Lt. Bob
Spread the Daf!
Beat me to some of that.
If the overall signal is increased , then the fact that the noise floor increases has no real effect. The ratio or proportion of noise to desired signal is exactly the same.
Beat me to some of that.
If the overall signal is increased , then the fact that the noise floor increases has no real effect. The ratio or proportion of noise to desired signal is exactly the same.
SouthSIDE Glen
independentrecording.net
Not quite accurate, in two ways.
First, during the quiet parts of the song - rest beats, reverb tails, etc. - the noise floor will be more audible.
Second, the dynamic range between the peak and the noise floor does not change, that much is true.
But as I understand it, the S/N ratio *does* change, because one needs to consider the noise floor not as a zero point, but rather as an amount of noise above the theoretical digital floor. Because the total range (the bit depth of the mastered signal) remains static, the *percentage* of total signal that the noise takes up increases, and therefore the S/N decreases.
To use the numbers I used before, with a starting peak of -5dBFS and a noise floor of -65dBFS, let's assume a working master of 16 bits depth (it's going to CD.) That means a total usable range of 0 to -90dB before one gets to the theoretical digital noise floor. That puts the recording's noise floor at 25 dB (-65dBFS - -90dBFS) and the peak level at 85dB (-5dBFS - -90dBFS) above digital ground zero. That yields a signal to noise ratio of 3.4 (85/25). If you add 5dB to everything, the S/N is now 90/30 or a decreased ratio of 3.0.
G.
Last edited:
Lt. Bob
Spread the Daf!
cool info ..... thanks,
but it makes me glad I don't do recording in the 'puter.
Don't have to worry about such things and normalization isn't even an available option.
Ultimately, for me, the only thing I ever worry about is what my ears hear but I guess it's easier to go that way when you don't have a 'puter screen full of spectral graphs and endless options.
And let me add the obligatory old-man comment of "Loud loud LOUD might be the current trend but it sucks the dynamic life right out of recordings.
And you kids get off my lawn!
SouthSIDE Glen
independentrecording.net
The details described don't don't really change much over in analog. The wallpaper changes, but the room remains the samecool info ..... thanks,
but it makes me glad I don't do recording in the 'puter.
Don't have to worry about such things and normalization isn't even an available option.
Ultimately, for me, the only thing I ever worry about is what my ears hear but I guess it's easier to go that way when you don't have a 'puter screen full of spectral graphs and endless options.
And let me add the obligatory old-man comment of "Loud loud LOUD might be the current trend but it sucks the dynamic life right out of recordings.
And you kids get off my lawn!
.Our tape decks and mixers may not have a "normalize" button, but because all peak normalization actually is is a global gain increase, on the analog side it's the same thing as pushing a fader up by xdB.
And S/N calculations, though using a different type of decibel scale than digital uses, still requires assigning an absolute decibel (volume) value to both signal and noise, because a noise level of zero would mean having to divide by zero, which (as anybody with a calculator can tell you
), cannot be done and can only yield an answer of zero.The things there is anytime you take a higher value (like signal) and a lower value (like noise) and add a fixed amount to both, that fixed amount will always be a greater percentage increase to the smaller value than it is to the higher value. For example, 5dB is a 20% increase to a noise level of 25dB, but it is only about a 6% increase to a level of 85dB. This changes/reduces the overall ratio of signal to noise, whether it's done via digital normalization or an analog fader push.
Again, the actual dynamic range - the distance between peak and noise - does not change. Nor does the crest factor - the distance from RMS to peak - change. So in that way, yes the signal seems to remain the same, and it does.
But when considering S/N, one must consider the absolute levels of the signal related to the medium, and not just the properties of the signal in and of itself.
And those kids better stay off the lawn; I've had to double my blood pressure medication this week. Only 50 and the pill parade is already starting ...
.G.
D
dintymoore
Guest
Personally, on Cubase, I'd just put the volume of the tunes that were soft up.
Am I missing something?
What's next - phone a plumber to flush your toilet?
If all you really want is the volume changed on a few songs you could have done it by the time you read these posts.
C
Chibi Nappa
New member
C'mon. You know what he means. He wants the volume raised past the digital ceiling. No, he can't just raise a fader in Cubase.
I'll say what I always say: Don't do it and let your song remain "quiet". Even if you have $10,000 to send it off to Bob Ludwig, don't do it. There is no transparent way to raise the volume past the ceiling even with $200,000,000 of equipment made from unicorn horn operated by Goldenears McMastersalot.
If your fans like your song, they won't avoid it because they have to take 1/2 a second out of their life to turn the volume knob.
If you fans don't like your song, they won't go out of their way to hear it because they don't have to take 1/2 a second to turn the volume knob.
The whole thing makes no sense. Take it to a mastering guy. Tell him to make it sound good. Let the volume be.
D
dintymoore
Guest
R
Richard Monroe
Well-known member
Looks like there are three main issues here:
1. Does gain staging in normalization create a noise floor problem? That depends on the intervals of no signal, or not, in the music. If you play rock with a noisy humming amp with a Fender, it's not a problem. If a string quartet stops, and you hear an audible hiss, it sucks. We all have to decide how much *absolute* noise we are going to live with.
2. Is normalization a useful tool for mastering, or in fact, for anything at all? No clue. It looks like pro engineers don't agree on it. I'm just a tracking tech. What do I know?
3. Should the kid crank up his own recordings, or pay an engineer to do it, and secondly, should he pay a mastering house to do the whole deal? If you just want it louder, it shouldn't cost $400. More like $50-$100. But Drew, you argue for and against the same thing. *You* want to hear *your* songs the best they can sound, to hold that gold pressed master in your hand, with the liner notes, the release party, the whole 9 yards, and are willing to shell out to do it, but this kid should worry about how many copies he sells on itunes? His dreams are as big as yours (and mine). No, kid, you won't make the money back, but you won't regret owning properly mixed and mastered recordings either. Pay the nice engineer.-Richie
1. Does gain staging in normalization create a noise floor problem? That depends on the intervals of no signal, or not, in the music. If you play rock with a noisy humming amp with a Fender, it's not a problem. If a string quartet stops, and you hear an audible hiss, it sucks. We all have to decide how much *absolute* noise we are going to live with.
2. Is normalization a useful tool for mastering, or in fact, for anything at all? No clue. It looks like pro engineers don't agree on it. I'm just a tracking tech. What do I know?
3. Should the kid crank up his own recordings, or pay an engineer to do it, and secondly, should he pay a mastering house to do the whole deal? If you just want it louder, it shouldn't cost $400. More like $50-$100. But Drew, you argue for and against the same thing. *You* want to hear *your* songs the best they can sound, to hold that gold pressed master in your hand, with the liner notes, the release party, the whole 9 yards, and are willing to shell out to do it, but this kid should worry about how many copies he sells on itunes? His dreams are as big as yours (and mine). No, kid, you won't make the money back, but you won't regret owning properly mixed and mastered recordings either. Pay the nice engineer.-Richie
Waltz Mastering
New member
If your talking about a very limited release to friends and family, I would agree, but if you are talking about a release destined for any kind of distribution or to be taken seriously outside of a home studio environment, I think your missing the whole point of professional mastering.
SouthSIDE Glen
independentrecording.net
He wants to sell/distribute his stuff to the public on iTunes. Which unfortunately means that the wants end users to be able to stick his songs into their playlists without his songs sounding "wimpy" compared to the rest of the over-smashed, over-mastered crap out there. While completely misguided, it's an understandable position for a newb to take.
@14 (probably now 20 or 21 ); If you can leave your ego at the door and objectively and honestly say that your recordings and mix are actually pretty high quality, then by all means, take it to a *real* mastering engineer like the ones that hang out here. Nice BMWs deserve nice detailing.
If in all honesty you can say that you have newb recordings and a newb-quality mix, then don't waste your money on a pro mastering job; head over to kvraudio.com and search for a free "limiter" or "finalizer" plug, and learn how to use that to slam your mix against the wall, and then just move on. You're stuff will then sound pretty much just like the rest of the user-submitted '92 Ford Taurus' out there.
G.
@14 (probably now 20 or 21
); If you can leave your ego at the door and objectively and honestly say that your recordings and mix are actually pretty high quality, then by all means, take it to a *real* mastering engineer like the ones that hang out here. Nice BMWs deserve nice detailing.If in all honesty you can say that you have newb recordings and a newb-quality mix, then don't waste your money on a pro mastering job; head over to kvraudio.com and search for a free "limiter" or "finalizer" plug, and learn how to use that to slam your mix against the wall, and then just move on. You're stuff will then sound pretty much just like the rest of the user-submitted '92 Ford Taurus' out there.
G.
D
dintymoore
Guest
One good thing about doing yourself is that you might learn how to do it.
My guess is that most people starting out who don't have a lot of money would be better off doing that.
It doesn't have to sound bad if you use your ears.
My guess is that most people starting out who don't have a lot of money would be better off doing that.
It doesn't have to sound bad if you use your ears.
Waltz Mastering
New member
My 97 Mitsubishi Montero begs do differ, of course my wife gets the nice car, but it's no Beemer : )
N
NYMorningstar
Recording Modus Operandi
That's pretty funny
C
Chibi Nappa
New member
???
I said to tell the mastering guy to make it sound good and not push the volume past the ceiling. How is that missing the point of professional mastering?
I think especially if you are making it for wide distribution you don't want to push the volume. Every body does it, so you won't stand out if you do it too. You will just be one more in the multitude. And your song will sound worse than it could have. You lose on both fronts.
Waltz Mastering
New member
When it comes to loudness or pushing the volume, there is usually a point that you reach where things start to fall apart. Knowing where that point is and maximizing the potential for improvement over degradation is what makes a good ME worth the while. Part of the trick is to get loud and good. (insert sarcasm here)
It all up to the client though, when they order a steak well done you don't serve it rare.
adamvelvetu
New member
I have to go with Tom on this one. I used to work as a mastering engineer, have since left and gone into teaching English, and have finally put together a home studio. Even having had a history as an ME I would not attempt to master my own stuff in my home studio and would gladly pay someone to do it (providing I heard samples of their work), especially if it were going on iTunes and I was hoping to build an audience.
Lt. Bob
Spread the Daf!
well, I see that ...... but in the 'day' we never talked about having ANY db of noise floor. It mostly never came up in that sense. It would be more a matter of, "Well crap ...... that's too noisy". And as long as it was quiet enough we didn't think about it further.The details described don't don't really change much over in analog. The wallpaper changes, but the room remains the same
I mean .... we'd take steps to make sure it was quiet enough ..... record up against the red and into it for some instruments ...... but I think it ended up being more of an intuitive thing .... kinda like playing. I never think, "Ohh, I'm gonna play in Gm over his E" .... I just do it.
It seems to me that there's a lot more electronics engineer mindset that has to be applied nowadays.
I'm not criticizing ........ just commenting ......
DrewPeterson7
Sage of the Order
Um, not to over-intellectualize a tossed off comment, shouldn't that be "G over his Em? A Gm triad would be, respectively, the b3, b5, and b7 of an E major, which is (to quote a passage a certain forumite is also fond of) almost exactly unlike an E major triad.
SouthSIDE Glen
independentrecording.net
You're right, I don't disagree with that. But noise floor is always in the back of the mind, even in an "old guy's studio", at least as keeping it low is a major goal of good analog gain structure.well, I see that ...... but in the 'day' we never talked about having ANY db of noise floor. It mostly never came up in that sense. It would be more a matter of, "Well crap ...... that's too noisy". And as long as it was quiet enough we didn't think about it further.
I mean .... we'd take steps to make sure it was quiet enough ..... record up against the red and into it for some instruments ...... but I think it ended up being more of an intuitive thing .... kinda like playing. I never think, "Ohh, I'm gonna play in Gm over his E" .... I just do it.
It seems to me that there's a lot more electronics engineer mindset that has to be applied nowadays.
I'm not criticizing ........ just commenting ......
The only reason I even brought it up was because of the context of the argument as to whether peak norming increased the noise level or not (it does), and regarding the use of the term "signal-to-noise", which was being referred to a bit incorrectly by a few folks here.
But yeah, we almost never talk about S/N in the studio, that's more for datacom or telecom engineers or spec marketers to worry about. In our signal path, it's really little more than a numbers game. It's not whether we care if the S/N is 3.4:1 instead of 3:1, the specific measurement isn't something we ever really take, we just try to keep the noise low and the dynamics high (unless we're a newb or a producer in the 21st century, in which case we don't give a shit about the S/N *or* the dynamics, as long as it's all loud enough
).G.
C
Chibi Nappa
New member
But people shouldn't push it until just before it falls apart. You should stop pushing before it gets worse in any way. Most of what I hear today stops just short of the "fall apart" line. Some stuff goes over it. Almost nothing chooses no damage at all. For what? 6-8 db of gain? Not worth it. Not even sane.
This is what has me scratching my head. Improvement over degradation: What improvement? There is nothing to be gained by pushing for volume that the consumer's volume knob can't do (much) better. It is all degradation. Just don't do it.
Send it to the ME for EQ. Send it to the ME for dithering. Send it to the ME for 2-buss compression (way different than pushing for volume). Send it to the ME to make it sound it's best. Volume is in the hands of the consumer anyway. Don't even bother.
Agree 100% Which is why I'm posting in a thread to a client trying to persuade him to order his steak rare. The ME certainly has to do what the paying customer instructs.
Last edited: