Normalizing tracks (or not)

[boomer]

It sure seem like it now takes much longer to normalize audio tracks using R9 and Mac Sierra than it did with R8 and earlier OSX.

So my question is, other than the obvious small differences in level and assuming no overloads, what technical differences with regard to sonic quality are there between normalizing a track and simply raising its level?

[selig]

It sure seem like it now takes much longer to normalize audio tracks using R9 and Mac Sierra than it did with R8 and earlier OSX.

So my question is, other than the obvious small differences in level and assuming no overloads, what technical differences with regard to sonic quality are there between normalizing a track and simply raising its level?

No difference, except for that you probably don't ever want all audio tracks normalize to max level. The reason is that when you mix them together you are literally "adding" the levels. So adding two tracks that peak at 0 dBSF (normalized) will possibly clip the output if they both peak at the same time. Add 10-20 more of these tracks and you will certainly clip.

So if you DO normalize tracks, I suggest the very next thing you do is reduce their clip gain (select all clips, type in a value) by 12 dB or so to leave headroom for mixing.

My philosophy is to have ALL audio signals peaking around -12 dBFS, which leaves enough headroom to avoid clipping in most all cases, and means you don't have to lower all the faders when mixing (or lower the master). There are other advantages, but this is the main one IMO.

[submonsterz]

- there is no clipping in the digital domain

- normalize just adds gain until the loudest peak hits zero.. same with raising the volume

But although that is true .
You monitor from the master if as selig says which is correct normsalizing tracks ie more than one being added together will more than likely clip at out put and sorry but you are then clipping .
Just giving a no clipping in digital domain can to some people mean they can carry on regardless and sorry but that is a bad way to portray the info to especially an audiance that may have begginners that do not understand how clipping at out put can damage thier kit along with totally screwing thier mix in the process .
Maybe add in the where it does not matter in your answer to where it becomes a problem and can cause all dorts of destruction and confusion. Selig at least explained it as it is .

[submonsterz]

Lol post i responded to dissapeared .....wtf ???

[selig]

Lol post i responded to dissapeared .....wtf ???

Sometimes it takes a second for a new post to appear - you have two posts in this thread, does that include the "disappeared" post?

Is there something I should be responding to in your initial post?



Sent from my iPad using Tapatalk

[spacehorse]

hi there,

i removed my post because i did not want to confuse anyone but at the same time submonsters responded to my post

hope that clarifies ;]

[Kalm]

My philosophy is to have ALL audio signals peaking around -12 dBFS, which leaves enough headroom to avoid clipping in most all cases, and means you don't have to lower all the faders when mixing (or lower the master). There are other advantages, but this is the main one IMO.

I've found -12 dBFS to be the magic number as well over my years of doing audio. That number just magically works.

[selig]

My philosophy is to have ALL audio signals peaking around -12 dBFS, which leaves enough headroom to avoid clipping in most all cases, and means you don't have to lower all the faders when mixing (or lower the master). There are other advantages, but this is the main one IMO.

I've found -12 dBFS to be the magic number as well over my years of doing audio. That number just magically works.

FWIW, same here - AND it's the level that the Reason User Manual suggests for recording audio (the meters change from green to yellow starting at -11 dBFS and above, AND it's what the more recent FSB patches are set to (I worked on some of the R6 FSB additions).

Exceptions would be if you routinely work with 64-128 tracks (yes, it happens), in which case you'd adopt a lower peak reference level, or if you routinely record solo instruments with no overdubs, in which case you'd adopt a higher peak reference level.

[boomer]

Thanks guys, that's a good strategy (although I'll mess with the numbers a bit).

[olive6741]

-12dBFS seems a useful recommendation. it's not the first time I see this 'magic' number hanging here or there. But just to avoid confusion in my mind, we are talking about PEAKS, is it? So it means there is no direct way to measure it on each channel on the mixer directly? (as the indicators are RMS, is it?). Any recommendation to do a quick measurement then, channel by channel? Any thoughts? thanks.

[Djstarski]

-12dBFS seems a useful recommendation. it's not the first time I see this 'magic' number hanging here or there. But just to avoid confusion in my mind, we are talking about PEAKS, is it? So it means there is no direct way to measure it on each channel on the mixer directly? (as the indicators are RMS, is it?). Any recommendation to do a quick measurement then, channel by channel? Any thoughts? thanks.

You can use the meter on your master channel .

[olive6741]

Yes, it was what I was thinking. By soloing the tracks, it seems to be the fastest way (assuming the master is at 0.0 dB)...

[Djstarski]

Yes, it was what I was thinking. By soloing the tracks, it seems to be the fastest way (assuming the master is at 0.0 dB)...

That is correct .

[Koncide]

Useful info here guys, thanks'a'many.