Normalizing the Mix

[tonemangler]

Hello,

I am currently recording and mixing a bunch of my tracks for a CD. This is a hobby or mine but when they're done I would still like to have them professionally mastered. As each track is completed I've tried mastering myself to hear them on different systems. After mixing the track I have a 24 bit 44.1 KHz stereo file. Then to get it loud I normalized %100. Then fed thru a compressor-eq-limiter. I know the processors will depend on the track, but my question is should I be normalizing at all? I've read on other posts that Mastering Engineers like to receive mixes with ample headroom, am I limiting the effectiveness of the mastering processors by normalizing first?

Thanks,

Paul

[Jae B]

No, you should not normalize.

Get the mix levels worked out with the master volume down at around -3 to -6 db to ensure adequate headroom. Then on your master bus put your compressor, eq, limiter. With these plugs active, and sounding good, beginning bringing the volume up to 0db. (this is where your limiter really comes into play). This will give you an idea of how the finished mastered track(s) will sound. If all sounds go, bypass the plug-ins on your master bus and bounce down the file to stereo (again, make sure your at least down -3db to give the mastering engineer space to play with).

[James Perrett]

If you are going to be doing further processing then there is absolutely no point in normalising as most processing will work just as well, if not better, on non normalised material. If you normalise to full scale, there is a chance that further processing may cause clipping.

James.

[Dave Blackman]

Don't normalise. You're adding another process where you don't need to, certainly if you're then going into a limiter - bring the level up there if need be.

Also, as James mentioned, if you're normalising at 100% you may very well be introducing clipping. Which is bad.

Dave

[Exalted Wombat]

Hello,

I am currently recording and mixing a bunch of my tracks for a CD. This is a hobby or mine but when they're done I would still like to have them professionally mastered. As each track is completed I've tried mastering myself to hear them on different systems. After mixing the track I have a 24 bit 44.1 KHz stereo file. Then to get it loud I normalized %100. Then fed thru a compressor-eq-limiter. I know the processors will depend on the track, but my question is should I be normalizing at all? I've read on other posts that Mastering Engineers like to receive mixes with ample headroom, am I limiting the effectiveness of the mastering processors by normalizing first?

Thanks,

Paul

You're doing it backwards. Compress/limit to make it loud, normalise (if needed) afterwards. Otherwise you're processing in an environment with no headroom.

Don't forget to send the mastering engineer the mix BEFORE you do all this stuff. Then do your thing, andyou can compare his attempt at mastering with yours.

[narcoman]

Do not normalise.
Do not touch a limiter.

Leave it for the mastering chap.

[tonemangler]

Thanks for all the advice. I learned two things: #1 Do not normalise during mastering #2 Normalise is spelt with an 's' when dealing with audio.

Usually while tracking I will normalise takes at around %70 to get a healthy waveform but maybe this also is not beneficial, though I haven't really noticed a problem with my mixes.

Thanks again,

Paul

[Moses]

Naw, we're just crazy Brits, who spell stuff right!

Mo

[narcoman]

Thanks for all the advice. I learned two things: #1 Do not normalise during mastering #2 Normalise is spelt with an 's' when dealing with audio.

Usually while tracking I will normalise takes at around %70 to get a healthy waveform but maybe this also is not beneficial, though I haven't really noticed a problem with my mixes.

Thanks again,

Paul

You're Canadian. Use an "s"

There isn't any advantage to normalise anything. YOu just have the same s/n ratio but with a different set of numbers. HOWEVER - there is a case for a gain of poor recording technique so that plugins get an appropriate level for optimal operation. Something like the Softtube Trident plugin wants a level in accordance with (something like) -18 dB(FS) being equivalent to 0dB (VU).

In this particular instance a normalisation could be beneficial - but more beneficial would be recording it "better" in the first place!