Fixing variable sound level (DJ Mix)

[steve]

This is getting seriously surreal. Is there any reason we're not using the multi-band compressor built into Audacity 1.3.5?

Effect > Apple: AUMultibandCompressor

That's the trouble with posting in a general forum section like this rather than a Windows/Mac/Linux board. The AUMultibandCompressor is only available on Macs (proprietary Apple effect).

[kozikowski]

<<<The AUMultibandCompressor is only available on Macs (proprietary Apple effect).>>>

Oh. So you're using the wrong machine.

Koz

[steve]

Oh. So you're using the wrong machine.

Koz

I'm not, it's you two
(and I bet DJCraft is on Windows)

[kozikowski]

I still want that sample track.

The blue waves are nice, but I can't mess with them--other than Photoshop.

Koz

[DJCraft]

The blue waves are nice, but I can't mess with them--other than Photoshop.

I'm almost to the point where it feels like I'm editing in photoshop, j/k

Here's a few minutes cut up http://arkive.us/MixCutup.wav

[kozikowski]

Excellent. That will be fun to play with when I get a second.
Koz

[steve]

Chris's Dynamic Compressor works pretty well on that. I used the following settings:
Compression Ratio: 1.0
Maximum Amplitude: 0.8
Everything else at the defaults.

If you want to turn it up to 11, use the fast lookahead limiter after the compressor with these settings:
Input gain: 2.0
Limit: -0.3
Release time: 0.1

[kozikowski]

I did notice one thing on a first listen. There are portions of the presentation that have significant energy in the 3-5KHz range (bright and forward) and this throws a hitch into things. No matter how much you compress and squash the waveforms, those segments will always sound louder. This, in fact, may be one of the things the poster is struggling with without entirely realizing it.

That dumps us immediately into the question of how much changing are we allowed to do before we start to step on the artist's intentions.

Koz

[DJCraft]

Chris's Dynamic Compressor works pretty well on that. I used the following settings:
Compression Ratio: 1.0
Maximum Amplitude: 0.8
Everything else at the defaults.

If you want to turn it up to 11, use the fast lookahead limiter after the compressor with these settings:
Input gain: 2.0
Limit: -0.3
Release time: 0.1

Thanks, I'll try both out and give it a listen

I did notice one thing on a first listen. There are portions of the presentation that have significant energy in the 3-5KHz range (bright and forward) and this throws a hitch into things. No matter how much you compress and squash the waveforms, those segments will always sound louder. This, in fact, may be one of the things the poster is struggling with without entirely realizing it.

That dumps us immediately into the question of how much changing are we allowed to do before we start to step on the artist's intentions.

Koz

Would you point out in the timeline in the sample you're hearing this? It's just cut up with the levels that came straight from the live mix. There are certain points where songs are quite and I'd like to keep them that way. I'm really just wanting to bring all the loud spikes and louder songs down to the level of the average song on the mix. Currently I've been doing this in my spare time to see if it helps:

View -> Show Clipping
Ctrl+A -> Effect -> Normalize -> Remove any DC offset, Normalize maximum amplitude to 0.0
Envelope tool to average out the clipping to the waveforms next to it
Repeat.

I've done about 20 passes with this so far and I can see some of the clipping gone but it looks like there are a ton more. I guess priority wise what I want to do is this:

1) Lower obvious peaks/clipping without sacrificing (by rounding off) the sound quality
2) Average the sound level out between songs so the mix as a total is more average while leaving any quiet spots in songs quiet

I heard something about an L1 Ultra Maximizer (look-ahead limiter?) would that be what I'm looking for?