LUFS: How loud to master songs for streaming (& why it’s not -14 LUFS)

How loud should my master be?

The loudness of tracks is an area of mastering which has evolved a lot over the years. In this lesson, we’ll discuss the issues surrounding loudness in mastering in the era of streaming. We’ll cover the loudness war, loudness normalization, and measuring loudness in LUFS. We’ll also look at what LUFS to aim for. Plus, we’ll talk about why, contrary to popular belief, you shouldn’t be aiming for -14 LUFS.

Before we go any further, if you want to make better music in your home studio, then be sure to download your free home studio bundle here (EQ cheat sheet, compression cheat sheet, & vocal guide).

Loudness normalization

A key factor that has changed the way we master music in recent years is loudness normalization. Since the late 2010s/early 2020s, streaming services have used loudness normalization to manage the playback level of the songs on their platforms.

Loudness normalization in streaming is important. That’s because, at the touch of a button, you can now listen to tracks from any genre, and any decade. This poses a problem. That’s because songs from different genres and decades have been mastered to greatly varying levels of loudness. Say you’re streaming a song from the ’60s, followed by a song from the 2000s. The 2000s track would likely playback substantially louder.

As such, streaming platforms apply gain adjustments so that songs play back at a consistent level. That way, listeners don’t have to adjust the volume at the beginning of every song. Nor will they damage their hearing if the song that they had already turned up fairly loud is followed by a track that was mastered significantly louder.

Many cite the use of loudness normalization in streaming as something which has ended ‘the loudness war‘…

The loudness war

The loudness war as we know it today was a trend that started around the mid-’90s, whereby tracks were mastered at increasingly loud levels. This loudness often came at the cost of both dynamic range and sound quality.

Tracks were produced loud in an effort to compete with all of the other tracks that were also being mastered loud. Typically, we tend to think that louder music sounds better. The problem is that the more you apply compression/limiting to achieve such loudness, the more you compromise the dynamics.

Since the advent of loudness normalization in streaming, however, this has become less of a problem.

Is the loudness war over?

Because of loudness normalization in streaming, it’s no longer possible for someone to try and make their songs louder than everyone else’s. Even if someone did master their songs really loud, their efforts would be in vain. As the streaming platforms would simply turn the track down to play at the same level as everything else. The benefit that this gives you as a music maker is that you no longer have to compromise the dynamic range of your songs to ensure that they’re not quieter than others.

How much compression/limiting you apply is a subjective decision. It’s worth keeping in mind though, that if you compare a dynamic song to a heavily compressed song, both played at the same level, the compressed song is likely to sound smaller or weaker in comparison to the song with a greater dynamic range.

So how do streaming platforms determine how much they need to alter the level of songs by…

What is the LUFS standard?

LUFS (loudness units full scale), were developed by the International Telecommunications Union to create a standardized way of measuring loudness.

Meeting specific LUFS values is now a requirement in TV and radio broadcast, as per the EBU R128 standard in Europe and the ATSC A/85 standard in the USA.

LUFS are also used by streaming services to measure the loudness of songs. Each platform has a target LUFS value for its playback level. So if your music is louder than its target, it will lower the level of your music to meet the target. On many platforms, though not all, your music will also be turned up if it is lower than the target.

Target values have changed over the last few years. There was also a time at which a few different loudness measurements were being used other than LUFS. We’re now at a point, however, where the vast majority of streaming platforms are using LUFS, and most are using a target level of -14 LUFS integrated.

LUFS meters

Because LUFS are a standardized measurement of loudness and are used by streaming platforms, it’s highly beneficial to incorporate LUFS measurements into your mastering process. Many DAWs now come with LUFS meters built-in. But if yours doesn’t, you can download a great free loudness meter from Youlean.

LUFS can provide you with a number of different measurements. The three main measurements are ‘integrated LUFS’, ‘short-term LUFS’, and ‘momentary LUFS’.

Integrated LUFS are a measurement taken over a period of time. This is usually used to measure the whole song from beginning to end to give you the song’s overall loudness value. It’s a measurement of how loud the peaks are compared to the average level. So the more dynamic the track, the lower the number. The less dynamic the track, the higher the number.

Short-term LUFS are a measurement over the last 3 seconds. Momentary LUFS are a measurement over the last 400ms.

What LUFS to master to

A common question is this: what LUFS should I master to? At present, there is a lot of advice online stating that you should master your songs to -14 LUFS integrated. The rationale behind this is that -14 LUFS integrated is the target value of most streaming sites. But this isn’t necessary. You’re not being asked to meet this target value. That value is the streaming platform’s target. So they will alter the level of your music to meet their target. There are also two other reasons why simply mastering to -14 LUFS integrated is actually quite a bad idea.

The problem with mastering to -14 LUFS

Firstly, there’s nothing to say that -14 LUFS integrated will make your songs sound their best. -14 LUFS integrated is a pretty conservative level. So it may not be suitable for the kind of sound that you want to achieve from your songs.

The amount of dynamic range reduction applied to a song can be as much of a stylistic decision as it is a means of achieving loudness. You might prefer the kind of compression/limiting found in tracks that are mastered louder, like a rock or pop track with an integrated loudness of -10 LUFS, or a hip hop track with an integrated loudness of -8 LUFS. If so, you’re unlikely to be happy with the sound of your tracks when mastered to -14 LUFS integrated.

Of course, many would say that heavily compressed songs should be avoided in favor of dynamics. But at the same time, many producers are still mastering very loud tracks. They know their songs will be turned down, but they do so because they like the sound. So whether you retain the dynamics, or you really squeeze your tracks, it’s important to do what you think sounds best for the music. Doing this is much more important than simply trying to make your song playback at a particular integrated LUFS value.

Integrated loudness

The second problem with aiming for -14 LUFS integrated is that integrated LUFS gives an average measurement for the whole song. The problem here is that the content of songs can vary greatly. On an album, you might have some songs with consistent levels between each section. But there might also be songs that were intended to be more varied, with verses that are quieter than the choruses. Then there might be a song that starts quietly and builds throughout the song to one final loud chorus.

The logical thing to do here would be to master these tracks so that the loudest parts of all songs play at the same level. Any quiet parts of songs will playback quieter, just as intended. The song which starts quiet and builds throughout would start quieter. Then it would build and finally reach the same level as the other loud parts on the album. There would be a consistent level relationship between the songs. If you measured the integrated LUFS for each song, you’d find that each had a different value. The songs with quiet parts would have a lower integrated LUFS value because the quieter parts would lower the average loudness of the song.

If however, you simply mastered every song to have the same integrated loudness, you would lose the level balance. Songs with quieter sections would have to be turned up to reach the same integrated loudness as those with consistent levels. So a song with softer verses might end up louder than a ‘wall-of-sound’ type track. A song that builds over time might start too loud. By the final chorus, it may end up being significantly louder than anything else on the album.

Album balance

At this point, you’re probably thinking, but isn’t this exactly what the streaming platforms are going to do anyway? Don’t they make all songs playback at the same target level? Well, in most cases, the answer is actually no. Whilst some platforms normalize songs on a track-by-track basis, most of the streaming platforms will honor the level balance between tracks when they are all part of one release. So when someone listens to the songs on an album or EP, the level balance will remain intact. One way that this is achieved by streaming platforms is by measuring the integrated LUFS of the whole album and then applying the same gain change to all tracks to make the whole album meet the platform’s target level.

So if some of the songs from your album have lower integrated LUFS values than others, that’s ok. This level difference will be reflected when you listen to the album on most platforms. It’s not until your songs play as part of a playlist or on shuffle alongside songs from other releases that songs are normalized track-by-track.

What LUFS should I aim for

Having established that an integrated LUFS value is not a great value to aim for, what should you be aiming for? Personally, I’ve had the best results using the recommendation that Ian Shepherd gives in his free Home Mastering Guide, which is to have the short-term loudness hitting no higher than -10 LUFS at the loudest parts of your songs. Then, adjust the levels so that the loudest parts all playback at the same level. This should give you a healthy loudness, but should also retain a good dynamic range.

Of course, you can make the loudest parts quieter or louder if this achieves the sound that you’re looking for. Remember, you shouldn’t just master your music to meet a specific LUFS value. You should instead master your songs to make them sound their best. If you’re mastering the loudest parts of the songs to hit -10 LUFS short-term, the integrated loudness may well be above -14 LUFS. But that’s OK! The streaming platform will adjust your playback level if necessary.

When setting the levels of your tracks during mastering, you should also make sure that your audio is peaking no higher than -1 dBTP. Click here for a full breakdown on dBTP (decibel true peak). At present, Spotify states that songs louder than -14 LUFS integrated should peak no higher than -2 dBTP. Personally, however, I’ve found that if you’re mastering so that your short-term loudness is no louder than -10 LUFS at the loudest parts of a song, then -1 dBTP is adequate.

Mastering using integrated LUFS

So given that you shouldn’t be aiming for a specific integrated LUFS value, are integrated LUFS still useful? Yes, knowing the integrated loudness of your songs can still reveal a couple of things to you. If your song is above -14 LUFS integrated, then you know that the song is probably going to be turned down. This allows you to experiment with less limiting. That way, you can see if things sound better with more dynamics.

Remember also that whilst all of the major platforms turn louder music down, not all of them turn quieter music up. So a song quieter than -14 LUFS integrated may play back quieter than other songs on some platforms. If you want to get really specific, you could use the loudness penalty website. There, you can audition your songs to find out how much they’ll be turned down (or up) by different streaming platforms based on their current targets.

What LUFS measurement do you use when you master your music for streaming? Leave your comments, questions, and suggestions in the comment box below.


FREE GUIDES: Get the best results from EQ, compression, and vocals

4 thoughts on “LUFS: How loud to master songs for streaming (& why it’s not -14 LUFS)

  1. Can you advise how the end listener will perceive the volume and dynamics discussed above if using ear buds of varying quality and hence compromise the whole LUFS discussion above!

    1. Hi David, thanks for your question!

      In general, my approach when mastering is to first get the music to sound the best that it can on my studio monitors in terms of dynamics, level balance, and all other aspects. Then, I audition the masters on the kinds of playback systems that I think listeners are likely to use. This is often a decent pair of speakers, a Bluetooth speaker, ear buds, car stereo, etc. I then go back and make adjustments to create a master that translates as well as it can across the multiple systems. So my advice to you would be to do the same. See how your masters sound on different sets of ear buds, and then tweak your masters to translate the best that they can across the different pairs. And always keep in mind that your aim is to make masters that sound their best to you, regardless of what LUFS value that may result in.

      I hope that this helps!

  2. Thank You so much for taking the time to share Your findings. This really brings light to a lot of creative possibilities that I wasn’t aware of before.

Leave a Reply

Your email address will not be published. Required fields are marked *