Mastering: Levels & Eq on Multiple Tracks Question

[Alex Y]

Hello guys,

Ok so I just finished my album, and I was hoping that there is an easy way to level multiple tracks up to the same volume. Same goes to eq'ing, I mean I got them all pretty much close, but is there is a way I can make them all sound in the same range?

Really hope someone helps!

Thanx!

[il7mago]

First if you think the high frequencies are too quiet, lower the low frequencies and vice veersa. Then you want to have roughly the same dynamic range across all songs. Then you want to make sure the peak of each track is near 0 db.

[Alex Y]

thanks for that one, is there any way you could explain that practice, the one with the vice versa part?

[il7mago]

Say you have a pad sound that you like, but you think the bass part is too quiet. You can add and EQ on it and either increase the bass or decrease the treble. It doesn't really matter as long as there is no clipping (going beyond 0db), especially on the master track.
So the best way to avoid clipping is to lower the treble. Then if the pad sounds quiet compared to the beat or other elements of the song, lower the gain of those elements too.

When all this is done, you might find that, on the master track, you never get close to 0db, but instead the highest it peaks at is -6.7 db.
You then add a utility effect on the master track and increase the gain until the song peaks near 0db. It's better to do that than to just normalize when exporting the audio, because you take full advantage of the bit depth of the file. (If it peaks at -12db, and you export as 16 bits, there will be 12 db (or 2 bits) of unused depth, essentially making it a 14 bit audio file, with the size of a 16 bits one (larger)).

[invol]

Hello guys,

Ok so I just finished my album, and I was hoping that there is an easy way to level multiple tracks up to the same volume. Same goes to eq'ing, I mean I got them all pretty much close, but is there is a way I can make them all sound in the same range?

Really hope someone helps!

Thanx!

EQ is key, but the question you ask is a lot more complicated than you might think. Volume is not a useful term at all - see links below. You are really talking about Equal Loudness, which includes many subjective characteristics of perception and variables of monitoring system. Assuming you have great speakers in a well designed acoustic space, it still take time to get good at "leveling" - especially for it to translate across multiple systems.

Here is a good start.: http://en.wikipedia.org/wiki/Equal-loudness_contour

Then here - Chapter 5 from Mastering Audio by Bob Katz
http://www.focalpress.com/uploadedFiles ... 376_cs.pdf

Then here: Bob Katz Article on K-System http://www.aes.org/technical/documentDo ... m?docID=65

This is also a very interesting article by John Chowning -
http://profs.sci.univr.it/~dafx/Final-P ... owning.pdf

Good luck and have fun : )

Cheers,
Brian

[mharris]

You then add a utility effect on the master track and increase the gain until the song peaks near 0db. It's better to do that than to just normalize when exporting the audio, because you take full advantage of the bit depth of the file.

Is this really the case..? What magic trickery can the Utility perform when boosting the gain that a normailzation process cannot?

(Sorry to hijack the thread btw.)

[Alex Y]

invol, very good read indeed, thank you! mharris, it depends on how you record stuff i guess, i over compress everything, so that makes me record a song in 3 phases, and of course i encounter loss of quality.

[il7mago]

You then add a utility effect on the master track and increase the gain until the song peaks near 0db. It's better to do that than to just normalize when exporting the audio, because you take full advantage of the bit depth of the file.

Is this really the case..? What magic trickery can the Utility perform when boosting the gain that a normailzation process cannot?

(Sorry to hijack the thread btw.)

The question is whether the normalization process happen before or after everything is rendered. I don't know the answer so I chose to the safe way.

[invol]

Normalization would occur after it is rendered, otherwise, how could it analyze a peak.

FYI, Normalizing will not help with the loudness leveling he is talking about, only the PEAK level. A good brick wall limiter will handle this better for you during mastering anyway.

Normalizing is for sound design, not mastering.

[il7mago]

Normalization would occur after it is rendered

If that's the case, then it's always better to normalize manually before rendering. That way you maximize effective bit depth and don't waste file size with unused db levels.

[invol]

Normalization would occur after it is rendered

If that's the case, then it's always better to normalize manually before rendering. That way you maximize effective bit depth and don't waste file size with unused db levels.

There is no need to "maximize" bit depth anymore. That is left over from the old 16 bit world of the 90's. Normalizing is not necessary, ever, when mastering.

edit: p.s. All bits are used no matter what the signal level. When converting from A to D with old 16 bit systems there was a trend to record "hot" to maximize signal to noise, not to "maximize bits". Also, file size is not going to change unless you change the sample rate and bit depth settings, size has nothing to do with the info in the file when talking about audio - e.g.,16 bit, 44.1 kHz, mono is 5 MB even if silent.

[invol]

Normalization would occur after it is rendered

If that's the case, then it's always better to normalize manually before rendering. That way you maximize effective bit depth and don't waste file size with unused db levels.

Everyone should read this.

http://www.hometracked.com/2008/04/20/1 ... alization/

[il7mago]

This talks about the fact that normalizing a digital recording does not affect dynamic range. Of course it doesn't, because it cannot invent parts of the signal that weren't already present!
However we're debating two different things here. With a live set that hasn't been rendered yet, it's entirely different, much like when you record from analog to digital.

I think you should read this instead http://tarekith.com/assets/Leveling.html

[Mage2k]

First if you think the high frequencies are too loud, lower the low frequencies and vice versa. Then you want to have roughly the same dynamic range across all songs. Then you want to make sure the peak of each track is near 0 db.

Er, don't you mean, "if you think the high frequencies are too *quiet*"? The idea being that, rather than raising the highs, you lower the bass or mids.

[il7mago]

First if you think the high frequencies are too loud, lower the low frequencies and vice versa. Then you want to have roughly the same dynamic range across all songs. Then you want to make sure the peak of each track is near 0 db.

Er, don't you mean, "if you think the high frequencies are too *quiet*"? The idea being that, rather than raising the highs, you lower the bass or mids.

Exactly. Thanks for noticing it!