when to normalize?

[Farview]

There is no reason to back anything off if the main buss is only peaking at -5dbfs. That's fine, you don't need any more headroom at the main buss. There is certainly no reason to back everything down by 15db.

If you are trying to get peak levels of -18dbfs, you are going way overboard with the headroom thing. -18dbfs should be your average level, not your peak level. If your peaks are hitting -5dbfs, your average level is probably just about right.

Also, if you insert the compressor on one of the bottom 2 channel inserts, those are post fader, which will be a problem. Insert it in one of the top ones, then the compressor will act on the prefade signal.

 

[sixer2007]

Thanks for that Fairview. After I did my first mix or two some time back and thought they sounded absolutely terrible, I read some things about digital clipping and how it's a total No-no, and sure enough, it was everywhere on those tracks. So then now I feel like my tracks can be as low as possible while mixing as long as my monitors still have some range in the vol knob haha. It keeps me from touching red in any place including plugs which can be a bit of a pain sometimes. Yeah, probably a bit overboard. Thanks!

Oh, and so far in my mixing existence I've only used one of those bottom inserts one time.. I haven't had a need yet, and i can usually get a similar thing with sends. I'm sure that will come though.

 

[chuckduffy]

Also, if you insert the compressor on one of the bottom 2 channel inserts, those are post fader, which will be a problem. Insert it in one of the top ones, then the compressor will act on the prefade signal.

Wow - I learn something every day. been using cubase forever and never knew this. I was racking my brain for an explanation how inserts could be operating post fader...

 

[mixsit]

There is no reason to back anything off if the main buss is only peaking at -5dbfs. That's fine, you don't need any more headroom at the main buss. There is certainly no reason to back everything down by 15db.

If you are trying to get peak levels of -18dbfs, you are going way overboard with the headroom thing. -18dbfs should be your average level, not your peak level. If your peaks are hitting -5dbfs, your average level is probably just about right.

Also, if you insert the compressor on one of the bottom 2 channel inserts, those are post fader, which will be a problem. Insert it in one of the top ones, then the compressor will act on the prefade signal.

You must spread some Reputation
around before giving it to Farview again.

grinns..

 

[Washburn100]

One problem with normalizing is you also raise the noise level. Say you record a track and have some ambient noise, there always is. If the recording level is low, you normalize and bring the level of everything up including the background noise. Now if you just recorded a little hotter, you have a better level, but the ambient noise isn't increased as much as normalizing.

 

[Farview]

One problem with normalizing is you also raise the noise level. Say you record a track and have some ambient noise, there always is. If the recording level is low, you normalize and bring the level of everything up including the background noise. Now if you just recorded a little hotter, you have a better level, but the ambient noise isn't increased as much as normalizing.

But normalizing doesn't change the signal to noise ratio any more than just turning the volume up does. (because that is what normalizing is) So it doesn't do any more 'harm' than any other method of turning the volume of the track up to get it to sit in the mix properly.

So the background noise is a non-issue.

 

[DrewPeterson7]

But normalizing doesn't change the signal to noise ratio any more than just turning the volume up does. (because that is what normalizing is) So it doesn't do any more 'harm' than any other method of turning the volume of the track up to get it to sit in the mix properly.

So the background noise is a non-issue.

That goes back to my question from a few pages ago, though - is the general disdain for normalizing due to the fact that it doesn't discriminate between signal and noise, while carefully applied upwards compression potentially could?

 

[mixsit]

One problem with normalizing is you also raise the noise level. Say you record a track and have some ambient noise, there always is. If the recording level is low, you normalize and bring the level of everything up including the background noise. Now if you just recorded a little hotter, you have a better level, but the ambient noise isn't increased as much as normalizing.

Hmm. Actually I think you have it wrong there
Any level change moves a wanted signal and your recorded noise proportionally (equally. Sig-to noise would remain the same.
(I guess there's a bit of math error added in with any process? )
As well, record level would not effect sig-to ambient level.

Changing the dynamic range is another thing all together. I need a review. Is upwards compression = expansion?

 

[RAMI]

One problem with normalizing is you also raise the noise level. Say you record a track and have some ambient noise, there always is. If the recording level is low, you normalize and bring the level of everything up including the background noise. Now if you just recorded a little hotter, you have a better level, but the ambient noise isn't increased as much as normalizing.

That defies logic. If you record hotter, you're also raising the noise. It's the exact same thing. It's not the same as turning up to make your signal louder than tape hiss. In digital, there's no tape hiss to over-come. Ambient noise is part of the signal being recorded. Record hotter and the ambient noise gets louder, just like it would if you normalize.

 

[Bobbsy]

With modern gear, noise shouldn't be much of an issue any more. Modern mics and interfaces are very quiet--S/N ratios in the high 80s or low 90s are not uncommon. I'm talking about system noise here--if we're referring to background noises in your studio, normalising leaves the ratio between programme and the noise the same anyway.

...and that's the nub of things. Compression changes the ratio between the loudest and quietest bits; normalising just raises everything by the same amount and uses the dynamic range the same. It's exactly the same as pushing a fader up a bit--except that you can decide where the loudest bits end up. If you push the fader on a quiet bit, the loud bits could clip.

In my workflow, sometimes the fader push is convenient; other times normalising works for me. It's down to individual preference. The only main point I was trying to make earlier is that there's nothing in normalising that automatically messes up your signal--and it's a totally reversible process (subject to the integer vs. floating point thing I mentioned earlier but that rarely comes into play). It's the misuse of normalising that can get you into trouble--but that applies to every effect in your tool box.

 

[mixsit]

"Upward compression increases the loudness of sounds below a threshold while leaving louder passages unchanged"

So you'd be raising the noise floor and the lower level portion of your wanted signal (depending on threshold and where 'noise and low sig' is.

 

[miroslav]

...is the general disdain for normalizing due to the fact that it doesn't discriminate between signal and noise, while carefully applied upwards compression potentially could?

As a process, it's no different than anything else in the DAW that uses mathematical calculations to adjust levels.
I think the concern or caution with normalizing is that it uses peak levels VS average levels...so if you just pick some reference level to normalize all your individual tracks or your album tracks to...you can mess up their average level balance even though they are all hitting the same peak level after normalization.

You just have to watch for that trap, and spend more time listening to the average level balance instead of just shooting for a numerical value...otherwise, normalizing is just another tool that has its uses.

 

[Washburn100]

But normalizing doesn't change the signal to noise ratio any more than just turning the volume up does. (because that is what normalizing is) So it doesn't do any more 'harm' than any other method of turning the volume of the track up to get it to sit in the mix properly.

So the background noise is a non-issue.

I understand what everyone is saying but that assumes if you record hotter, you are also recording more of any ambient noise. Technically you are, but it is not directly proportional because of distance to the source. Lets say I sing 6 inches from a condenser mic and record at a low level. Then I do it again with a higher level. My voice is 6 inches away and the volume increase a certain amount. But if say I have a low level of noise from my tower. Recording at a low level picks up a certain amount of this noise. Recording at a higher level, I will get more of this noise, but not near as much as the increase in my voice.

If I normalize the low level recording, the background noise is raised significantly more than what I get from recording at a higher level.

 

[miroslav]

I understand what everyone is saying but that assumes if you record hotter, you are also recording more of any ambient noise.

Not really...depends on how you are arriving at "hotter".

If you crank the gain on your preamps or recorder inputs...then yes, you will pull in more ambient noise too.
However, if you simply increase the level of you *signal source* (turn up your amp, sing a little louder, etc)...then you have a hotter signal with less ambient noise.

 

[mixsit]

I understand what everyone is saying but that assumes if you record hotter, you are also recording more of any ambient noise. Technically you are, but it is not directly proportional because of distance to the source. Lets say I sing 6 inches from a condenser mic and record at a low level. Then I do it again with a higher level. My voice is 6 inches away and the volume increase a certain amount. But if say I have a low level of noise from my tower. Recording at a low level picks up a certain amount of this noise. Recording at a higher level, I will get more of this noise, but not near as much as the increase in my voice.
If I normalize the low level recording, the background noise is raised significantly more than what I get from recording at a higher level.

This is just not the case. I'd bet that you could be falling into the similar false logics where people seem to think they'll get different amb-to sig levels adjusting gain (input trim) on a PA vs raising the fader thinking one does more to background spill than the other.

Or.. be carful you're not ignoring that (perhaps) when varying the record gain you're not also changing the micing distance, and/or singing/playing louder (unconsciously?
These are where amb-to sig ratios actually would change.

 

[mixsit]

Not really...depends on how you are arriving at "hotter".

If you crank the gain on your preamps or recorder inputs...then yes, you will pull in more ambient noise too.
However, if you simply increase the level of you *signal source* (turn up your amp, sing a little louder, etc)...then you have a hotter signal with less ambient noise.

I like the fun way to do it is- How does it know? (How would it know if the sound being brought up or down, is the louder/closer/stuff we want.. or the quieter/farther stuff we don't?

 

[Bobbsy]

This is just not the case. I'd bet that you could be falling into the similar false logics where people seem to think they'll get different amb-to sig levels adjusting gain (input trim) on a PA vs raising the fader thinking one does more to background spill than the other.

Or.. be carful you're not ignoring that (perhaps) when varying the record gain you're not also changing the micing distance, and/or singing/playing louder (unconsciously?
These are where amb-to sig ratios actually would change.

Exactly this. If you sing or play louder you change the "performance to ambient" ratio. If you simply crank up the gain on your mixer or pre amp, the ratio stays the same and the effect is the same as normalising or pushing up the virtual mixer on your DAW.

 

[DrewPeterson7]

"Upward compression increases the loudness of sounds below a threshold while leaving louder passages unchanged"

So you'd be raising the noise floor and the lower level portion of your wanted signal (depending on threshold and where 'noise and low sig' is.

Maybe I'm describing it wrong then - using a graphic compressor to expand rather than compress sounds above a certain threshold?

 

[mixsit]

Maybe I'm describing it wrong then - using a graphic compressor to expand rather than compress sounds above a certain threshold?

There you go then- you actually got down to your specific action.
I can know exactly want I might want in a function- I just have to keep looking up which they're called
Here's one, doesn't show the graphic for it but you'd be doing upward expansion then.
compressor, gate and expander
Signals above the rotation point on a graphic comp would increase.

 

[Washburn100]

I disagree. Remember, sound power is an inverse square.
If you stand beside a jet and measure the sound level (power) say you get 189 db. As you move away from the source, the sound level drops by a factor of two. So, if I sing 6" away from a mic, then turn up the preamp gain, my voice will be amplified. Any other noise will also be amplified, but if it's farther away than my voice, the amount of amplification is much smaller than my close voice.