bitflipper

-0.2DB is now the worldwide mastering standard...

Can you cite an authoritative reference to support that statement?
 
AFAIK, there is no standards organization for mastering, but -0.2 dB could still be a de facto standard by consensus. I've only checked a small handful of recent releases personally, so I wouldn't know.
 
However, I do have a great many CDs that peak well under -1.0 dB, so if this is indeed the worldwide standard then it must be a recent development. Or was your statement perhaps aimed at a specific musical genre or market segment?

I was wondering the same thing, bit, or thought maybe it was "aimed" at me just because. In my experience though, even if there were some sort of "etched in stone" value as stated in the "constitution of mastering" handbook, I'd still say use what works for a particular project regardless of what setting it ends up being. For me, nothing hotter than -0.1, nothing lower than -0.3 dB peak is what I prefer for most projects. However, there are times when I'll close my eyes and set something to what it sounds like. Wherever it ends up, it ends up. It depends on the genre, how the song was recorded, mixed/handled and of course what limiter I use.
 
I have an album here that David Rosenthal was involved with in the 90's. It's by far the loudest album I have ever heard from start to finish. It's even louder than Metallica's St. Anger. For some reason, this album that David did sounds great even though it's insanely loud. I wouldn't have even attempted to make such a fine album so loud, but I must say....it's definitely a loud one done right. It even shows clip points yet doesn't SOUND clipped. Go figure..
 
-Danny

It strikes me as silly that a difference of 0.1 dB will make an audible difference, in any context.
 
For that matter, 1.0 dB usually won't make a noticeable difference when you're talking peak values in reasonably dynamic music. Try it yourself: make two mixes of the same song, normalized to the same RMS but with one with peaks limited at -1.0 dB and the other at -0.1 dB and see if you can distinguish the two in a blind test. 
 
You may or may not be able to; it depends on how dynamic the mix was to begin with. If you started with a dynamic range of 4 dB, then that single decibel represents 25% of your range. But if you have a crest factor of, say, 14 dB, it's far less likely you'll notice if peaks are increased or decreased by 1.
 
Perhaps all this will eventually become moot when R128 is universally adopted. I know I'd have to remaster most of my own stuff to meet the new broadcast requirements, and my stuff is far from squashed, at least by current conventions. (Want to have your eyes opened? Get an EBU loudness meter, check your material and see how much of it would be flat-out rejected for European broadcast.)

bitflipper
Get an EBU loudness meter, check your material and see how much of it would be flat-out rejected for European broadcast.)

Why would they reject it?

Much of it is too loud.

I am told that sound for television is particularly picky about EBU levels, and have heard stories about people's submissions being kicked back that were just close to the limit but did not actually exceed them.

I work in radio broadcast full time.  Usually we expect varying degrees of loudness in music submissions and is my job in the production department to bring the loudness to spec.  Too loud and it will sound just awful on the air after further pre-transmitter compression,  too soft and the audience gets upset having to turn up their dials (gasp!).  At any rate, it is good to know the target.  But don't worry too much if your music isn't at the correct specs for broadcast, I still need to earn a paycheck.  

rumleymusic
I work in radio broadcast full time.  Usually we expect varying degrees of loudness in music submissions and is my job in the production department to bring the loudness to spec.  Too loud and it will sound just awful on the air after further pre-transmitter compression,  too soft and the audience gets upset having to turn up their dials (gasp!).  At any rate, it is good to know the target.  But don't worry too much if your music isn't at the correct specs for broadcast, I still need to earn a paycheck.  

Daniel, just curious....can you tell me what "spec" actually is for where you are? And, is it different for each place that has a guy like you taking care of that stuff? For example, does WXYZ have a different spec than WVFN?
 
Next, one of my friends was the creative services program director in Philly at the old Y-100 station. Basically, he did all the voice work, all the commercials, little jingles, all the weird sounds like stations changing etc and of course, he had to bring things "to spec". This was a long time ago in the lat 90's to mid 2000's....but if you were doing what you do now during those times, has this "spec" changed from then to today, and if so....how much of a difference is there? Thanks in advance.
 
In my experience when getting things ready for radio for my clients, I do my best not to squash anything because I know the station will do quite a bit of that. I have a few limiter settings that simulate what *could* go on at a station so I sort of use those as my guides. Of course none of them are probably correct, but it shows me how something too squashed on my end could sound like absolute dog crap on a radio station if I hammer the limiter too much.
 
That said, with the stuff I do at my studio, I can get away with hammering if I needed to and the end results isn't bad at all. It depends who records the material in my opinion. A guy that may not have a grasp on how to really record or mix something is not going to get the same results as a guy that has a clue. That reminds me of a good quote I came up with to a student the other day...lol!
 
"The more we try to process a turd, the more it tries to find its way back to the sewer"
 
I find the same to be true trying to master something that isn't recorded very well from the start. Or you get a client that says "it has to be louder...no, louder still..no, it's still not loud enough!" What they fail to realize is, you don't just grab something and master it loud. It has to be MIXED to be mastered loud. It has to be recorded in a good way (notice I said "good" and not great) for the loudness to come across the right way. But people don't get it. Anyway...sorry, I drifted a bit there.
 
Like I mentioned before in my post, whatever sounds the best to me that is not showing clip points or allowing me to HEAR clip points is what I use on a project. If I smash a limiter to -7 dB and it works for that material, so be it. I try not to, but anyone that knows anything about this field knows that you can end up with some really freaky things going on at the end of the day. Stuff you wouldn't normally be down with, ya know? Whatever works, makes the client sound great and makes you not ashamed to have your name on the project as the engineer is what is best in MY opinion. :)
 
-Danny

In the USA, radio is still mostly analog transmissions so when it's all said and done the FCC has the final word on that. They don't want your licensed broadcasting stepping on other entities licensed broadcasts.
 
In the USA, TV is now mostly digital transmission so, for the most part, the loudness levels are regulated in house to set some perceived standard of quality that the broadcaster wants to be known for.
 
best regards,
mike

The ITU-R BS 1770-2 measurement standards are what most stations aim for.   Usually around -14dB RMS before it goes out, but that depends on the settings of the broadcast audio processor.  (in my station it is closer to -20dB RMS since we are classical and need as much dynamics as possible)  This is mainly for dialogue, but the music in radio needs to be matched dynamically as well.  I doubt most stations have more than 5-6dB of dynamic range by the time it reaches your car.  Pre-processor digital and online streaming make the need for in house standards imperative as Mike said.  Everything goes in the computer now of course, the days of tape, cards, and even CD's are mostly over, which makes control over levels a much easier task.   
 
Many popular music stations though will not bother with levels on their music, much of which is taken directly from online content sources directly to their automation systems.  They can rely on their preset processors to squash the audio down to an acceptable level in comparison to the talking, even before online streaming or HD transmitters.  If you want to submit music that will sound good on radio, make sure you adhere to the -14dB RMS target, or it could be just a muddy mess and sound nothing like your mix once it reaches your listeners.  My advice anyway.
 
 

Anothe reason why the K system approach to production is just so good. -14 dB rms is one of the ref levels and so is -20 dB rms  with -12 dB rms being the third ref level.  -14 is such is a nice level. It is reasonably loud but still retains quite a lot of dynamics and transients. (For a lot of music genres and I agree -20 is nice for classical)
 
The great thing about -14 is that you can easily reach it without any mastering (loudness wars that is) required. Of course you an still apply mastering processes such as EQ and compression but the limiter is not needed under those conditions.
 
All the more reason as well to get more VU meters into your DAW's and start using them.