Before you publish your audio or video content, you need to be aware if it complies with worldwide audio volume loudness standards.
Although this article is primarily about controlling volume loudness in Ableton Live, this applies to YouTube videos as well. I am in no way an expert at any of this, but rather this subject peaked my interest as I came across the need to do mixing for my audio projects. Immediately what came to mind was “How loud should the audio in my videos be?” That led me on a learning journey through the subject of volume loudness.
Spilling Popcorn All Over The Floor!
To put some practicality in all of this, lets assume you are at home watching a movie on your television. It is an action packed story loud in some places, quite in others. This difference between quite and loud is the movies dynamic audio range. Your hearing dynamic range maybe different. You get a sense of this and “know what to expect on extremes” and what is “normal” over time. Watching these kinds of movies on your tv with audio hardware you know basically what to expect. You’ve adjusted your TV volume in a soft spot to accommodate the changes. For your audio perception is trained and built to respond to wild audio highs and lows.
Now, on a quite Friday night, you sit on the couch with your girlfriend watching an action movie. An hour into the TV movie, you are watching an action scene that is loud and then, the scene cuts quickly to a young couple who are hiding in a barn whispering and making out. You and your girlfriend snuggle closer to each other. The TV broadcaster cuts to a commercial. Up comes that annoying loud, and obnoxiously rude used car salesman screaming at the top of his lungs! The volume is even louder than the action scene! Both of you jump out of your seats spilling your drinks and popcorn all over the family room floor. So much for making out!
Ok, I think you get the point now about why there needs to be rules on being responsible in creating audio with appropriate volume. :O)
Global Worldwide Audio Loudness Standards
So one of the big things today in audio is how much volume should you pump out to your listeners. Over the years, audio engineers have forced volume levels up to crushing highs in attempt to get attention. For music concerts and outdoor venues, that may be fine. But when it comes to broadcasting across radio, television, and even YouTube, you have to follow audio standards. Worldwide broadcasters have been forced to play audio to conform to government rules and regulations.
Movies play at a certain perceivable level of loudness. Rock music on DVDs play at a level of loudness. A podcast has its own appropriate level of loudness. Extend this idea across all kinds of things you listen to and you will realize, there is a whole lot of variety. You need a standard measure to go with and a way to adapt the level of sound if it is too loud or too soft. To determine volume level appropriateness, international broadcast standards use the normal loudness in voice with something called dialog normalization.
Dialog Normalization (Dialnorm)
In the United States, the CALM act punishes content broadcasters that create loud works of audio that annoys its viewers. In other countries, they abide by the ITU-R BS.1770 standard. Both laws requires audio to be based relative to the normal level of spoken dialog within that audio session. Technically, videos and audio are streamed content stored in digital files. Most use Dolby technology. Within a Dolby AC-3 encoded stream is a Bit Stream Information frame that has 5 bits set aside for dialog normalization (dialnorm). This number has a range between -1 and -31 that tells the interpreter (a decoder) how much to adjust the volume. It only exists in Dolby Digital or Dolby-E digital audio streams. It is not supported in analog audio.
The important thing to recognize here is that the value represents your content created audio’s dialnorm. I emphasize “your” to make sure the value you measure and report is your level of audio content loudness. If there is audio which doesn’t have human dialogue, say like music, you still need to measure, compare, and set this value accordingly to other works of music in the same category.
You determine the level of your audio content by using a software loudness plugin (no, not earplugs!) or purchase Leq(A) hardware to measure this value. You will be given three different readings:
- Short form
- Medium form
- Long form
All relate to the time frame of the audio content. You then take that value and set it appropriately within your audio creation software and store it in the audio/video file. This value will then be read by audio playback decoders and treated as a gain to adjust the volume higher.
Dialnorm is how much the audio’s dialogue (speech) level is below 100% digital full scale (0 dBFS). Think of 0dBFS as the point of unbearable loud obnoxious sound. The point where your ears bleed and things get way too distorted to be unrecognizable. Even if it doesn’t seem that way, treat it as such. A -31 dBFS level is the baseline level used by Dolby playback decoders.
According to Dolby’s Metadata Guide, the value for the dialnorm property ranges from -1 to -31. Each is 1dBFS of step of increased volume loudness gain.
- A -31 dialnorm means the Dolby player should make no adjustment in volume (-31 + 31 = 0)
- A -21 dialnorm means the Dolby player should increase the volume by 10dB (-31 + 21 = -10)
- A -11 dialnorm means the Dolby player should increase the volume by 20dB (-31 + 11 = -20)
When a commercial comes on, the broadcaster cannot blast its volume more than dialnorm of the television show. If it does, viewers can report this to the FCC and the station can be fined. To be on the safe side, if you are the content creator, you need to be aware of this and create audio content that always falls below 100% digital full scale and, even lower yet. So thinking ahead, if someday you have an excellent piece of work that goes TV mainstream, you need to abide by these rules or your work will be rejected.
Audio Metering In Ableton Live
There are many ways to measure loudness in an audio signal. In the analog world, you use VU meters. In the digital world, you can use Peak level and RMS meters.
Ableton Live has a Track Volume meter that shows RMS and peak levels. A Track Volume meter consists of a left and right channel with a default vertical unit scale in decibels, ranging from -60db to +6db. It has a peak hold level to show the maximum value that was generated. Track Volume meters are shown in both Arrangement (Track) and Session (Grid) view.
A peak meter represents a real time reading of an audio signal’s left and right stereo channels. As audio is played, a current level is displayed. A peak hold level (a real time “max” bookmark) is shown for the maximum decibel level obtained over time.
Root mean square (RMS) meters take the square root of the mean of the squares of the audio signal. Basically, its mathematics to get the average from a negative and positive range of numbers. It normalize all the voltage values to positive numbers first. Remember, a volt can be a floating point number between -1.0V to +1.0V in range. After these voltage values are squared to be positive numbers, you sum them up and take their arithmetic mean (average), and then take the square root of the positive mean to revert back from a squared average to a non-squared average. This gives one a sense of the average magnitude over a range of time. Notice it says nothing about the extremes, only the average. The extremes is where the peak meter comes in by Live constantly monitoring over time the highest RMS value.
So, think of the Track Volume meter as keeping track of the extreme high value and in real time, the current average volume over time.
Give Some Headroom
Now we get to the discussion of what is an acceptable loudness level. There’s two things to consider in doing this:
- Dynamic range – the low to high range of the output signal
- Peak level – the highest value the output signal achieves
Ableton Live is able to calculate large input/output ranges due to internally, it uses 32-bit floating point numbers. These values from a software algorithm standpoint, will not result in Live clipping positive values. The meter just reflects what the values are and pictorially show them in the Track Volume meter. This means, that even though you’re audio exceeds 0dBFS, Live won’t remove hot values. You or a professional audio engineer can still tweak things down.
Generally though, don’t let your track volume values exceed 0dBFS. It is always best to leave more headroom than less headroom. Never hot in the red zone..
Many people will give you different responses to what amount of headroom to leave. Some say -4dBFS. Others -6dBFS. There really is no set standard. The important thing is, just leave headroom. Watch each track to see if the peak level exceeds 0dBFS and goes red. Adjust as necessary in the framework of your audio structure. If it doesn’t make sense for those congas to be blasting away, tweak them down.
Its best to mix your audio project with a duller sound to it. Then, when mastering, increase track volumes where needed.
If you run into a MIDI or audio track that is getting too hot, adjust that track. If the master is getting too hot, everything should be blended in together in relation:
- Multiple select all tracks
- Reduce all of them by -3dBFS
- Keep reducing until the master is no longer hot
When you create your audio file, remember that it can be handed over and played anywhere. There are many plugins for measuring loudness. Here’s a list that measures and analyzes in ITU BS.1770, EBU R128, and ATSC A/85 standards:
- NUGEN VisLM 2: Many including Avid Pro Tools, Steinberg Nuendo Price $449
- Dolby Media Meter 2 – Avid Pro Tools, Steinberg Nuendo, Sony Sound Forge: Price $689
- TC Electronics LC2n – Avid Pro Tools: Price $299
- iZotope RX Loudness Control – Avid Pro Tools, Adobe Premiere: Price $349
- WLM Loudness Meter – Many hosts including Ableton, Avid Pro Tools, Cubase, Adobe Premiere: Price $399
If you want to comply to worldwide and US FCC regulation, it is best you use one of these plugins to analyze your audio work. Ableton Live’s Track Volume does not conform to ATSC A/85 CALM as it is RMS and Peak level only.
Also, I strongly recommend that you test the plugin against ITU compliance files. It is of no use to be spending hundreds of dollars for something that may not even give you the correct loudness reading against the ITU standard’s test comformance files. Every plugin product I’ve seen does not mention whether they can prove conformance. Make sure you get it in writing that they do conform and offer a 100% money back guarantee if they don’t.
Worldwide loudness laws have nothing to do with RMS dBFS and peak units but rather LKFS units as mentioned in ITU BS.1770. You cannot use what is in your default DAW metering to conform to loudness laws. You have to use meters that are ITU BS.1770 compliant. I would recommend making sure your loudness plugin conforms to ITU BS.1770-3 Leq(RLB) algorithm. Faulty implementations of Leq(RLB) algorithm can result in way different LKSF readings. Always test the software plugin against the set of WAV data used to conform.
Remember, the point of these loudness plugins is to make sure you conform to local government audio laws. If your plugin produces different results than the standard LKFS results for each WAV file, your wasting money, your time, and may not even comply with your local broadcasting laws.
If you choose to use Track Volume metering, give yourself some wiggle room (also know as headroom) by keeping the track volume below -10db for each track. You do this by fiddling with the Volume control (the little left arrow triangle icon) so that it maxes out (i.e. peaks) at -10db.
TC Electronics LC2n is used in Adobe Audition and Premiere Pro CC. So if you have an Adobe CC subscription, you can feed your Ableton Live audio into it for further mixing. Adobe Audition is a great tool for mixing, cleaning up audio, restorating, conforming to loudness standards, sound effects, among many other features. Its useful not only for audio but for video as well.
- Make sure every track doesn’t go over 0dBFS
- Create enough headroom for adjustment by leaving -3dBFS to -6dBFS at peak level
- Do not compress or limit the dynamic range of a track to be less than -3dBFS in range. It needs fluctuated movement.
- Take a file snapshot of your mix and save it as an Ableton project along with all the tracks
- Hand over your Ableton project to a professional mastering engineer
- Use a dialnorm loudness plugin or Leq to measure the short, medium, and long form of your audio content
- Know your government’s loudness laws and standards
- Set your dialnorm value using what you read from a hardware Leq or loudness plugin
- Use equalization, loudness, limiter, compressors
- Have the mastering engineer give you back the project as an Ableton project