Normalizing audio waveforms?

  • Thread starter Thread starter jndietz
  • Start date Start date
J

jndietz

The Way It Moves
Should I normalize all of my waveforms to a certain volume before I sent them into the mixing process?
 
This subject has been discussed thoroughly a few times, so use the search function. :)
 
Not sure if it's been mentioned. But, no, you should not. :)
 
nah thats how i messed up many of my own projects.... i would silence all the dead space, and then normalize b4 incorperating any mixing amongst tracks following after one of these "reject guides" someone was posting about making enhanced vocals via CEP/AA

it was almost like using a fucked up compressor

i didn't realize how retarded my mixes was sounding til i stopped following that format and did more of a individualized trial and error process.

i still don't know what "normalizing" theorhetically does, but i know since i stopped using it, the clarity of my mixes improved from sounding like i recorded in a cave to stadium status....(yeah right, i'm still learning)
 
Sinistah said:
...i still don't know what "normalizing" theorhetically does...
Click to expand...
It's nothing more than raising the overall level of the file until its highest peak reaches a specified point, often 0 dBFS or just slightly under.
 
  • Like
Reactions: XLR
XLR said:
It's nothing more than raising the overall level of the file until its highest peak reaches a specified point, often 0 dBFS or just slightly under.
Click to expand...

cool, so now i see why it would be done after a mixdown instead of before hand, and also why it sounded like bootleg compressor, it was because i was following the steps of an idiot........
 
Sinistah said:
i still don't know what "normalizing" theorhetically does,
Click to expand...
There's two types of "normalization". The first one, called "peak normalization" is the kind that XLR described. It's really nothing more than a volume level conrtol that sets the volume not by amount of gain like a normal volume control, but rather by seeting the highest volume level for the loudest peak in the wave.

The second kind is called "RMS normalization". This is also a fancy volume control, but rather than setting the volume by peak volume, it's setting it by the average (RMS) volume level. These normalizers usually include brick wall compression at the top of the scale to avoid clipping.

Neither one should be used before mixing because there is no reason why every track should be at the same volume levels - either peak or average. If that were the case, there'd be no need for mixing, would there? :p And raising the volume without reason is just taking away headroom that could be better used downstream.

G.
 
Glen summed it up perfectly.

Also, in digital recording, you want to keep the processing to a minimum.
 
Naaa, don't listen to any of these guys. You should normalize, then compress, then limit your mix before sending it out to the ME. Make sure to limit your mix so that when you load the exported wave, it looks like a solid black square brick. Don't forget to add some multey-band compressin and then truncate the file down to 16bits.

That's how all the modern pop mixes are done.

:D ;) :p
 
noisewreck said:
That's how all the modern pop mixes are done.

:D ;) :p
Click to expand...
:D I don't think even those boneheads do that before the mix.

Ssshhhhh...don't give them any ideas! :rolleyes: Otherwise we'll have to unleash a Kurzweil beast on you....

G.
 
SouthSIDE Glen said:
:D I don't think even those boneheads do that before the mix.
Click to expand...
Duh... Somehow I missed that part... Still, even then I think I am only slightly exaggerating... I know that limiting the drum submix at least is becoming increasingly routine :mad:

SouthSIDE Glen said:
Ssshhhhh...don't give them any ideas! :rolleyes: Otherwise we'll have to unleash a Kurzweil beast on you....
Click to expand...
Eh, that's not a threat :D It's more like a prayer... Oh man... the more I read about it, the more excited I get. Talk about pure synthesis nirvana!
 
SouthSIDE Glen said:
There's two types of "normalization". The first one, called "peak normalization" is the kind that XLR described. It's really nothing more than a volume level conrtol that sets the volume not by amount of gain like a normal volume control, but rather by seeting the highest volume level for the loudest peak in the wave.

The second kind is called "RMS normalization". This is also a fancy volume control, but rather than setting the volume by peak volume, it's setting it by the average (RMS) volume level. These normalizers usually include brick wall compression at the top of the scale to avoid clipping.

Neither one should be used before mixing because there is no reason why every track should be at the same volume levels - either peak or average. If that were the case, there'd be no need for mixing, would there? :p And raising the volume without reason is just taking away headroom that could be better used downstream.

G.
Click to expand...
Jackpot..... perfectly understood bro, thanks, so with the first one it sets the volume via the loudest peak of the track to i'll assume make sure it doesn't rise above that high-point

and the second one is a culmulation of the whole track down to an average peak level and then compressing it all to make sure it doesn't go past the designated average.........

so i was kinda right with my first statement referring to the idiot-steps of the normalization making it sound like a shitty compressor was used....

i'm taking babysteps to a better mixdown each time i log on...thanks guys!
 
noisewreck said:
I know that limiting the drum submix at least is becoming increasingly routine :mad:
Click to expand...
Good point. :)
noisewreck said:
Eh, that's not a threat :D It's more like a prayer... Oh man... the more I read about it, the more excited I get. Talk about pure synthesis nirvana!
Click to expand...
Kinda early to be tugging on Santa's pantleg, isn't it? Just increases the amount of time you have to be a good boy :p
sinistah said:
i'm taking babysteps to a better mixdown each time i log on...thanks guys!
Click to expand...
Keep up the good work and keep trying new stuff :).

G.
 
SouthSIDE Glen said:
Kinda early to be tugging on Santa's pantleg, isn't it? Just increases the amount of time you have to be a good boy :p
Click to expand...
Oh man, Santa's real afterall? And here I'm trying to work the woman. :rolleyes: Barking up the wrong tree I guess :eek: :D
 
noisewreck said:
Oh man, Santa's real afterall? And here I'm trying to work the woman. :rolleyes: Barking up the wrong tree I guess :eek: :D
Click to expand...
You could do what one of my buddies did this summer...

He purposely dropped and damaged one of his old keyboards so that his gal would let hin get the new one. A dangerous gambit, perhaps, and he'll probably burn in hell for it, but he got the new keys ;).

Actually I think he's already burning in hell for it. Now that I think about it, they were just living together when he started the gambit. By the time he got the new keys he was already on his way to a 115° Las Vegas to get married. (You can have your keyboard if you finally marry me?) I don't think he's had sex since....

All in all, I think Santa is the cheaper bet. You may not get your new toy, but at least you'll get to keep your jewels and save on the tanning lotion.

G.
 
SouthSIDE Glen said:
..."RMS normalization". This is also a fancy volume control, but rather than setting the volume by peak volume, it's setting it by the average (RMS) volume level. These normalizers usually include brick wall compression at the top of the scale to avoid clipping.
Click to expand...
To each his own re terminology, as audio is so fragmented that way.:) I don't consider matching average levels to be normalization... even if there are processors now that do it (after a fashion anyway) and call it that.:p
 
You must log in or register to reply here.
Back
Top