Z
Zeppe
New member
Forgot about this kid?
Forgot about this kid?
philbagg
Just Killing Time
Who's this n00b tellin us we don't know what shit is?
Z
Zeppe
New member
Explain this? You raise your average level and it raises your noise floor. Its physics man.
Z
Zeppe
New member
I was wondering how long that would take. lol, Might be a newbie here but definitely not in the audio world.
philbagg
Just Killing Time
Raising your noise floor is not a problem as long as the signal sounded good beforehand.
Its raising your noise floor, and raising your signal by the same amount.
Signal to noise ratio stays the same.
It's maths man.
philbagg
Just Killing Time
I've been hearing shit since I was a baby too.
bozmillar
New member
Zeppe, normalizing does not increase the noise any more than it increases the signal level, so your SNR is the same before and after normalizing. There's no cumulative noise effect. I'm not quite sure where you are getting that from.
And the OP stated that he doesn't want his tracks mastered, he wants them all to be the same level. This is a very simple process where normalizing can be a first step.
Everything that DrewPeterson7 has said so far has been valid and answering the OP's question.
If all you want to do is get them to sound as loud as each other, I'm sure there are 100s of people on here that would jump at the chance to do that for $400. If you want to get them mastered, there might be a couple people that would do a good job at that price range.
i think everything Drew said is right. What is your ultimate goal here? Once you start forking out money to get something done, it changes from a hobby to an investment. If you think you'll make your money back, then great. If you don't think you will, then why spend the money to get it mastered?
And the OP stated that he doesn't want his tracks mastered, he wants them all to be the same level. This is a very simple process where normalizing can be a first step.
Everything that DrewPeterson7 has said so far has been valid and answering the OP's question.
If all you want to do is get them to sound as loud as each other, I'm sure there are 100s of people on here that would jump at the chance to do that for $400. If you want to get them mastered, there might be a couple people that would do a good job at that price range.
i think everything Drew said is right. What is your ultimate goal here? Once you start forking out money to get something done, it changes from a hobby to an investment. If you think you'll make your money back, then great. If you don't think you will, then why spend the money to get it mastered?
mattr
Resident moody teenager
What is this obsession with normalisation that seems to have arisen recently?
Lets get this straight... normalisation does nothing useful.
Its not doing anything that you can't do simply by raising a fader, or some other kind of gain stage.
And I don't quite understand what the argument is even about... noise floor? Blah Blah Blah.
Yes - "it brings the noise floor up" - but only along with the signal itself... the SNR remains the same... reduce the gain again and you end up with exactly the same as what you started with.
Aside from the problem that there need be no argument in the first place, we're probably talking 32-bit floating point audio here, which you can put through a horrendous ordeal of gain adjustments way out from anything you would ever realistically throw at it... and so long as you bring it back to a 'normal range' by the time it comes back into the fixed-point realm then it will come out the other end unscathed.
Any kind of compression will actually 'destructively' increase the noise floor.
Lets get this straight... normalisation does nothing useful.
Its not doing anything that you can't do simply by raising a fader, or some other kind of gain stage.
And I don't quite understand what the argument is even about... noise floor? Blah Blah Blah.
Yes - "it brings the noise floor up" - but only along with the signal itself... the SNR remains the same... reduce the gain again and you end up with exactly the same as what you started with.
Aside from the problem that there need be no argument in the first place, we're probably talking 32-bit floating point audio here, which you can put through a horrendous ordeal of gain adjustments way out from anything you would ever realistically throw at it... and so long as you bring it back to a 'normal range' by the time it comes back into the fixed-point realm then it will come out the other end unscathed.
Any kind of compression will actually 'destructively' increase the noise floor.
DrewPeterson7
Sage of the Order
Oh, no arguments here, I made the same comment myself several posts ago:
And for the gain stages, not be needlessly argumentative, but you categorically stated a while back that you shouldn't normalize because that raises your noise floor, but rather you should use the output knob on a multiband compressor, which seems to me like it would raise the noise floor just as much (ignoring philbagg's comment about the impact of the compression, which is a separate issue IMO and not one related to a pure change in volume), if not even more due to the added analog gain stage. I agree with you that normalizing individual tracks is a bad idea, and that normalizing to 0dB before you send it off to be mastered is pointless and if anything makes the ME's job harder, but I don't think you can categorically say normalizing is a worse way to get a signal louder than using the output on a compressor simply because it's actually a cleaner signal path (i.e - no signal path at all).
philbagg
Just Killing Time
Zeppe, you debated me on this, then told us we don't know what a mastering engineer does. You sure you've been around long in the audio world?
Just curious.
DrewPeterson7
Sage of the Order
Though, I will grant him that you make the noise louder when you normalize, by virtue of the fact that you make everything louder. That's just as true with using a normalizer to boost a mix 10db as it is using the output on a compressor to boost it 10db, although with the added caveat that if we're talking about an analog piece of rack mix, that could actually increase the amount of noise in the mix ever so slightly.
Z
Zeppe
New member
You're right, and in a stereo situation when it is a finished amateur that is fine. I feel like I said this before but... my points were...
-Don't send normalized tracks to a mastering engineer.
-Don't normalized individual tracks in a multi-track recording.
-Normalizing is no where near the same thing as mastering and should not even be uttered in the same breath.
-Send your work to get mastered if you think normalization is mastering!
Thank you, I hate normalization even more after today.
-Don't send normalized tracks to a mastering engineer.
-Don't normalized individual tracks in a multi-track recording.
-Normalizing is no where near the same thing as mastering and should not even be uttered in the same breath.
-Send your work to get mastered if you think normalization is mastering!
Thank you, I hate normalization even more after today.
Last edited:
philbagg
Just Killing Time
A mastering engineer might normalize a track once it's been finished, just so that it DEFINITELY peaks at a certain level.
There's nothing wrong with it at all. But hell, if you feel that way, you might not want to touch any faders on a DAW or mixing console, or volume knobs on a stereo.
DrewPeterson7
Sage of the Order
I could tell it was a purely semantic debate about halfway through, anyway, but whatever. 
A normalizer is a great tool - it's a very quick, easy way to take a mix and nudge it up to exactly the peak volume you want it to be. I think as long as you know there's a lot more to mastering than making a song louder, and think about WHY you're making something louder rather than just assuming everything should be loud, it's a pretty damned useful digital effect.
I fully intend on paying someone to master my stuff when I finish it, but for the time being, if I mix down a demo and want to see how it sounds, or if I want to share it with a couple friends for feedback and criticism, I kind of like to get it up to the level of a "pro" CD just because that's what people (myself included) are used to hearing. So, I definitely normalize finished demos all the time, and then run them through a volume maximiser - I just have no illusions that what I'm doing is "mastering" them.
........aaaaand I think where this tangent started was if all the OP wants to do is get his mixes louder, this is a pretty good way to do it, and if he doesn't expect to make at least $400 on iTunes, it's better to do this than pay $400 to have his songs mastered.
mattr
Resident moody teenager
...the same would also be true when you turn the volume up on your playback system to compensate for the quieter file.
I would rather apply clean gain in the digital domain than gain in the average cheap circuitry and average cheap amp of your average cheap playback device.
I would rather apply clean gain in the digital domain than gain in the average cheap circuitry and average cheap amp of your average cheap playback device.
Z
Zeppe
New member
The OP intends on selling his songs. I'm opposed to just normalizing your tracks and considering them a finished professional product. Normalization is fine for demo's and such but not a professional product.
I didn't mean to cause such a heated debate, but ill take the beating you normal lovers.
Phillbagg - Take it easy buddy. I thought this was a forum where you can learn and debate topics on...??? If you want my resume I'll be more than happy to email it to ya.
That is all.
I didn't mean to cause such a heated debate, but ill take the beating you normal lovers.
Phillbagg - Take it easy buddy. I thought this was a forum where you can learn and debate topics on...??? If you want my resume I'll be more than happy to email it to ya.
That is all.
philbagg
Just Killing Time
It's one L.
I'm quite calm. Just saying I know what mastering is, and that you don't seem to have the right idea of normalizing, which is why I questioned your experience in the audio industry.
Also, I'm dying for a shit but I'm too lazy to stand up. So I'm gonna take out my anger on people here on the boards. I thought that was the right thing to do?
SouthSIDE Glen
independentrecording.net
Hold on guys, you're about to reach the point where your mother will turn this thread around and take you back home, and there'll be no burgers and ice cream for anybody 
.
The fact is, you're both mostly right.
Peak normalization (I wish people would stop just calling it "normalization", there are a dozen ways to "normalize" any given signal) is nothing more than a linear volume gain. Period. All it does is look at the highest peak on the signal, determine the number of dB between that peak and the user-selected dB level, and boost the volume of the entire signal by the difference. You want math? Here's an example:
Lety's say the kid's signal has a maximum peak level of -5dBFS and a noise floor at -65dBFS. In other words, fairly typical. If he runs that through peak normalization set to, say, 0dBFS, all the normalizer will do is increase the gain of the entire track by 5dB.
Now, that means that the *dynamic range* of the signal remains unchanged. I don't think it's quite exactly right to say that the S/N ratio remains the same, because the ratio of noise to the entire signal from -inf increases by *a larger percentage from the original* than the peak signal does, but neither the dynnamic range nor the crest factor of the signal is affected. And I think this is what one side was really trying to say.
At the same time, the other side is also correct in that the normalization will also increase the noise floor by 5dB, because the entire signal - including the noise - will be boosted by the volume gain.
If one is that worried about the noise floor (which is probably the least of the OP's problems, quite frankly), peak normalization is not what concerns me the most; it's the throwing the signal against the limiter without any other processing, because *that* will significantly decrease the S/N and bring the noise floor up in a mich more audible way than peak normalization will.
In either case, the best way to deal with that noise floor in post would be yo apply some heavy downward expansion with a threshond set just above the noise floor. This can be done before or after either the normalization or the limiting, but my personal preference (YMMV) is to get the expansion out of the way first, knocking it down to near or below the digital or bit depth conversion limit right off. and keep the amount that I have to worry about it in final mastering to a minimum.
G.
.The fact is, you're both mostly right.
Peak normalization (I wish people would stop just calling it "normalization", there are a dozen ways to "normalize" any given signal) is nothing more than a linear volume gain. Period. All it does is look at the highest peak on the signal, determine the number of dB between that peak and the user-selected dB level, and boost the volume of the entire signal by the difference. You want math? Here's an example:
Lety's say the kid's signal has a maximum peak level of -5dBFS and a noise floor at -65dBFS. In other words, fairly typical. If he runs that through peak normalization set to, say, 0dBFS, all the normalizer will do is increase the gain of the entire track by 5dB.
Now, that means that the *dynamic range* of the signal remains unchanged. I don't think it's quite exactly right to say that the S/N ratio remains the same, because the ratio of noise to the entire signal from -inf increases by *a larger percentage from the original* than the peak signal does, but neither the dynnamic range nor the crest factor of the signal is affected. And I think this is what one side was really trying to say.
At the same time, the other side is also correct in that the normalization will also increase the noise floor by 5dB, because the entire signal - including the noise - will be boosted by the volume gain.
If one is that worried about the noise floor (which is probably the least of the OP's problems, quite frankly), peak normalization is not what concerns me the most; it's the throwing the signal against the limiter without any other processing, because *that* will significantly decrease the S/N and bring the noise floor up in a mich more audible way than peak normalization will.
In either case, the best way to deal with that noise floor in post would be yo apply some heavy downward expansion with a threshond set just above the noise floor. This can be done before or after either the normalization or the limiting, but my personal preference (YMMV) is to get the expansion out of the way first, knocking it down to near or below the digital or bit depth conversion limit right off. and keep the amount that I have to worry about it in final mastering to a minimum.
G.
C
Chibi Nappa
New member
Just so we're all clear on this: Turning up the volume on your home stereo raises the noise floor.
I don't think anybody would advise against turning up the volume on your stereo (neighbors aside).
That said there are plenty of other reasons not to normalize all willy-nilly.
I don't think anybody would advise against turning up the volume on your stereo (neighbors aside).
That said there are plenty of other reasons not to normalize all willy-nilly.
Waltz Mastering
New member
Using peak normalization in the context of mastering will do very little/nothing for you.
It can be useful for other applications, but not as a process used in any part of mastering ime.
Also wanted to mention Tom at Master House is another ME to consider that was not previously mentioned.
... and a good de-esser that is free is spitfish.
http://www.digitalfishphones.com/main.php?item=2&subItem=5
and Massey has a full functioning demo.
It can be useful for other applications, but not as a process used in any part of mastering ime.
Also wanted to mention Tom at Master House is another ME to consider that was not previously mentioned.
... and a good de-esser that is free is spitfish.
http://www.digitalfishphones.com/main.php?item=2&subItem=5
and Massey has a full functioning demo.