How to Master Super Loud: Peak Shaving Technique for Professional Results
I'm going to show you a secret source used by professional mastering engineers that lets you make tracks super loud without compromising the mix. This is the technique I use to get competitive loudness on every master. The results at the end of this process will impress you, and it starts with understanding why removing stray peaks actually adds loudness overall.
The Peak Shaving Foundation
Peaks are the first thing that hits the threshold on any compressor or limiter in your mastering chain. By removing those stray peaks without affecting the overall sound, you free up headroom. That headroom becomes usable gain. You can push the track another 2 to 3 dB louder without triggering the limiter on the body of the mix. This is how you get master loudness without obvious compression.
The technique is simple but requires precision. Export your finished mix as stems. In my case I have a bass bus, drum bus, effects bus, instruments bus, and vocals. Each stem becomes a track in a fresh mastering session. On each one, I place a FabFilter Pro-L2 limiter and configure it as a clipper.
Turning a Limiter Into a Clipper
The settings matter here. Attack fully round to the right. Turn true peak off. Release super fast. Channel linking at zero. Output at zero. Oversampling at 16 times or as much as your computer handles. The all button should be aggressive. These settings turn the limiter into a clipper that shaves peaks without affecting the body of the mix.
The crucial setting is the one-to-one button on the Pro-L2. This is sometimes called unity gain or gain match on other plugins. What it does is keep the output the same as the input. If you apply gain to the input, the output stays the same level. The limiter just shaves off peaks without fooling your ears with level tricks. You hear what the plugin is actually doing, not just the loudness increase.
Finding the Sweet Spot
Start with the loudest section of your mix and play it through the mastering session. Gradually add gain to the limiter on that stem. Watch the meter to see how much peak reduction is happening. I'm usually aiming for 5 to 6 dB of peak reduction on individual stems. If I go too far, the kick starts distorting. That's the signal to back off.
With the one-to-one button on, you're hearing the limiter work without level fooling you. You add 12 dB of gain but the output stays the same. You hear the sound getting slightly compressed but not obviously so. That's the goal. Shave the peaks without changing the tone or character of the mix.
Processing Every Stem
Once you've set the gain on the first stem, repeat the process on every other stem. You're not changing the sound. You're just removing stray peaks from each element. When you combine them back on your stereo output, something beautiful happens. All the peaks are shaved off so the mix bounces cleaner through your mastering limiter.
The meter on your output shows you the before and after. The mix sounds exactly the same as it came in but the meter shows 2 to 3 dB less energy. That's your gain. That's 2 to 3 dB of extra headroom you just created without changing the sound.
Gaining Extra Loudness in Mastering
Now that you've freed up 2 to 3 dB by removing peaks, your actual mastering chain has more headroom to work with. When your mix hits the mastering limiter, it's not getting caught by stray peaks anymore. The limiter responds to the body of the mix instead. That means you can push another 2 to 3 dB louder than anyone else mastering the same session.
This is the difference between a master that just fits loudness standards and one that stands out. You're getting 3 dB louder than the competition purely through smart peak management, not obvious compression. The mix still sounds like a mix. The dynamics still breathe. But the perceived loudness is competitive.
Why This Works
Peaks are noise from an engineer's perspective. They're not the actual music content. They're stray transients, clicks, moments where the sum goes a bit too high. By removing them cleanly without affecting the rest of the content, you're not changing the mix. You're cleaning it up. You're making it more efficient for processing.
This is especially powerful on stems because each one has its own stray peaks. The kick might peak 2 dB higher than the body. The vocal might have a click at a specific moment. By processing stems individually before recombining them, you catch all these issues. The final stereo file is cleaner and more transparent, which means it can be louder without hitting hard.
