Loudness level

[mairomaster]

There are lots of topics for mastering in the forum soon, but I think my question is important.
People always talk about commercial levels of lets say -9 dB RMS. Do they usually mean the average loudness of the whole song or the average loudness of the loudest moments (choruses for example) or what? I've tried to find some information on the web but I didn't have much of success.
I know it's always best to use your ears, but I'm pretty new to mastering and prefer to use some reference measurement tools for better results. I use Izotope Ozone 5 Advanced. Do you think its loudness meters are good? What tools do you use for measuring the loudness?

 

[Terminus]

Average loudness of whole song. Yes to the Ozone thing, at least from what I've demo'ed and heard about it's one of the better plugs of it's type, though I don't own it and ftmp people tend to only use parts of it and not the whole deal together ftmp. I use the Brainworx TT dynamic range meter for my loudness metering.

 

[schwinginbatman]

Yeah, because of how many variables can affect the loudness of my stuff, from the loudness of the volume on the interface, the volume on Windows, etc. etc. I use reference materials to judge the loudness. I aim for an average of -7 RMS for metal, and it typically peaks somewhere in the -6 to -5 range. Mastering kicks my ass, haha.

 

[waltz mastering]

People always talk about commercial levels of lets say -9 dB RMS. Do they usually mean the average loudness of the whole song or the average loudness of the loudest moments (choruses for example) or what?

Ime, when I see people give an rms value for their loudness it is for the top rms reading of the song and not the average rms throughout.

To add to the confusion some digital meters read the AES- 17 standard which is 3 dB difference from convention digital metering.

If you're going to use metering for loudness, It's better to go with V.U. or Dorrough meters because of the ballistics involved and how they react to the music...a bit more revealing and helpful, but still not as good as your ears.

 

[Jarkko Mattheiszen]

Ime, when I see people give an rms value for their loudness it is for the top rms reading of the song and not the average rms throughout.

To add to the confusion some digital meters read the AES- 17 standard which is 3 dB difference from convention digital metering.

If you're going to use metering for loudness, It's better to go with V.U. or Dorrough meters because of the ballistics involved and how they react to the music...a bit more revealing and helpful, but still not as good as your ears.

Pretty much everything he said. If you take the average RMS for the whole song, you might get "normal" readings due to the more silent passages in the song (intro, outro, etc.) even though the louder parts are completely smashed. Things get even worse if you compare song A with several rather silent parts to a desired, commercial song B that's full-blown blasting all the way and try to match the numbers.

These are part of the reason why I dislike mastering music to a given RMS guideline unless it's clearly specified and benefits the music. "Make it sound like album X" is a better reference than "the average RMS has to match song X."

 

[mairomaster]

Thanks for the replies guys! I haven't been around for a while but I'm back now. As most of you are saying, it sounds logical to me to measure the loudest part of the song and state that value for the loudness RMS. Because if you can make the loudest part sounds good the rest of the song will be OK (dynamic wise). I think the biggest problem is that there isn't a single standard method for measuring the loudness. These days I was working on the mastering of one of my mixes and I got really confused. I used 3 different tools for the rms - ozone's rms meter, ozones BS.1770 meter and Span rms meter. With all the tools I was measuring the rms of a short period of the song. And the results I've got were absurdly different - about 6 dB. It was something like:
Ozone BS.1770 -6dB
Ozone RMS -7dB
Span RMS -12dB

I was also using songs that I know well for reference, but the song I was mastering was pretty strange and different compared to them and the loudness comparison was difficult.
Any suggestions?

 

[BLUElightCory]

-7 dB RMS? Seriously guys?

 

[mairomaster]

It wasn't by far as loud as it sounds (if that's what you mean). Compared to different modern songs (ear wise) and not so modern ones it was quieter (maybe with a few dB).

 

[CubanDude]

here we go again.. perceived loudness vs "actual" loudness... i'm quite sure your tracks will bei clipping at -7 dB RMS.. many of those commercial clips you're referring to aren't near that level, they just sound like they would

 

[schwinginbatman]

If you do it right, it shouldn't clip at -7. I've gotten things up to -6 and peaking at -5 without any noticeable clipping. But the -7 I'm referring to may not be the -7 you're referring to.

 

[mairomaster]

I checked the levels again, more carefully this time. For the loudest part, Ozone RMS and Ozone BS.1770 shows almost the same: -7dB. The Span gives -10dB.
I found why is the difference - they are just calibrated differently. Ozone is calibrated to give 0db RMS with sine wave that peaks at 0dB. Span by default shows -3dB for a sine wave that peaks at 0dB, although you can change the option from the menu. Maybe that's the difference that waltz_mastering mentioned.
According to Bob Katz, the first way of calibrating is the right one, mainly because it's been used for 70 years. However from purely scientific point of view, the second one is more correct, because the real RMS of a sine wave is smaller than the peak value.
So obviously everybody is using different meters and it's hard to speak about the loudness of a song without using the same meter...

 

[mairomaster]

That's the song by the way (the link in the last post):
http://www.ultimatemetal.com/forum/rate-my-mix-tone-threads/747116-first-mix-i-post-here.html

You are welcome to drop a comment in that topic, too.