Mastering full album, making songs the same volume?

[twister]

There have been a lot of discussions here about mastering songs, but I have one about mastering album and making sure all the songs are at the same level. After mastering my songs, I run Statistics analysis in Sound Forge on the loudest part of the hook to get RMS Level. I do that on one song and use it as a reference. While mastering other songs, I use iZotope Ozone, I play with different settings of Loudness Maximizer threshold and than check it again with Statistics analysis to make sure I get RMS Level within 0.3dB of my reference. So in theory it all should be close in overall level. But when I play these songs in a row in my car I can clearly hear volume variation that sounds to my ears like 1-2dB in difference.

Is there a trick you guy use to even out the sound? I never had to deal with this before because all my productions in the past were pre-mastered with plenty of overhead room and no mastering EQ - basic requirement of any record label that likes to master their own stuff for compilations. Now, I'm trying to figure out how to do it myself. Any suggestions? Tricks? Thanks!!!

[C00kie]

Yes, the trick is to make them equally loud according to your ears.
So make notes which are too soft and which are too loud.
Maybe the loudness now differs because you took the loudest part as reference, and not an average part.

[TankEyes]

using some kind of average or RMS volume data can help too. A/b them and use your ears And don't worry about it too much. Don't sqaush stuff just to match loudness.

[twister]

The problem is that I don't trust my ears What I'm afraid is that a song which is more bass driven will not sound as loud as the other song which has more high frequency content. That just goes along with that famous frequency curve they have in all mastering books. Our ears can't hear lows well enough as highs. So, I'm relying more on numbers of RMS Level. I might try doing analysis of the whole track to get average RMS but since I produce dance music and songs have quieter intro and also breakdowns, I'm afraid it will throw the average off. Someone mentioned to Normalize the song first to scale up everything with a highest peak to 0dB, and than apply mastering setting. Is that a good idea? Of course, I'm trying to be very careful not to squash the track and ruin dynamics. Didn't realize its going to be such a PITA to get everything leveled off and loud at the same time.

[thecontrolcentre]

Someone mentioned to Normalize the song first to scale up everything with a highest peak to 0dB, and than apply mastering setting. Is that a good idea?

No ... bad idea. Use your ears to judge which of the tunes need to be louder or quieter. Load them all into a WAV editor and listen and compare ... it's the only way to get an overall balanced album. If you don't trust your own ears, then send the project to a mastering engineer.

[mauseoleum]

RMS analysis will only get you in the ballpark IF the songs have very similar frequency content and are also very similar in their dynamic structure.

I had to mix/match various snippets in multimedia context (years ago, commercial presentations) which was a different proces tcompared to mastering an album, but even before the latest outbreak of loudnes war the program material intended for small speaker playback was squashed beyond belief, with heavily eq-ed parallel processing etc to make sure there wasnt any unwanted dips or distracting transients on playback etc.

What I wanted to say is that you need a program material that is already extremely coherent, very similar, to start with, if you want to use automatic tools AND don't want to kill the songs with extreme processing.

[eduardo_b]

The problem is that I don't trust my ears What I'm afraid is that a song which is more bass driven will not sound as loud as the other song which has more high frequency content.

I think you actually can trust your ears. It's fairly obvious when some tracks are louder or softer than others. There's nothing wrong with using rms metering to confirm what you're hearing. Keep in mind that mastering engineers will EQ vocals or bass if they have to to even out both volume and tone issues. That's why mixing engineers don't do mastering, which requires a different approach and a focus on the details you're concerned with. It takes considerable skill to do mastering well.

[twister]

I guess in our day'n'age of most of the music producers being "jack of all trades", its inevitable to wear different hats Since I do everything from scratch anyway, I spend a lot of the time on mixing to get track dynamics sorted out. Alright, will trust my own ears I'm thinking that if I hear one track being louder than another, and even if I get inaccurate average RMS Level reading of the whole track - all that matters after mastering level reduction/increase to see a relative level change according to my ears. Forgot to mention that unfortunately all my music production at home is done with Sony MDR7506 cans (2 little kids under age of 5 at home, can't blast anything on speakers), so my car is the only true listening environment.

[eduardo_b]

I guess in our day'n'age of most of the music producers being "jack of all trades", its inevitable to wear different hats Since I do everything from scratch anyway, I spend a lot of the time on mixing to get track dynamics sorted out. Alright, will trust my own ears I'm thinking that if I hear one track being louder than another, and even if I get inaccurate average RMS Level reading of the whole track - all that matters after mastering level reduction/increase to see a relative level change according to my ears. Forgot to mention that unfortunately all my music production at home is done with Sony MDR7506 cans (2 little kids under age of 5 at home, can't blast anything on speakers), so my car is the only true listening environment.

I agree that bass can be an issue. I've heard albums with some tracks having far more bass than others even though all other frequencies were the same as the other tracks. The mastering engineer could (should) have cut the bass to rebalance the tonal qualities.

The Sony 7506s have strong bass, so that might have an effect when listening on other systems. The car system is not necessarily flat in response given the acoustics and noise of that environment, so having another music system would be very helpful. I think you might try loading up all the tracks in your host or editor and switch back and forth among them to detect any significant differences in perceived loudness. Keep in mind that a change of 3 dB will double of halve the sound level.

[Kim Lajoie]

It's not enough to adjust the level - you must also adjust the tone.

Note that I used the term 'adjust' - not 'match'. Like matching level, sometimes it's not appropriate to match the tone exactly to your reference tracks. There is some judgement that must be exercised.

For an overview of a more complete mastering processing, see here:

http://www.kvraudio.com/forum/viewtopic.php?p=3305233

-Kim.

[eduardo_b]

It's not enough to adjust the level - you must also adjust the tone.

I don't think I emphasized that sufficiently. It really is as important as the level. It's why there are mastering engineers...well...one of the reasons.

[donkey tugger]

Yes, the trick is to make them equally loud according to your ears.
So make notes which are too soft and which are too loud.
Maybe the loudness now differs because you took the loudest part as reference, and not an average part.

Best advice there. If you're just a wee hobbyist (like me) and you can't afford any fancy mastering engineer, listen to your songs over and over again comparing loud bits and quiet bits, fast songs and slow songs, songs in the same key and different keys and after a bit you will no doubt hate them all. After that you might have an idea of the running order. Then it's time to balance up the overall volume levels by a bit of judicious limiting/eq here and there, perhaps reducing the overall level for other songs.

One of the best parts of doing an 'album' though, for me anyway; you've got all yer songs then you have to present em.

[synthgeek]

The only "trick" I use is this: I put all the tracks on in order, as if I were listening to the cd, and I set the volume during the first track to a point I consider loud enough to hear the details and feel it, but not actually "loud". Then I just listen. If I find myself feeling the urge to reach for the volume knob, I make a note of which track and when, and skip to the next one (in case there is a real difference, so I don't get used to that level). Then I go back through the stuff I marked to see if & where the problems are.

[donkey tugger]

The only "trick" I use is this: I put all the tracks on in order, as if I were listening to the cd, and I set the volume during the first track to a point I consider loud enough to hear the details and feel it, but not actually "loud". Then I just listen. If I find myself feeling the urge to reach for the volume knob, I make a note of which track and when, and skip to the next one (in case there is a real difference, so I don't get used to that level). Then I go back through the stuff I marked to see if & where the problems are.

I like it!

[eduardo_b]

The only "trick" I use is this: I put all the tracks on in order, as if I were listening to the cd, and I set the volume during the first track to a point I consider loud enough to hear the details and feel it, but not actually "loud". Then I just listen. If I find myself feeling the urge to reach for the volume knob, I make a note of which track and when, and skip to the next one (in case there is a real difference, so I don't get used to that level). Then I go back through the stuff I marked to see if & where the problems are.

Excellent! And so obvious...not!