When mastering for streaming, you'll need to keep loudness normalization in mind, as well as how each streaming service utilizes normalization is a slightly different way. When in doubt, master between -14 and -10 LUFS and set your output level to about -1dBTP to avoid clipping during encoding.
When mastering for a streaming service, loudness normalization is an important topic - since it’ll cause the amplitude of your track to potentially be altered during playback. Most services utilize downward normalization, meaning your track is turned down if it goes above a pre-determined LUFS.
I’ll create a -14 LUFS signal, and a -8 LUFS signal. Then I’ll use -6dB of clip gain on the louder track to match them and simulate normalization. Let’s see if we can hear a difference between the 2.
Different streaming services like Apple, Spotify, Tidal, YouTube, and more normalize audio to different LUFS settings, and may utilize upward and downward normalization, or just downward. For example YouTube only uses downward normalization, and will only turn audio down if it’s too loud, not up if it’s too quiet.
Just about all of these normalize to -14 LUFS. Let’s listen to a signal pre and post normalization.
Another thing to consider is how encoding, or the translation of a lossless file to a lossy file changes the amplitude. When the resolution gets reduced, so too does the accuracy of the amplitude - causing clipping distortion; to avoid this it’s recommended you set your ceiling to -2dBTP.
I’ll simulate how a track would clip after encoding, so you can hear the effects.
When you’re trying to avoid clipping distortion after encoding, like when mastering for streaming, oversampling can be helpful, since it’ll reduce the amount of clipping distortion. To avoid clipping distortion post encoding, I’d recommend lowering the ceiling of the limiter, using a little lookahead, and 4x or more oversampling.
Let’s listen to a pre and post-oversampled signal to see if we notice a difference.
Although you may be using a true-peak meter to measure the level of your signal, this is different than true peak limiting. True peak limiting utilizes true peak metering to determine where clipping would occur, but then compensates for these peaks, sometimes to the detriment of your transients.
With that in mind, it may be better to simply lower your limiters ceiling and use oversampling instead.
Let’s compare the two.
Although more dynamic is typically thought of as better, when the audio is normalized, this can become an issue. For example, if your LUFS is -18, and you have a -2dB peak, your signal will be increased by about 4dB, causing the streaming service to utilize a limiter.
This situation isn’t ideal, since you can control how that limiter will alter your signal.
Let’s take a listen to a track with these levels for reference.
Many engineers ask, why go above -14 LUFS if it’ll be normalized? There are a lot of answers to this but in short, you never know when normalization settings will change, the timbre of a louder master is different and can be beneficial, and some listeners turn normalization off.
Furthermore, some services don’t normalize so you’ll need to keep that in mind.
With this in mind, let’s listen again to the -14 LUFS and normalized -8 LUFS master to see if we hear a difference in their timbres.
Although target loudnesses should typically be avoided, some limiters have target loudness functions, which can be convenient if you want to preview how that loudness will sound. Izotope’s Maximizer and TC Electronic’s brick wall limiter are examples of these, but more will probably be released in the near future.
Let’s listen to a signal being normalized by the Maximizer and TC Electronic’s brick wall limiter.
If you want your master to sound aggressive like a loud master, but you don’t want normalization to affect your peaks, utilize low-level compression. This will increase the detail of your track from the bottom up, leaving your peaks alone, and somewhat negating the need to master loudly.
Let’s listen to a master with low-level compression at -14 LUFS, and one without low-level compression at -14 LUFS.
Let’s face it — trying to create professional mixes without the correct knowledge, tools, services, and network is like playing a concert with a broken instrument to an audience of no one.
That’s where the Sage Audio membership comes in — the membership gives you everything you need to create professional mixes consistently.
The membership is the secret to your success — it’s the difference between the 1,929,999,999 of poor-sounding mixes and the .00000002% of mixes that make people say 'WOW!"
Yes, I want to join the membership now >
Yes, I want to join the membership now >
Yes, I want to join the membership now >
Yes, I want to join the membership now >
Yes, I want to join the membership now >
Yes, I want to join the membership now >
