Part 1. How to set the limiters ceiling for the absolute peak of loudness…
and how to push that limiter for the average loudness to make sure that the song won’t be too quiet or too loud on the services, where people will be listening to it. These are the questions that worry and stress a lot of people, but we’re going to settle all of the myths. And for that we won’t give you any rules to follow. Instead, we’ll tell you exactly what to do.
Almost any DAW app has its own tools that function pretty much the same way. Of course, they may sound a bit different, but what is incredibly important, all of them have a free loudness meter. We’ll be using the Ozone Limiter.
The first thing on the limiter is the ceiling, which also might be called ‘max output’ or, actually, anything else, depending on the limiter type you use. People usually set it to minus 0.1 or minus 0.3, and their decision is based on absolutely nothing at all. Instead, you should just go straight to Spotify and check their FAQ, where you’ll find how to upload the tracks and what they’re going to do with it.
- If you upload a WAV file, they’re going to transcode it into a different delivery format, depending on the device and location of the listener. This encoding process needs to be done as successfully as possible to make sure there’s no distortions or any other possible issues.
- One more thing to take into consideration is that Spotify has loudness normalization, which means that if you upload a loud track, it will just turn it down, and if you upload a quiet track, it will turn it up.
If you don’t want your track to be turned down, push it much louder than this. But in that case, you should have an integrated LUFS value (loudness measurement of -14DB).
As you can see, people that just set it to minus 0.3 or minus 0.1 are those who don’t quietly understand what they’re doing, or maybe they’re using advanced techniques so that the track distortion is minimized. But mostly, we’re not that advanced. That’s why it is much better to start at minus 1DB (at least).
Engage true peaks. Say, I take the ceiling back up to minus 0.3DB, you’d think that it would limit it there. But let’s look at this loudness meter, specifically the short-term max. If the true peak max is 1.1DB above zero, then there’s not anything we can do about the fact that it’s going to clip. If I engage true peak, the loudness will be much closer to the actual value the limiter says it will be. So, the first thing you should do is take that down to minus 1DB.
This loudness meter gives you all kinds of analysis for the short-term, integrated loudness, as well as dynamic range. There’re presets for Spotify and other services, including YouTube. And the average at minus 14DB will be already labeled there for you so that whenever there’s a peak higher than Spotify recommends, the red line will appear.
But even though we set it to minus 1, it’s actually a true peak of minus 0.9. So, we recommend to push the ceiling down just a little bit more, to minus 1.1 DB.
If your mix doesn’t sound good at minus 1.1, pushing it up to minus 0.3 will not suddenly make it sound good. You definitely shouldn’t push the ceiling higher, hoping that a listener will be more impressed by the track. Spotify will set the loudness so that it matches everything else.
Part 2. How much should I push the limit to make sure that the track loudness is correct?
Let’s use the Spotify guidelines and a free tool called the loudness penalty analyzer. According to them, the loudness normalization is applied to all of the tracks. Make sure that listeners won’t end up hearing from one track being too loud and the next one being too quiet. If the track is quieter than -14DB, they’re going to apply their own limiting.
LUFS are the loudness units relative to full scale. It is just another way to measure the loudness. This would be turned down quite a bit on Spotify. Set the limiter to the point where it’s pushing the minus 14. And then once you’ve exported the file, go on to loudness penalty analyzer. It will show you how much your track will be turned down. This is why it’s called a loudness penalty. These are relative values, but they’re pretty accurate. Such apps may be kind of an analysis tool for you.
After exporting the needed version of track, just drag it into the DAW and send it to your mixer. Say, the track would be turned down to 5.3DB. Then, just turn it down by minus 5.3 And now if you turn off all mastering effects, you’ll be able to listen to it the way it would be on Spotify. Once you get up to that loudness, you can just explore lots of different versions, then match their level and then just see which one you actually prefer the most.
Overall, be sure to check with Spotify, with YouTube, keep checking before you master something and see what they’re actually asking for.
All these standards changed a lot over the last six years, and they’re likely to keep changing. There are plenty of ways to affect the perceived loudness of mixes. Most of the time when you hear a song on Spotify, you just think it sounds so loud. The loudness actually comes from the mixing.
Here’re the things that you should be aware of:
1) tonal balance, which is the distribution of the energies and frequencies in the mix. If it has more low end, less high end, it’s going to sound quieter. By adjusting ше, you can make the perceived loudness being increased quite a lot.
2) macro dynamics. Making a verse quieter than a chorus will make the chorus feel much louder.
3) vocals. When the track is turned up, all the instruments become louder as well as the track itself. Although it is something to experiment with, be careful not to make things worse.
4) mix saturation. Adding tube saturation within the mixing phase, or in mastering itself, subtle saturation and distortion throughout the whole track can make it a lot more harmonically rich.
Try not to stress out about it. Just remember that loudness normalization rules help to make your track sound better. They don’t aim to discriminate against you as an artist.