IOW, isn't normalization just a shortcut to setting a volume peak manually? 

As far as my limited knowledge goes, then yes, this is exactly what normalizing does. It rasies the WHOLE of the waveform up to a predetermined level

I've never really understood why a ME is put on such a high pedestal if they are initially handed a great mix

Indeed, a great mix will require no 'mastering' other than matching it in level to other songs on the same project/album.

If you use Wave Lab, or any software like it,,, you will use Normalizing all the time... They have very clear instructions in the help files about using it and HOW and WHEN to apply it. 

Sonar's normalizing function is terribly implemented if you compare it to the same process in Wave Lab. So just to qualify-- Yes,, don't use Sonar's Normalization , you are working blind.  But if you do your own Mastering and wave editing with a Wave Lab ( or other like it) it normalizing is a very important tool. 

I just finished a 2 hour live recording of a Choral Concert. 32 songs. They just want the recordings for there own critique. I don't have to Eq the tracks as that was done at input stage. But between using the normalize " get peak level" and global analyse  " Loudness" 
I quickly produce a CD that is well balanced, lot's of dynamic range and as loud as it can be without needing any compression. There is a little hand work done to reduce stray spikes that happen when say someone sneezes.. If you learn how to work in this environment it's all in a days work and normalizing is the most important tool I have to get the job done. 

 I set it to - 1 Db.  this puts the RMS average at around -20Db if it was Pop music I strive for RMS average of -14.  I would imagine head banger stuff would average -5 RMS

Depends on what medium and end result the Master is for. Some of the best mixes I've ever heard would never meet radio standards or sound big enough to compete with louder Masters, which is tragic for the industry. But this gets very subjective. In a perfect world, when a mix is done, that's it--it's done, and the only thing an ME needs to do is merely encode it or transfer it to its destination medium. Classical music should be a great example of this. Hard Rock? Never happens anymore. That was the '70's.

Fortunately, we've reached the top of what is possible, and a real example of carnage left by a bad mastering job was Rush's Vapor Trails (the first pressings). You can audibly hear clipping on many CD players except the newest hardware. It's horrible and fatiguing on the ears. So, we know where our limits are, and what people don't like when it goes too far.

So, I'm happy to say that I want good, loud masters, but I no longer feel the need to keep hunting for the loudest ME I can find. 

Bob Ludwig raved to me about his work on the last GnR album, in that he was asked NOT to smash the mix to oblivion for loudness. But in reality, it likely was compressed a little, or limited to get it up to a level where people would still think it was a rocking CD. It's a good Master, for sure. 

There's very few (many none) mixes left that really don't need a final touch from a Mastering Engineer. To do that really entails the Mixing Engineer is engaged in modern Mastering practices, most likely. That's not a terrible thing, but it's happening despite what the recommendations are. 

 

I don't get why one leaves gain headroom for mastering.

 


I do it because it's what all the cool kids are doing.


HTH 
Mike

Normalization is a useful tool, not a cure all.

I've just been transferring an LP (actually 2) of Jesus Christ Superstar to mp3s.  Even flat out I was getting a weak signal into Sound Forge.  I used the normalize function to -16 dB (the preset) and it worked fine to bring up the vol. to something respectable.  Did it for each side file and then edited it into the songs I wanted to keep.

Worked fine.

@

I highly recommend a book by a guy who calls himself "Mixerman", Zen and the Art of Mixing.

Thanks Brian, I took your advice and ordered it yesterday.