]

Scorpions' "Humanity: Hour 1" hits -5.7dB average for some songs. And how I love that album.
I've been using Sonnox Limiter and reaching great loudness with no clipping - something around -7dB. Do you guys use it?
I think the mix, tones, etc, play a HUGE part on loudness. Not only in numbers, but in the 'hugeness' of the sound - again, latest Scorpions album.

EDIT: I may be deaf or my monitoring is bad (HS80M + ATH-M50), but I don't hear clipping (at least bad clipping) in the cases I mentioned (my "masters" and Scorpions')

 

[Plec]

Every mix has its own loudness limit, and that limit varies a lot depending on how the song was mixed and produced. There are hundreds of things happening on a single track that determine how much it can take in mastering, loudness wise.

Some mixes can start to sound strained at -12 even, others reach -8 without a problem. Arrangements, tempos, frequency content.... everything affects how hard you can push it even in different parts of a single song and different songs on an album will be able to take more beating than others. There are x amounts of techniques to master in terms of getting loudness, but when learned they pretty much only offer different "tastes" of loudness and not that one technique would just be "better" than another one. So it's important to select a "taste" that fits the mix and production. If you think you've found "the" way to get good loudness.... think again. It always changes from project to project.

The thing that still makes the biggest difference is the mix, the production and the song.

-7 is a tough benchmark though!
Great job Joey!

 

[Ben Johnson]

great post, Plec

 

[KillFrenzy]

Yeah. Plec pretty much explained my brief thought.

 

[waltz mastering]

Every mix has its own loudness limit,

Some mixes can start to sound strained at -12 even, others reach -8 without a problem.

This only true if everybody mastered the same exact way, using the same technique.

The perceived loudness of the same track at the same rms, mastered by two different ME's using different techniques could vary greatly.

 

[Plec]

This only true if everybody mastered the same exact way, using the same technique.

The perceived loudness of the same track at the same rms, mastered by two different ME's using different techniques could vary greatly.

I don't actually know if you're disagreeing or not?

Perceived loudness differing at the same RMS pretty much sums up to different choices of EQ and/or saturation in the midrange and very little else, but a mix being pushed hard seldom takes large EQ changes gracefully so for things to "vary greatly" at the same RMS the mastering engineers must be diggin' in quite deep with EQ or saturation in various ways which would alter the mix in a quite profound way and that usually isn't what's preferred IMO. Other techniques than actually changing the frequency content might affect the perceived loudness a very small amount, but it won't make say a given -10 master seem like a -8.

 

[waltz mastering]

I don't actually know if you're disagreeing or not?

Both... Saying a track has it's own loudness potential might be partially true, but from ME to ME the loudness potential would change depending on how they go about achieving it.

Perceived loudness differing at the same RMS pretty much sums up to different choices of EQ and/or saturation in the midrange and very little else,

It really depends on the track. Imo/e mid range is a big part of it, but low and high end can't be discounted.

It really depends where the energy is, and where it's needed and most importantly how it's achieved and distributed.

but a mix being pushed hard seldom takes large EQ changes gracefully

Agree.

so for things to "vary greatly" at the same RMS the mastering engineers must be diggin' in quite deep with EQ or saturation in various ways which would alter the mix in a quite profound way and that usually isn't what's preferred IMO. Other techniques than actually changing the frequency content might affect the perceived loudness a very small amount, but it won't make say a given -10 master seem like a -8.

I do not think rms numbers tell a whole lot about the actual perceived loudness of a track.

How someone goes about getting level through gain staging, eq distribution, compression and limiting has a big effect on the perceived loudness, even though the end number might be the same.

In other words, given the same exact track, one ME might be able to get that track to a average rms level of -10 dBFS before the sound starts to go south, where another ME using different techniques might be able to get it to -11 dBFS before it starts to go south, yet depending on how the energy is distributed, the track at -11 could sound louder than the track at -10.

As another example, say you have a mix and simply pull the threshold down on a bw limiter until the rms reads -10 dBFS. Take the same mix that someone like Ted Jenson mastered and is reading the same exact rms. They would sound totally different in perceived level, even though they read the same rms.

One of the tracks might sound like ass at -10, where the other might retain/gain punch and clarity at -10. The perceived loudness would be different.

I think picking a number and saying a track is "that" loud actually tells little about how loud the track actually is...

 

[El_Gato]

I'm not a ME but I noticed also the speed is a huge factor to take into account (?¿)

A 230bpm track needs less rms than a slow one. I perceive the faster a lot louder, so they need less rms in general. A -11rms fast track can be louder than a -10rms slow one.

 

[Ermz]

Ozone + a good mix makes this kind of thing entirely achievable. The loudness maximizer portion is an immensely powerful tool, easily abused and such. I like to mix with a very aggressive rough mastering chain in place, so that I am forced to dynamically control and balance my mixes in a way that would avoid clipping in a final master. I never meter RMS levels. Just listen, compare and gauge what feels right for the project.

 

[ttrentt]

I like to mix with a very aggressive rough mastering chain in place, so that I am forced to dynamically control and balance my mixes in a way that would avoid clipping in a final master.

I might have to try this, but ozone is a huge hog. I am pushing 75% already...are you running something like L2 just to slam it and "work" with the dynamics?

 

[Ermz]

Nah, I run Ozone. The sessions are usually at around 90% these days. Not overblowing the CPU usage is a real balancing act.

 

[colonel kurtz]

I'm not a ME but I noticed also the speed is a huge factor to take into account (?¿)

A 230bpm track needs less rms than a slow one. I perceive the faster a lot louder, so they need less rms in general. A -11rms fast track can be louder than a -10rms slow one.

in addition to that, cutting out low end can definitely increase the levels you can squeeze out of a mix...but this is a fine, line, of course, because you still want the mix the bring the BOOM

 

[darthjujuu]

late to this party... was OP album "we came as romans"?

 

[Sinister Mephisto]

I have no idea, but the last we came as romans album is SUPER loud. I was crying trying to match it.

 

[Plec]

Both... Saying a track has it's own loudness potential might be partially true, but from ME to ME the loudness potential would change depending on how they go about achieving it.

It really depends on the track. Imo/e mid range is a big part of it, but low and high end can't be discounted.

It really depends where the energy is, and where it's needed and most importantly how it's achieved and distributed.

I do not think rms numbers tell a whole lot about the actual perceived loudness of a track.

How someone goes about getting level through gain staging, eq distribution, compression and limiting has a big effect on the perceived loudness, even though the end number might be the same.

In other words, given the same exact track, one ME might be able to get that track to a average rms level of -10 dBFS before the sound starts to go south, where another ME using different techniques might be able to get it to -11 dBFS before it starts to go south, yet depending on how the energy is distributed, the track at -11 could sound louder than the track at -10.

As another example, say you have a mix and simply pull the threshold down on a bw limiter until the rms reads -10 dBFS. Take the same mix that someone like Ted Jenson mastered and is reading the same exact rms. They would sound totally different in perceived level, even though they read the same rms.

One of the tracks might sound like ass at -10, where the other might retain/gain punch and clarity at -10. The perceived loudness would be different.

I think picking a number and saying a track is "that" loud actually tells little about how loud the track actually is...

I agree with your points.

To have any idea about what "loudness" you might expect from an RMS value you need to know what type of genre/sub-genre and type of song you're dealing with and that will only give you a rough idea.

A slow, sparse country ballad at around -8 will come across as a lot louder than a thrash metal, 170BPM reading the same RMS value.