Now that I'm getting a bit more into mastering, not only in my own work but for the short film I'm working on and also a project that Mike Allen has given me, it's important to understand what exactly the proper industry standards are for how loud things can be. While mastering isn't all about making things loud, it's an important part of the process and there are some guidelines that need to be followed. These kinds of guidelines are set depending on the medium (film, music, etc.) and the target platform (Spotify,YouTube, Netflix etc.) so let's have a look at some different loudness standards for music and film, and the various platforms on which they're consumed.
Introduction to Loudness
To understand exactly what is required of us as audio engineers when it comes to loudness, we first have to understand a couple of key terminologies and theories that govern the way standards are set and measured.
Dynamic Range
Dynamic range in film mixes is also being reduced in the mastering process in order to get a seemingly loud mix while remaining under the required thresholds. Dynamic Range is effectively the difference measured in dB between the quietest and loudest sound in a piece of audio, and as Emmanuel Deruty from Sound on Sound puts it; "Intuitively, we feel that dynamic range ought to measure how 'variable' or 'mobile' the music level is."
According to Mike Thornton of ProToolsExpert.com, there is an unfortunate consequence that comes from film mixers pushing the limits of how loud they're mixes are; "Due to complaints from cinema goers, playback volumes in cinemas have consequently come down. As a result, many cinemas keep the playback level low for all movies, which puts pressure on post-production facilities to adjust their playback level." This is a pretty familiar story with the 'loudness wars' being a pretty big talking point in music over the last few decades.
Sample-Peak (Inter-Sample Peak)
This is a fairly in-depth topic and I'll do my best to explain it without getting too in over my head. In a digital recording system 0dBfs is assigned to the maximum possible digital level (Rietvink. A, 2016). Digital audio is measured in samples and the sample rate (44.1Hz, 48Hz, etc.) is the rate per second at which digital audio software takes a sample of the audio you're hearing. So needless to say, the higher the sample rate, the higher quality of the audio. Again, this is a very basic description so as I don't spend a hours on this one topic. When we actually listen to music, it has to of course be converted out of the digital domain into analogue which can then be played from your speaker system, and because samples are essentially blocks of information, this conversion smooths it out into a waveform that is appropriate in an analogue format.
Because of the way a digital to analogue converter works (as demonstrated in the diagram above), the "smoothing" process can often reach above the digital sample peak of 0dB and cause distortion. Meaning that if you master to absolute 0dB within your DAW, you're more than likely going to get distortion once you bounce and listen on consumer hi-fi (especially that of low quality). Now this doesn't sound that important when it comes to loudness standards but it is an important thing to understand when you're mastering, especially when it comes to the different formats and the way they process your audio.
LUFS (LKFS/LU)
"Rather than trying to measure audio level by counting samples (sample-peak level), circumstantial research has proven that even though two pieces of audio may be measured to be equally loud using the sample-peak method, they may very well be perceived as being very different in terms of level" (TC Electronics).
This is where most of the standards are set when it comes to loudness standards, because a LUFS (Loudness Units Full Scale) Measurement is the most effective way of measuring the average perceived loundness of a piece of audio rather than the now other averaging measurement. In a nutshell, Loudness Units are the unit of measurement used in the process of quantifying a piece of music’s perceived loudness by analyzing the average level over time (Anderton, C. 2018). There’s a further refinement called LKFS, which uses something called weighting to account for differences in frequency response; however, for all practical purposes the two terms can be used interchangeably (Waves, 2017).
Because LUFS are calculated using an average, you can monitor the measurement in different time-frames;
Momentary measures 400 ms of signal, sufficient for the shortest loudness timescale. Because this measurement is quite fast, it is usually displayed as a bar meter with smoothed release ballistics.
Short Term averages loudness over a period of 1 to several seconds, with some overlap between the time slices.
Long Term (sometimes called Integrated Loudness) averages the entire program, very often providing the most useful value.
True Peak provides sub-sample accuracy for measuring intersample peaks—those that occur between two consecutive samples and so cannot be detected by traditional meters. This is critical because when program material is later loudness-normalized (something that digital broadcast systems and online delivery services do as a matter of course), there is, in theory no chance of digital clipping or distortion, even after conversion to the kinds of lossy formats used by those systems.
(Waves, 2017.)
When it comes to meeting LUFS standards we want to make sure we're monitoring the Long Term or Integrated Loudness, as that will give us an over all measurement for the entire piece of audio and is generally what is referred to when a LUFS standard is required.
Now that we've got a bit of knowledge under our belt, let's look at the different standards.
Film, TV and Broadcasting Standards in Australia
In the days of 35mm film, someone from Dolby had to be present during the mastering of a cinema release. Masters were always printed with the volume set to 7 on the Dolby box. The trend at the end of the 35mm era was already towards very loud mixes. People present during the mix of movies like Tron were advised to wear hearing protection. The audience does not appreciate this level anymore, there are lots of complaints (Thornton, M. 2017).
For any kind of Broadcasting media there are certain standards you have to adhere to depending on what country you're in, however the standards that are set are generally blanket rules for all broadcasting mediums. OP59 is the Australian loudness standard for broadcasting mediums. The document which explains OP59 a lengthy one that can be found here; in the interest of LUFS, the target perceived loudness is -24 LUFS. They also make a rule of the True Peak being no greater than -02dBFS, if a meter that measures in this way is used, however it is more accurate to measure loudness in LUFS or LKFS (which are identical units of measurement).
Interestingly enough, "theatrical mixes do not have any official limitations or guidelines regarding loudness. Of course, this isn’t license to mix the film as loud as you can make it but rather to afford you the widest dynamic range as an artist to explore your auditory world of your film" (Simple DCP, 2019). As I mentioned earlier, cinemas have had to dial down the level at which they show films because audience's over the years have begun to complain more and more that it's too loud. This has resulted in mixing engineers having to change their habits when it comes to mastering their work. Dolby suggests as a standard that their Dolby boxes should be set to 7, however because of the audience's complaints, they have been forced to turn them down to 6 and often even 5.5. Dolby sets these levels as they ensure that the theatre sound system will output a specific dbSPL in C weighting as examined in this graph, provided by Mike Thorton of ProToolsExpert.com;
Streaming services like Netflix do have guidelines that they recommend when it comes to theatrical mixes, which can be found of their Partner Help Centre website;
2.0 Theatrical Mix - Mastering and Archive
Theatrical Mixes should follow the Near Field specification, listed above, with the following exceptions:
Use 85db spl as your standard reference level for mixing.
Theatrical mixes have no LKFS loudness requirement and may peak at 0db True Peak.
Provide Dialog, Music and Effects Stems at Theatrical spec
The "Near Field" specification is as follows in regards to loudness only;
1.0 Nearfield Audio Prerequisites for Mix Facilities
Average loudness must be -27 LKFS +/- 2 LU dialog-gated. Peaks must not exceed -2db True Peak. Audio should be measured over full program according to ITU-R BS.1770-1 guidelines.
So while they do mention that theatrical mixes don't need to hit a predetermined perceived loudness, they do have to follow their Near Field guidelines, which do in fact suggest that there is a necessary LKFS/LUFS they must meet. Confusing to say the least but I suppose it allows films mixers to have a bit of freedom with their theatrical mix while still being offered guidelines.
YouTube uses a compression algorithm that normalises all of the audio on videos posted to their site, in order to keep everything on a level playing field. Much like OP59, this kind of normalisation is in place to stop competing YouTube channels, content creators or advertisers from essentially "1uping" each other with the loudness of their content. A support article I read on YouTube's Help forum mentions that avoid having your audio squashed by YouTube, "So while you can't stop YouTube from doing this, it seems that you can master your audio prior to uploading it to YouTube somewhere in the -12 to -14 LUFS range in order to avoid YouTube's manipulation of it which is out of your control" (YouTube Help Form Post, 2019).
Music and Music Streaming
Music mastering levels are somewhat easier too follow than those for film and TV. If you're creating music for radio or TV, guess what, you're following the broadcasting standards of OP59, but if you're target platform is something like Spotify, then you'll have to follow their standards accordingly. For starters, Spotify does a few things things to process your audio;
Check they aren’t corrupt and that their format and container is known.
Convert to WAV 44.1 kHz (keeping bit depth).
Transcode the file into the following delivery formats for the quality options available to listeners:
- Ogg/Vorbis [96, 160, 320 kbps]
- AAC [128, 256 kbps]
- HE-AACv2 [24 kbps]
Calculate the loudness using ReplayGain.
ReplayGain is a fairly straight forward software that calculates the perceived loudness of an audio file, and also offers information about the true peak, short term loudness, integrated loudness, etc.
Spotify also explains their reasoning behind using this particular software:
"We currently use ReplayGain, which was the most recognized standard for calculating loudness when Spotify first started.
In the future, we plan to use a new standard for calculating loudness, called ITU 1770 (from the International Telecommunication Union). This defines the integrated LUFS (Loudness Units Full Scale) measure, and it’s what we recommend you use to measure the loudness of your tracks.
ReplayGain doesn’t specify a measurement unit for loudness, so we’re unable to give an exact measure in LUFS used by ITTU 1770. However, we adjust tracks to 3 dB higher than ReplayGain algorithm specifies, which is roughly equivalent to -14 dB LUFS, according to the ITU 1770 standard."
And they're even nice enough to offer some mastering tips for their platform;
Mastering tips for Spotify
Target the loudness level of your master at -14 dB integrated LUFS and keep it below -1 dB TP (True Peak) max. This is best for the lossy formats we use (Ogg/Vorbis and AAC) and will ensure no extra distortion is introduced in the transcoding process.
If your master is louder than -14 dB integrated LUFS, make sure it stays below -2 dB TP (True Peak) max to avoid extra distortion. This is because louder tracks are more susceptible to extra distortion in the transcoding process.
With all this information readily available from Spotify, it's pretty easy to follow their standards and make sure your music gets the best treatment in their processing, they even offer suggestions for file types and further reading on why you should master to these standards, and what will happen if you don't; here.
Website, Mastering the Mix has a small table on the standards across a majority of popular streaming sites;
That's a lot of different mastering in order to get your music onto multiple platforms, but it's worth it in the end. "Club Play" in this list suggests the mastering level for the version of your track intended to be played in a club, and unsurprisingly it's quite similar to CD level. I personally wouldn't think this is a big issue, because the club will most likely be playing something from Spotify, and if there's a DJ, they have control over the level entirely so your over all level could be changed at will, much like the Dolby issue with cinema's turning their playback levels down.
The reason I think it's easy to follow music mastering conventions is probably because it's all very similar to Film ad TV in a lot of ways; just make sure you've hit all the standards for the platform it's going to.
Rietvink, A. (2016). Audio levels for mastering and inter-sample speaks - Digital Mastering. Retrieved from https://ar-onlinemastering.com/blog/2016/09/22/audio-levels-inter-sample-peaks/
Mastering Tip: What Are Inter Sample Peaks & Why They Matter. Retrieved from https://www.productionmusiclive.com/blogs/news/mastering-tip-what-are-inter-sample-peaks-why-they-matter
Thornton, M. (2017). Loudness And Dynamics In Cinema Sound - Part 1 | Pro Tools. Retrieved from https://www.pro-tools-expert.com/home-page/2017/6/21/loudness-and-dynamics-in-cinema-sound
Loudness Metering Explained | Waves. (2017). Retrieved from https://www.waves.com/loudness-metering-explained
FREE TV AUSTRALIA OPERATIONAL PRACTICE OP- 59. (2010). Retrieved from http://ticklemedia.com.au/downloads/OP59_Measurement_and_management_of_Loudness_in_Soundtracks_for_Television_Broadcasting_-_Issue_1_-_July_2010.pdf
Sound levels in cinema - Simple DCP. (2019). Retrieved from https://simpledcp.com/sound-levels-in-cinema/
Nikolic, J. (2019). Loudness Standards - Full Comparison Table (music, film, podcast). Retrieved from https://youlean.co/loudness-standards-full-comparison-table/
Loudness Explained Page | Music Tribe - TC Electronic. Retrieved from https://www.tcelectronic.com/brand/tcelectronic/loudness-explained#googtrans(en|en)
Deruty, E. (2011). 'Dynamic Range' & The Loudness War. Retrieved from https://www.soundonsound.com/sound-advice/dynamic-range-loudness-war
6 Things You Must Do For Film Festival Audio - Cinema Sound. (2018). Retrieved from https://www.cinemasound.com/x-things-you-must-do-for-film-festival-audio/
[AUDIO] Normilized video ( Content Loudness) - YouTube Help. (2019). Retrieved from https://support.google.com/youtube/thread/3224290?hl=en
Netflix Sound Mix Specifications & Best Practices v1.1. Retrieved from https://partnerhelp.netflixstudios.com/hc/en-us/articles/360001794307-Netflix-Sound-Mix-Specifications-Best-Practices-v1-1