Mastering Audio for Streaming Music and Video-services!

Feel confused? Don’t worry, you’re not alone…

The process of mastering audio for all the streaming services out there can be quite a daunting task compared to delivering an audio master for physical replication in the pre-digital age that used to be quite simple and straight forward. Pretty much the only technical matter you had to worry about at the mastering stage was to make the final audio stream into 16bit 44k which has and always will be the standard for Audio-CD’s and leave a little headroom, just a fraction of a db at the very top of the dynamic spectrum to safeguard against the way some cheap D/A:s in CD-players would handle the audio so you would not get any unexpected distortion.

A plethora of audio forums!

CD was THE ONLY format that guys like me starting out in early 2000 had to think about for that entire decade except for the very odd vinyl release. Today, it’s a jungle out there! A jungle that even major label people don’t really know how to wrap their head around.

In the good old days of digital, you only needed to consider that any CD would be played back on any standard CD-player. Today, you need to consider that a mastered audio file will be played back on;

SoundCloud
Pandora
Spotify Standard
Spotify Premium
YouTube SD
YouTube HD
Apple Music
iTunes Standard
iTunes Plus
Tidal
Tidal Premium
Tidal HiFi
Facebook
Vimeo
Beatport

etc….

All of these music/video services and sub-categories of them use slightly different formats for streaming sound to the consumer and also in some cases use different ways of altering the sound through various dynamic processes or loudness schemes. In order to have each service sound its very best you need to at least cater to their specific use of audio codecs. An “Audio Codec” takes the full quality master and compresses the data by throwing away sound information that you would “not likely be able to hear or appreciate anyway”… Just to keep data transfers to a minimum. In the process, it creates distortion.

Depending on what type of codec a service chooses to employ you can make adjustments to the final master in order to have it sound its very best through that specific codec and encoding quality used by that codec for different sub-categories of the service. Sounding complicated?

A plethora of audio formats!

At the moment we only need to think about these major delivery formats for online streaming;

MP3
AAC/MP4
OggVorbis

FLAC

Each format has a different encoding quality depending on the service involved. Like for example, standard iTunes uses a 128kb/s AAC-stream while iTunes+ uses 256kb/s AAC, and the optimal master for those two qualities of the same codec will involve different decisions being made, and then consider 128kb/s MP3 which SoundCloud uses and you have a third option for creating an optimal master just for that or why not a 96kb/s AAC that Tidal uses for mobile?

In comparison it would be like making small adjustments in order to cater to a dozen different makes of CD-players and how they choose to play back the content of any specific Audio-CD in significantly different ways.

So what’s the solution?!

Because no one wants to master their music for the least common denominator, it’s standard practice to create up to five different sets of masters as final delivery;

* DDP (16bit 44k + meta-data formated as a DDP-fileset for CD-replication).
* LP Master (24bit XXk WAV-files optimized for vinyl engraving).
* CD Pre-Master (16bit 44k WAV-files for FLAC-encoding. Perfect copy of the DDP without meta-data).
* Standard Quality streaming and download (16bit 44k WAV-files, modified to sound as good as possible but not optimal for any streaming service using a lossy-codec encoded below 256kb/s).

* High Quality streaming and download (24bit XXk, modified to sound optimal on high quality streaming services using a lossy-codec encoded at 256kb/s or above).

These distinct sets of masters are then handed to the label or artist for distribution to all the different streaming and download services and physical mediums. As you can imagine, this creates confusion all the time for which specific masters to use for what type of service or forum, and everyone has different routines and options for handling their uploads as well which doesn’t make things easier.

In conclusion!

To ensure that you get the very best quality possible out of any streaming music or video service out there today, be sure to ask your mastering engineer exactly what formats to use in any single case for whatever situation you might come across. It can make a very big difference in the end!

All in all, it’s a very confusing digital world compared to a decade ago where you would just go… “Here’s the final master!”, and everyone knew what to do with it and it would sound great anywhere in the world.

The author is the founder of “The Panic Room” which is a multiple grammy awarded and gold/platinum selling music production company in Sweden offering Production, Recording, Mixing and Mastering services.

Thomas “Plec” Johansson – Producer / Recording, Mixing & Mastering Engineer

Advertisement

The Audio Quality of future music releases!

There has been this notion going around for a while concerning the audio quality of future music releases. The notion being that since music consumption is now happening mostly through digital downloads and streaming which ultimately means that music is played back on ear-buds, laptop speakers and iPads, that it isn’t that important for production quality to be of such a high standard anymore.

The idea being that you can get away with doing a lot less, sacrifice quality and not spend as much time and money creating the final result as everyone will be just as happy anyway since no one will be able to appreciate the differences in the end.

That doesn’t sound like such a bad thing, right?
 
We’ve already seen the worst of digital!
 

It would even be a great idea too if it wasn’t for this single fact, that we’ve already seen the worst of digital and it has been getting better and better for the last 35 years, except for that little bump in the road called the Loudness Wars that ruined digital audio for 25 years. Good news though is that it’s over, or it’s over quite soon and we owe that a lot to the death of CD and the revolution of streaming services like Spotify, YouTube and Apple Music.

Since the quality of digital audio on the recording side has skyrocketed during these 35 years and is now on par with the most high-resolution analog formats, the only thing that is of a concern nowadays is storage space. The streaming services are in the business of selling data transfers over the internet, and this data just happens to be music. Since digital storage costs money and takes up physical space, and while keeping a library of all the worlds music on file in a low-quality format versus a super high-quality format makes for about a 95% difference in the amount of storage space and internet speed needed to transfer it, so it might not come to anyones surprise that this now is the current state of things.

Digital streaming is gearing up!

Vinyl releases are seeing a huge boom while the CD is almost gone and digital streaming is gearing up slowly but surely and we’re not going back any time soon. The quality of streaming is going up at the same rate that the world is getting better and better connected and storage is getting more effective and cheaper. If we jump back to about 15 years ago to the dawn of Napster and say that this was the first ever global music streaming/distribution service, we can see that today the quality of sound has tripled or even quadrupled in comparison.

Listening to old MP3’s of what you might have downloaded through Napster back in the day but on todays modern ear-buds, laptop speakers or iPads and compare that to the quality you get from todays streaming services listening through these modern playback systems would come as a shock to most people and kills the notion of good production quality in this day and age not making any difference in the long run.

It’s only fair to assume that the quality of streaming music will improve drastically in the following years and for now it seems that Apple is the only company taking this seriously and is on the forefront of development in this area.

The future is bright, but the sound is warm!

If the same argument of dumbing things down would’ve been made in the early 80’s when the CD became the #1 source of music distribution, or even much, much worse had it been made at the birth of Napster, we would’ve had to listen to horrible sounding productions today. The people who thought CD was great back then only thought so because of the fact that it was much simpler and easier to handle compared to vinyl records while the people who thought MP3’s were awesome thought so mostly because they were much easier to handle than CD’s. Not because any of it actually sounded better at the time but in hindsight, pretty much no one would like to listen to an early 80’s CD or lossy late 90’s MP3 given the choice.

The sound of these formats were nothing short of horrible compared to what was actually created in the studio. Why? Because when CD came out the analog to digital conversion was completely new and it would take 10-15 years before things started to sound good on this format compared to the master tapes leaving the studios. The same thing is happening now, since when Napster let the world in on the glory of MP3’s, it would take the same 10-15 years before things started to sound OK in lossy encoded formats due to development, internet speed and the structure and standards now provided by the streaming services.

In conclusion!

Since record labels nowadays make their revenue off of streaming music and there mostly through huge back catalog, it would not be a good investment for anyone to produce music that no one would want to listen to in 10 years time and cringe in the same way one would hearing an old Napster MP3 as we know it today.

If we let the way things sound and the way people listen to music at the moment, dictate how we ultimately choose to produce, record and document music for the future, we can only look back at this in 10 years and go -“What the hell were we thinking?!”

The author is the founder of “The Panic Room” which is a multiple grammy awarded and gold/platinum selling music production company in Sweden offering Production, Recording, Mixing and Mastering services.

Thomas “Plec” Johansson – Producer / Recording, Mixing & Mastering Engineer

Email: plec@thepanicroom.se
Web: www.thepanicroom.se
Facebook: www.facebook.com/thepanicroomse
Instagram: www.instagram.com/thepanicroomse/
Twitter: www.twitter.com/ThePanicRoomSE
LinkedIn: www.linkedin.com/in/thepanicroomse
YouTube: www.youtube.com/user/ThePanicRoom
Vimeo: https://vimeo.com/thepanicroomse

EBU-R128 : What does it mean to the modern producer/engineer?

There seems to be a lot of questions and misinformation floating around regarding the EBU-R128 standard and its practical implementations. From reading up about it, having worked with it for a few months in different scenarios and noting its practical use, I find it proper to write a short article since it’s causing a lot of confusion out there in the music industry where myself and colleagues find ourselves having to sort out misconceptions from time to time.

What does this really mean to the modern producer/engineer?

The EBU-R128/ITU-R BS.1770 was developed as a broadcast standard together with the LU (Loudness Unit) way of measuring loudness and is a superb tool for analyzing and evening out the loudness of program material ranging from all genres of music to commercials and just all types of audio. This is a great thing since the most dynamic classical piece of music you can imagine suddenly plays back at the same perceived level as Metallica’s “Death Magnetic” without losing any of its dynamics. But implementing this for modern music production, meaning basically anything less dynamic than classical music, it needs to be taken with a grain of salt.

LU vs. VU!

The EBU-R128 standard has set a reference value for loudness called 0LU (Loudness Unit) which is the same thing as -23LUFS (Loudness Unit Full Scale). LU is a much more accurate way of measuring perceived loudness compared to the old VU-meter (Volume Unit) or in the digital world measuring “dBFS RMS”. Just for reference, 0VU corresponds to -18dbFS RMS which was the old reference for “how loud music should be”. Now, modern commercial music very seldom sits at a level below -10dBFS RMS and depending on if you’re measuring your RMS levels at AES17 or not (which most people still don’t it seems, so we’ll leave that out for clarity) -10dBFS RMS corresponds to a value of 15LU. This means that most currently produced music is AT LEAST 15dB too loud in order to meet the EBU-R128 broadcast standard!

In any type of broadcast situation, most modern music will have to be turned down by 15db or more before it hits the air. So if you would look at the 0LU reference in dBFS RMS terms, you’d have levels sitting at around -25dBFS RMS on your final master which in practice means that you’d pretty much need to half your level, then half that again, and almost half that once again for most currently produced music to end up at the reference level set out by the EBU-R128 standard.

As you can see that’s a pretty darn soft level!!

Setting the standard!

Most Digital Audio Workstations today have some kind of an LU-meter built-in pre-set to the EBU-R128 standard and for most modern production it’s pretty much unusable since you will be going off the charts and pinning the meters from the very first beat! So if you can’t change the reference level of your built-in LU-meter you will need to work at an extremely low-level just to be able to use it properly… and that will be around 15db lower or less than you’re used to. What this means in practice is that the LU-meter in itself is a superb tool for measuring perceived loudness of any type of audio material but implementing the EBU-R128 standard reference level of 0LU for most types of music is pretty pointless before all physical media is gone and all types of online streaming such as Spotify, iTunes, YouTube etc… conforms to this standard playback level without deviation.

Wait a minute, they just did! 😀

Or, they kind of did anyway…

As I said, the EBU-R128 standard is just too dynamic for most types of modern music since it creates A LOT of unused headroom and considering the tiny amplifiers that people tend to listen to music on today, this headroom doesn’t make much sense since you could never get a good level going through your iPhone or computer speakers. The good people at these streaming services realized as much and raised the bar by 6-10db. YouTube for instance has put a constraint on their audio to be played back at around -13 LUFS compared to the EBU-R128 level of -23 LUFS. This makes much more sense to the average consumer and streaming services seem to implement a standard of between -16 to -13 LUFS.

This means that most modern music would need none to very moderate limiting in order to reach this level of loudness.

Keep your resolution!

Fact of the matter is with most types of modern music (house, edm, rock, metal, pop, rnb etc…) that don’t rely on a huge dynamic range like classical music and film scores, you will need to struggle in order to get peak values to read over -10dBFS at the mix bus of your DAW when it’s totally uncompressed/unlimited and conforming to the EBU-R128 standard. In practice, this means that if you would mix and master to conform to the EBU-R128 standard of 0LU, for most modern music productions you will have 10dB or more of unused headroom on your stereo bus.

If you then mix this down to a 24bit WAV or AIFF file, you’re basically throwing away two bits of resolution at the top of your dynamic range, leaving you with a 22bit resolution in practice. This is why you don’t really need to bother with the EBU-R128 standard unless you’re working specifically towards broadcast or you’re doing classical or film, otherwise you are basically throwing away a small part of your sound.

Summary

Just because the EBU-R128 standard for broadcast and now streaming services even out all playback levels, it doesn’t mean you should stop compressing and/or limit your mix or master altogether. From what I’ve seen so far it’s not far-fetched we’ll see the loudness war turning into a dynamics battle! Instead of “how loud can we make it?” it will be “how dynamic can we make it?!”. I might be the first mastering engineer saying this but.. DYNAMICS AREN’T ALWAYS A GOOD THING!

Tools designed for loudness and digital level control has a musical impact, and used in the right amount and with good taste they also offer a musical contribution. Some genres just sound good or right when mastered and mixed for loudness, or to be perfectly clear, a limited dynamic range. What the EBU-R128 enables you do to is that you don’t need to produce, mix and master for any louder than you think sounds musical for any given purpose. You don’t need to compare your work to others and feel you need to put out a louder product. You can finally just choose to go with what sounds good and be done with it since the competition will still be played back at the same levels, which finally opens up the door for better quality sound!

The author is the founder of “The Panic Room” which is a multiple Grammy awarded and gold/platinum selling music production company in Sweden offering Production, Recording, Mixing and Mastering services.

Thomas “Plec” Johansson – Producer / Recording, Mixing & Mastering Engineer

Email: plec@thepanicroom.se
Web: www.thepanicroom.se
Facebook: www.facebook.com/thepanicroomse
Instagram: www.instagram.com/thepanicroomse/
Twitter: www.twitter.com/ThePanicRoomSE
LinkedIn: www.linkedin.com/in/thepanicroomse
YouTube: www.youtube.com/user/ThePanicRoom
Vimeo: https://vimeo.com/thepanicroomse