Real average level???

  • Thread starter Thread starter RAMI
  • Start date Start date
R

RAMI

Guest
OK, I hope I can ask this in a way that makes sense, because it makes sense in my head....but that doesn't mean much.

I'm pseudo-mastering a song (I say "pseudo" because I'm no mastering engineer) in Wavelab. For the first 3 minutes of the song, the average level is about -14db. At abut 3:02 a drum fill happens that makes the "average" level shoot up to -12db. The fill doesn't peak and doesn't sound out of place, but it's loud enough to change the average level.

So my question is: Should I still take my average level to be -12db, since that's what it shows me at the end of the tune. But for the first 3 minutes of the song, the level is -14db. It only goes up to -12db for the last 20 seconds, after that drum fill. I don't want my song to sound out of place on a CD because it's lower than the other tunes, even though it's supposed to be -12db. I'm wondering if one part of a tune can throw things off enough to give a sort of "false" average level.

Any thoughts? Thanx.
 
Do you want the louder part to remain louder, or are you trying to even that out some? (not sure how wavelab works exactly, so....)

When I want a whole tune to be a bit louder, I try and boost everything pretty even, so as to retain most of the dynamics....obviously some will reach a peak, and won't go any higher, but I try not to boost too much, so that it loses too much of the dynamics....I don't want the quiet parts to be 1db quieter than the loudest. If this makes any sense. If not, just tell me I'm drunk, and to come back tomorrow..... :D
 
Thanx Dog...Wavelab is a 2 track mastering program. It's where I bring my song after recording and mixing it down. This is where you apply the final eq-ing (if need be) and compression to the overall mix.

I'm not really concerned with taming that fill, it's fine. But until that fill, the average level of the tune is 2db lower than my other tunes. So, it sounds a little lower. But if I turn up the gain or apply more compression to bring up the level, then after that fill, it will be about -10db.
 
Yes, I think it does matter, while I agree with your point 100%. In this case, though, it's not really a matter of the sound so much as having a uniform level to a certain extent. But, even then, your point is valid.

I guess it's more a philosophical question (hehe) about what the "real" average volume wold be in a case like this.
 
Seriously, Rami, what do the numbers matter? They don't tell you how loud a song sounds anyway. All they do is measure an average measurement over a certain arbitrary period of time. This measurement has only a passing relationship to how loud the song is actually perceived by the human ear.

In my compression notebook I gave an example of a rock anthem ("Wired All Night" - Mick Jagger) and a modern blues/alt rock ballad ("Feels Like Rain" - Buddy Guy), both professionally engineered by the best in the business, both that when played back sequentially have the same perceived volume; i.e. one sounds naturally louder than the other because of it's content, but they don't sound like they're mixed or mastered quieter or louder than the other one. They sound "right" together. Yet their measured RMS levels vary by a full 4dB.

Use that drum fill incident as a way to learn how deceiving RMS measurements can be. To say that a song measures at -12dBRMS is akin to saying that the temperature yesterday was 75°F. That 75° may have been accurate at a certain specific location at a specific time of day, but that's all it really means. It could easily have been 82° an hour later and/or 2 miles away. It is entirely possible to have two different days that both had an average temperature of 75°F that felt entirely different because of different humidities, different cloud patters at different times of day, warm or cold fronts that come through at different times but wind up giving the same average temperature over 24 hours, etc.

Trust your ears.

G.
 
SouthSIDE Glen said:
Use that drum fill incident as a way to learn how deceiving RMS measurements can be. Trust your ears.
Click to expand...

Great advice all around, but those 2 statements really answer my question specifically. And pretty much confirm what I suspected. Thanx alot, G.
 
Yeah, I think the only valid question is "Is the drum fill too loud?" If the answer is no, it fits just fine, then you have a -14 tune. Which is a good thing :)

BTW Wavelab isn't just a 2 track program :cool:
 
Thanx MS...makes sense. And you're right about Wavelab...I meant that's how I'm using it in this case.
 
Well, don't worry about numbers, man. Just keep doing what you do. You're way above most people in the mixing part, and definitely got the music down. That's enough for me, man.... :cool:
 
You must log in or register to reply here.
Back
Top