coordinating levels of songs on a CD

  • Thread starter Thread starter dobro
  • Start date Start date
dobro

dobro

Well-known member
I've got quiet songs on this CD - just voice and acoustic guitar. I've got mid-noisy songs on this CD - guitar and layers of voices and bass and beats. So how does one coordinate the levels of the different kinds of songs? When I do a mixdown of a song, the peaks are hitting about -2 dB. Should I keep it that way for every song? Should I 'turn down' the quiet songs relative to the noisy songs? If so, how much? I've never done an album with this much dynamic variety before. How do you do it?
 
I hope someone's got a better answer than me, but...by ear. Meters are nice and obviously necessary. But in the lonely struggle that is "mastering" for those that can't afford Mastering, you make decisions by sitting through several passes and listening.

Listening sometimes critically, as an audiophile, and sometimes distractedly, as a consumer. Preferably listening on two or more systems. Make notes on each pass, then compare them before you tweak. Trial and error for me, and I still wish I could afford a pro. But I've gotten close to right a time or two.

PS-I'll be watching this for some good answers.
-kent
 
Yeah, that's how I proceed for each song I do. But I'm asking about levels of songs relative to each other on the whole album. What - listen to the whole album a number of times? No - I don't want to do that. I've listened to this album enough it seems to me. Gimme some presets. :D
 
Yes, Dobro, that is what I mean. Sorry, no presets. The only way I've found that works to listen to it through in sequence. I first write out the song titles in order and leave space between each to note immediate impressions of the next song's level with respect to the previous. Does it jar me with a sudden blast? Or am I now straining to hear?

If I'm lucky, perhaps one track will stand out as being my ideal median for overall level. If so, I note this as well as how far the others are from it, meaning, do any seem too far above or below?
-kent
 
A full release should have a "mood" per se. Song order is part of that mood.

One song may end very loud. That ending may be the loudest passage on the whole CD! Comparing an ending with a start of the next song is not a good way of course to go about matching up levels if there is that kind of disparity going on. You would ruin the whole deal.

Maybe that loud ending SHOLDN'T be followed by a quiet ballad. Maybe it should. Certainly, neither way is unprecendented on a release.

I usually go to the loudest part of each song and compare levels there. That is about the only fair way to do it.

I have noticed that these days, rock productions go for most every song being as loud as possible all the time. Why people like that I will never know, but I certainly don't try to emulate it when I master a project. I have heard many great albums that had volume differences from song to song.

While I try to get the loudest passages of each song to match, I will make exceptions. If a ballad just doesn't sound good loud, I could care less how loud everything else on the CD is. I will keep that ballad quiet. I try to do what is best for each song first, then consider the whole project.

I have seen mastering done where each song was maximized to it's fullest (not over maximized, just taken to the point where making it any louder was detrimental), then all the average levels were matched by turning down the louder tracks. This is a valid approach.

The whole thing starts getting hard when you start trying to compare your master to stuff that was made REALLY LOUD. Then you will start trying to make everything loud too. If you don't play that game, you won't play it eh? :)

Another approach is to just make the whole thing flow well. I can assure you that when you listen from beginning to end that you will not say to yourself "Gee, the loudest part of this song isn't as loud as the loudest part of 2 songs ago". If the volume of endings and starts of songs NEEDS to sound smooth and cohesive, you will probably have to make other sacrifices.

Again, I have heard CD's that had a wide range of dynamics from song to song, and those dynamics were left in place. THERE IS NOTHING WRONG WITH DOING THIS AT ALL!!!! Just make sure that the CD flows well. It can be hard if you are doing your mastering yourself to be subjective to how the CD flows. You have a tendency to think about many other things, and start sweating crap like "Is this loud enough compared to other new products" that will screw it all up if you fall into that trap. Listen as objectively as you can. Listen as a listener, NOT as the producer/engineer/songwriter/performer. If you need to make some overall adjustments from song to song in regards to how loud it is based upon how it flows, well, use the master output of the editor to do so. Always save your plugin settings after mastering a song so that you can go right back to it and make an adjustment later if need be.

Good luck.

Ed
 
If you can listen to it comfortably without having to adjust the volume every song you are doing it right.
 
Kent - thanks for the input - that's what I'll have to do, then.

Ed - nah, I'm not trying to make my stuff as loud as possible or even make it sound like something else - that's a dumb game. I'll listen to the loudest part of each song, I'll set it up so it flows.

Tex - I think you summed it up nicely. :)

I've never had to do this before. Hadn't even thought about it. I've got a lot of ballads on this album, and I like them right up front when I listen to them, but it doesn't make much sense when the rocker that follows the ballad is at the same level - first it sounds like the singer is five feet in front of you, then it sounds like the singer's up on a stage forty feet away.

Guess I'm going to have to listen. ;)
 
Lots of good answers so far.

If I try to do this (and usually I let the mastering engineer worry about it) I usually focus on the lead vox and drums. A soft song SHOULD have lower meter readings than a louder song, that's why relying on meters is useless for this.

If your software allows it, it's good to be able to mark points in each song - like where the vocal verse starts, and the loudest section of the song. Then you can use the markers to jump from song to song and see if they sound in the same ballpark.
 
(and usually I let the mastering engineer worry about it)
Click to expand...

right on....Mister ME is the man for this sort of job...why worry about it yourself....you'll probably have to get your CD mastered anyway...let him do it..it's part of the package anyway:D :D
 
Yes, very droll, and echoing my sentiments exactly. That's the truth. But jokes aside, littledog's post was useful for me. That's the truth.
 
maybe you should write another bunch of Ballads.. and some more rock tunes, too.. and put all the ballads on one record.. and all the rockers on another!

hehe

xoxo
 
The best answer is from Tex on this one. :>)

On the other hand, if you are looking for a tool to help, Cool Edit Pro has a "group normalize" function that takes a group of wav files and can normalize them so that they have a pretty darn good level relative to each other. They use a pretty good psycho-acoustic model to make it work.

The good news is that you can try CEP for free - the demo is time limited (30 days iirc), not feature limited, except for MP3 saving.

-lee-
 
laptoppop said:

On the other hand, if you are looking for a tool to help, Cool Edit Pro has a "group normalize" function that takes a group of wav files and can normalize them so that they have a pretty darn good level relative to each other. They use a pretty good psycho-acoustic model to make it work.
Click to expand...

That's a pretty frightening thought. What if you recorded a ballad on a different day, and your mic levels were hotter than on another session? What a wonderful mastering job will result!

Shit, why don't we just poke out our eardrums with sharp sticks, since everybody's looking for mathematical shortcuts to good sound so they can avoid listening to the music. :(
 
Yo Dobro:

It's kind of frightening to a former English teacher, me, how badly folks articulate when doing vocals. Not only in music, but the other night I watched a late sci-fi drama and noticed that the lead male actor could be understood, every word. The gal who was doing the scene with him, I COULD NOT UNDERSTAND A FRICKING WORD SHE UTTERED.

So, I guess we are back to our ears?

I like to put my vocals UP front; I like them articulated which is a sign of a talent who has good mic skills, as well as a good voice.

I would love to be able to use the mics that Sinatra used in many of his very, very clear and articulated vocals with the band plenty loud.

I know that the industry tends to put the "beat" up front and most don't really care if the words are understandable. It reminds me of the time I heard a little girl singing as she walked with her mother. She was singing the old song, Buttons & Bows.

But, each time she got to those words, she said, quite clearly,

BUTTONS AND BONES.

Green Hornet:D :p :cool:
 
Yeah, I'm paying more attention to articulation these days too, especially the ends of words - I tend to miss some final consonants. But mostly I'm in agreement with Van Morrison on this one - the feeling's more important than clear enunciation. But if I can have both - yes, please!

But what I was wondering about out loud in this thread was more about loudness levels of songs relative to each other on a CD.

laptop, littledog: I'll check out that group normalize function in CEP- it seems to me it might be a useful place to start before actually beginning the listen-and-tweak work.
 
Hmmm

Perhaps a good starting point would be to record each song into two tracks of whatever multitracker you use, with each one running concurrently. Then you could quickly A/B them by muting/soloing/whatever to get an idea of what needs to be done to each one.

OB
 
Not a bad idea at all, OB!

One thing I might add -- especially if you are trying for commercial radio airplay -- include a couple extra tracks of popular well mixed songs in the same genre to match your levels to. You need to be sure your music can hold its own right next to a song by someone else.

-lee-
 
I hear Sonic Foundry is soon to release an updated CD Architec and is supposed to have very good tools for this purpose, so maybe that would be a possibility that I'll be checking out. I noticed something interesting the other day. Not new, but interesting. Someone sent me an mp3 via email that was at 128. OK, a bit of file crunching there, but it sounded good. It sounded loud too. I opened it in sound forge to have a look. The wave file was ugly. Dense as hell and spiking, but no distortion, just looked ugly. I burned it to CD twice. Once without touching it. The other I simply brought the volume down until the wave looked much more respectable. I knew this track would be lower on playback. When playing the CD, at first the loud track seemed the winner hands down (on my car stereo). No distortion and I didn't have to crank the volume. But then I started playing the lower volume track. True, I did have to increase the volume on playback, but not that much. Like if the volume was on 5, I had to push it up to 7 to get the mix as loud. But the sound of the track was much better dynamically. My whole system handled it better. More clear, crisp, and defined. And if I cranked it up to 8 it still sounded excellent. Not so with the loud track, which went to shit at higher volume, at least my system did anyway. I don't know much technically about all this, but it drove home to me the point that loud mostly sucks and there is a sacrifice. The lower volume track gave me much more flexability in regards to volume and eq on playback and always sounded great. So I'm quickly changing the way I think about this loud thing. I typically buck the system anyway!
 
Shockwave said:
But the sound of the track was much better dynamically. My whole system handled it better. More clear, crisp, and defined. And if I cranked it up to 8 it still sounded excellent
Click to expand...

If the tune was mixed that way in the first place, down to the stereo file you were trying to edit (or bring its volume down), I don't think it would have sounded better dynamically when you "pushed" the volume down (made the wave look "nicer")...if the dynamics were stuffed in the first place, and song compressed, I don't think you can change it just by bringing the volume of the mixdown down....doesn't matter how much "thinner" your wave looked, the dynamics would still be the same....(I think :confused: )
 
You must log in or register to reply here.
Back
Top