What started as an update to our article that we wrote a few years ago on the topic of streaming and loudness has become a whole article in its own right.

While initially typing, it was clear that a mere update of an old article would not have sufficed. Too many things that should have happened didn’t.

To complicate things further, the topic of online streaming/loudness has been a hot favorite amongst social media and youtubers. Unfortunately, much of the info that was dished out as the Holy Gospel has become outdated or just wrong.



As we are approaching the end of 2021, we can summarise the current situation in the audio industry like this “everyone does as they please.”

All the new standards that were brought in back in 2015 have become optional, voluntary, obsolete.

Starting from the major players in the streaming world, they have never been unified on this front. 

Off the bat, the signs were on the wall. They couldn’t even agree on the same LUFS target and adopted values between -16 and -13 LUFS (or between -23 and -11 if we include the paid Spotify Loud/Quiet options) without any apparent logical explanation. 

Furthermore, Spotify (until 2021) did not implement their loudness normalisation algorithms on LUFS, but instead adopted replaygain (1.0) (based on RMS) and some others (all DJ stores and Soundcloud)…well, they didn’t bother at all!

Apart from YouTube, none of the others made the loudness normalisation compulsory, and this is where some of the problems started.


You might have heard some engineers saying, “ignore the numbers, just make it sound good!”
While there is a lot to be said for solely focussing on the music and doing just what is required (without looking at the meters), it is also true that, in the real world, mastering engineers have to keep an eye on “numbers” regularly.
Leaving the streaming conundrum aside, an ME will always consider what the competition is doing (loudness-wise, dynamic-wise, tonal-wise etc…). By competition, I mean other productions aiming at a similar core audience. Failing to do so, the ME could end up delivering an uncompetitive product.
For example, if I master a grime track, I can’t ignore the loudness of the last Stormzy or Skepta’s records. If I find that my loudness levels are way quieter, I’ll have no choice but to rectify it quickly, or it will bounce back like a boomerang!
So you see, an ME is always aware of what sort of loudness range he needs to master within. Although I sympathise with the initial statement “just make it sound good,” I also find it a bit of a cliché and a lazy answer to a highly complex issue.

The -14 LUFS

If you ask on social media what level you should master for streaming, half of Facebook will shout -14 LUFS!
It is easy to understand why, as you can see from the previous chart, it seems that -14 LUFS hit the sweet spot amongst the major players.

Moreover, Spotify actively encourages you to have it mastered that way, with a -1dB tp as a buffer (or even -2 dB tp for their Loud option).
And even if you read our old article about the loudness in streaming, you’ll find out, we were actively promoting it ourselves too.

Unfortunately, all these good intentions were subject for the big players (Spotify, Amazon, Apple, etc.) to play ball. Well, they didn’t, and that is why we are where we are.

As of today, Youtube is the only big player that makes loudness normalisation compulsory. But, not without any issues, (although mainly in the past).


E.g., In 2016, we mastered a cracking record for The Computers, on One Little Indian.
As you can see from the picture, the video has not been normalised at all.
The new Lorde video “Mood Ring,” on the other hand, has!

Albeit a tiny glitch, it is a significant one.
After all, their servers need to put up with 500 hours of video uploaded per minute! (yes, that’s 720,000 hours of videos uploaded daily) and all those videos need to be encoded, analysed, and then rendered.

I always wondered, if on the off chance their video slipped through the cracks, some indie artists, producers, and labels keep uploading loud masters to Youtube.


SpotifyOFF (*)
  • Spotify client desktop and mobile had in the past different default settings. On a mobile, it is usually OFF

As I have explained before, the only big player that made loudness normalisation compulsory is YouTube.
All the others leave it up to the listener. In other words, you can completely disable the loudness normalisation on any other service if you wish to do so.
More importantly, some platforms  don’t even set it ON by default. Again, the listener will have to navigate through the app settings in order to activate it.

Wait what? Yes, you read it right.

If you don’t know that is there, or, even what loudness normalisation or LUFS are, you will then be streaming music  with the masters at their original level!
If your music was mastered at -14 LUFS and it is streamed on Amazon or Spotify with the loudness normalisation off, it will sound a lot quieter than most! And that, simply is not good.


It seemed just yesterday when Youtube announced to introduce LUFS analysis in their streams. Back in 2015, the professional mastering community was pretty united in the fight of the loudness war. All mastering engineers I spoke to, were pretty excited and optimistic about moving away from the uber loud approach. Papers were written, many posts popped up all over the net by top-tier engineers, and the concept of mastering for streaming was born: to master without putting loudness in the equation.

What was the point of slamming a master, if it gets turned down by the algorithm of the streaming service? Right?
It made all perfect sense, until the first cracks soon started to appear.


In the meantime, the social media machine got put into motion. Many mastering engineers (more or less credited) flooded cyberspace with video guides/tutorials on how to master for streaming.
Before we knew it, half of the audio community started to google terms such as LUFS, loudness normalisation, true-peak, and loudness penalty. Some sites popped up to check your masters and how they would be “penalised” on a specific streaming platform (https://www.loudnesspenalty.com/).
Because of the term “penalised,” many independent artists started to shackle themselves with numbers and specs (mainly LUFS and dB true-peak values) without actually looking at what was happening around them.
To this day, aggregators will still only take only one set of masters, while many streaming platforms do not implement loudness normalisation at all.


Music aggregators such as Distrokid, Ditto, CD Baby, iMusician, Tunecore, Awal, etc… still to this day will only accept one set of masters (with the only exception made for Apple Digital Masters).
Because of that, you will need to choose if you want your ME to master for streaming or CD. If you want both, you will have to pay for two submissions to the aggregator of your choice.
But is it worth it?


  • Target the loudness level of your master at -14dB integrated LUFS and keep it below -1dB TP (True Peak) max. This is best for lossy formats (Ogg/Vorbis and AAC) and makes sure no extra distortion’s introduced in the transcoding process.
  • If your master’s louder than -14dB integrated LUFS, make sure it stays below -2dB TP (True Peak) to avoid extra distortion. This is because louder tracks are more susceptible to extra distortion in the transcoding process.
  • Program loudness for Alexa should average -14 dB LUFS/LKFS.
  • The true-peak value should not exceed -2 dBFS


As you might have gathered, right now, the whole streaming/LUFS situation is very murky and unsettled.
Albeit, there are guidelines, everyone implements them as they deem fit. One thing is for sure; there is no united front any longer.
Our advice is to wait for the situation to settle and always ask the ME.
Even right now, there are situations where mastering for streaming makes sense. Since Youtube is the sole crusader out there forcing their streams (well, most of them) to be at -14 LUFS, then if you are shooting a video, it makes sense to use an optimised master for that purpose. It will not require a double submission to an aggregator, and you can be pretty confident that the master will translate as well as it possibly can on their site.
The same applies to Vimeo or Vevo (albeit, Vimeo likes the masters to be at 48k).


Our current approach is to advise our clients on an individual basis. We do not reject the idea of optimising our masters for streaming as a whole, but we want to make sure it is right for you.

Our approach to streaming is to provide you with a master that is still reasonably dynamic, not squashed just for the sake of it, and with an excellent translation (with or without the loudness algorithm). It will have a nice buffer (-1dB tp), and it will be 24-bit wav.
Ultimately it will be a master that will still focus on what is
best for your music from an emotional point of view.