Do u normalize, then hard limit each mix?

  • Thread starter Thread starter Taylor_214
  • Start date Start date
T

Taylor_214

New member
Nubie to home recording here, after mixing all the recorded tracks to one file, i've been normalizing the mix at 95%, then hard limiting at -.3db, boost input by 6 db, look ahead time 7, release time 100, link left and right is checked.

The mix sounds ok but should i be doing this or not? I do this with every mix. I'm trying to understand these features, and what exactly i should and shouldnt be doing.

thx
 
It's a matter of taste, I don't think there's anything technically wrong with your approach except to say that it takes no account of comparative levels with other songs that you may put together on a CD. There is Group Normalise but I prefer to modify just those tracks that sound "out of place" when heard in playlist order.

Depends on your type of music of course, but I prefer a more natural level even for hard rock. A lot of comercial stuff at the moment sounds overcooked to my ears. I own Korns Untouchables CD, I love the music, but it wears me out - it's been mastered for maximum energy all the way through at the expense of dynamics.
 
lol...check out Audioslave's CD - I don't think EVERY song is supposed to have distorted drums :D

Anyway, about the normalization - a lot of people would say that it's an unnecessary step. If you think about it, all normalization does is take the loudest peak in the song and bring it up to wherever you tell to to (in your case, 95%), and it moves the rest of the file up relatively, with no compression. So if you've got one kick spike that's at -.4, your whole file will barely move when you normalize it, but it'll still make the file go through an editing process, and the fewer of those there are, the better.

Personally, I prefer to mixdown as hot as possible without clipping, and then use a limiter, adjusting the output gain until the drums start to lose a little bit of power...then I back off a bit. If I do this with all tunes on a CD, they come out pretty close to even in overall levels.
 
Put another way; If your using a limiter and it has a gain control, or even if you just use the master fader for level trim, then Abnormalizing is completely redundant. Why send it through two gain steps?
Wayne
 
i hard limit tracks that are uneven volume throughout (bass, vocals). i know this could be bad at times but i've never had a problem with it,. i usually normalize to 96.6% after i mix down. i think thats the most you can do without making it clip.
 
kylosius said:
i usually normalize to 96.6% after i mix down. i think thats the most you can do without making it clip.
Click to expand...
:confused:

What in the world are you talking about????????? Normalization is simply a gain multiplier.... if you normalize to 0dBFS, then the difference between the peak level of your track and 0dBFS is determined and the track's gain is multiplied by that amount.

It can potentially degrade the signal due to DSP round-off error, but unless the algorithm is very poor, it's not going to cause clipping............ (unless the algorithm allows you to override the gain multiplier and you increase it beyond the peak difference - which would be a very stupid thing to do.....)
 
i didn't explain to clearly, usually if i normalize above 96.6 it starts to "go into the red zone". there are times it doesn't distort when it goes over 0db for me, but there have been times it has and i just play it safe.
 
OdBFS has nothing to do with what the meters read.... 0dBFS as a level is an absolute value. It's where the entire digital word is filled.

If, for example, the normalize calculation tells you that there's a 1.55dB difference from peak to 0dBFS, then the algorithm will multiply the track's gain by 1.55dB, making the top value 0dBFS. It CANNOT clip in this case, unless you override the gain value and increase it beyond the calculated difference.

What you're saying indicates you're not very clear on the whole process and what's going on with your digital signal..........
 
You Know...

...it's always good to see you poking around in here.
 
I don't ever normalize... personally I don't think it's a necessary step.. and I wouldn't touch the hard limiter in cep with a ten foot pole.. (not that I'm down on cep.. cause I'm not) BUT: if it sounds good then what your doing is the right thing to do. If your trying to max out your levels as final step in your process the waves L1 ultramaximizer plug-in is excellent for that and it's got some very usable factory presets.
 
Hi Taylor, I remember in a previous thread you were complaining about background noise.... looks like your normalising is the culprit here. :)

In short, don't normalise.
 
I used to use Hard Limiting, but realized it wasn't needed. Hard limiting does exactly what it says... It's a gain limiter, and merely increases volumes. It's a redundant step for me...

For vocal editing, I have 3 steps. I use a very small noise reduction to eliminate small background noise. Secondly, I use a 4:1 compression to get my levels on the same playing field, and lastly, I EQ them. I used to hard limit them after the EQ, but that's not needed really. I only need to mix it correctly. :)
 
just adding one little note...
i don't ever use noisereduction, maybe its just me,
but a little bit of hiss sounds better than that mp3-isch sounding noisereduction effect...

vocals...compressor...eq....and maybe a tiny touch of reverb or other "toys.."

i rather try to hide the noize with a gate or expander,
but thats only for the silent moments...it doesn't extract noise...
my signal path isn't super high-Q, but i'm happy :)

one neat mic and good preamp...and forget about noise..
 
You must log in or register to reply here.
Back
Top