For many years home viewing television audiences have been frustrated by the loudness of the commercials. No matter how loud or soft the prior program was, the commercials always seemed to pop in louder – much louder. In the US it seems that the Congress (don’t worry, no politics here…) moves faster and more decisively when their constituents are highly motivated. This issue is one where the residents of the US made their displeasure loud and clear. In fact one of the Act’s sponsors said that in all of her years of creating law not one issue generated more heat with her voters than the loudness of TV commercials.
To explain what the Congress has done will take quite a few layers as well as provide plenty of acronyms – it is a government mandate after all. At the end of the day however, it’s only going to be about reading audio meters. I will have to offer some possibly unfamiliar terms here but stay with me, it should all be clear by the end.
The CALM Act requires that the US Federal Communications Commission (FCC) enforce that the networks and cable providers broadcast no commercials louder than the programming before or after it. While that may be a worthy goal it could only be achieved by first enacting a number of loudness standards. The FCC has adopted the Advanced Television Systems Committee (ATSC) white paper called “Recommended Practice: Techniques for Establishing and Maintaining Audio Loudness for Digital Television”. Also known as the ATSC A/85 RP, a full copy is accessible online. Have a look, its a fascinating and enlightening guide to the production, distribution, and transmission practices needed to provide the highest quality audio. The act went into effect on December 13, 2012.
What does the ATSC document specify and therefore the FCC require for the loudness standards for television programming? A few more acronyms to follow… specifically the ITU-R BS. 1770 audio measurement in LKFS (Loudness, K-weighted, relative to Full Scale) which is their agreed upon measurement for the dialog volume level within a full mix.
The idea here is to mix, deliver and broadcast the audio with a constant and agreed upon level standard in the BS. 1770 meter across the board. If everyone was mixing to the same specifications and those mixes were all broadcast with the same specifications then theoretically everything would be the same audio level and voila! no more messing with the remote control.
Hands up – how many people still have to constantly adjust the volume between programs, stations and commercials? All right, trick question, I can’t see any of your hands. For the record, mine is up. Besides, I don’t think that any regulation or standard will ever alter the dynamic content of television. Some programs are meant to be quiet and subdued and many commercials are designed to be loud.
Audio is a “relative” experience. It is altered by the quality of the broadcast and playback equipment, the quality, size and dimensions of the listening environment, how the source material was originally created and mixed and how the material is reproduced at the set top box.
Audio amplitude is measured in several meaningful ways and in the digital realm it is measured in decibels or dbs. The ATSC A/85 RP document specifies several specific audio measurements. Among other things the standards call for a decibel true peak “dbTP” scale and that the dialog element level be measured with the ITU-R BS. 1770 algorithm expressed in LKFS.
TV and film programming typically uses the dialog as an anchor element of the overall mix. The program is mixed for dialog clarity and level and then the other elements – music and effects are mixed to compliment the dialog. The ATSC A/85 RP document (hence the CALM Act) specifies the following audio requirements – a measure of the program’s dialog in LKFS using an ITU-R BS. 1770 meter to maintain a reading of -24, +/- 2 db over the course of the entire program length. Therefore the final mix output meter must measure between – 26 to -22 to be in compliance. Which is plus or minus 2db from the -24 standard. The document also specifies that the dbTP never reach above -2db full scale “dbFS”. There is also a short term interval reading of the LKFS (every three seconds) where there is an allowance for additional db over the 2db variation of the -24 standard for short intervals.
In their delivery documents broadcasters require various audio parameters for their audio master deliveries. This includes sets of the full mixes, those mixes split out into their individual components as well as various other things and adhering to the loudness standards specified and required by the CALM Act. In the real world of television however there are variations to the CALM Act that the broadcasters require.
The meter shown here is made by Dolby Labs. It is called the Dolby Media Meter and provides all of the measurements required to comply with the loudness standards of the CALM Act. The main windows can be programmed to show various things and here you can see the infinite term LKFS, the short term LKFS, and the dbTP measurements that have already been discussed. I use this meter daily in my TV mixing work. All of the key measurements are easily viewable in this one plugin to check compliance at each stage of the mix. I use many other meters for many other things, but for the CALM Act compliance and the LKFS measure this is the one I use most. There are a number of other LKFS meters currently commercially available.
Looking at that meter you may notice that the dbTP is at -12 and not -2. There is a sticky reason for this. For many years broadcasters have required that most TV mixes be limited to -10db. This is one of the main reasons why most TV programs have very little dynamic range. Dynamic range in audio is defined as the range of loudness to softness of amplitude. Since mixes are required to be delivered at -10 but also (usually) required to keep their average or RMS level to between -20 and -10db, that is only 10db of dynamic range. Listen to a well recorded Beethoven symphony and the dynamic range could be 40 dbs or more.
The standards recommendations of the ATSC are a good thing overall. These are serious, well considered concepts and conclusions by some very smart people. However they do crash into the reality of delivering content for broadcast television. Some of the broadcasters I deliver for require the indicated +/- 2db on the infinite measure. But several require +/- 1db on the infinite measure. Why they decided to take away that 1db I’m not sure. That is pretty tight spec over a long form program. The CALM Act specifies a true peak of -2 dbFS however many broadcasters still require the old limit of the -10db specification.
Until all broadcasters move away from analog and fully embrace the CALM Act in all of its digital glory we are stuck with this odd compliance – follow all of the CALM Act measures except for the odd variations of a specific network’s specs. The output of the old standard of -10dbFS now, even more than before, severely restricts the dynamic range of our television programming.
I am interested to see what happens when everything does finally become fully compliant and all mixes follow an output level of -2db true peak. At that point we may finally get back to decent mixes, filled with dynamic range. Mixers should then be able to rely much less on compressors and limiters to scale back the outputs and create more natural sounding mixes.
What might be one other consequence of true compliance? Maybe this example – a heartwarming moment, quiet and touching, coming just before a commercial break. You’re swept in the moment of the program and then the commercial audio slams in at -2 db true peak throughout its entire 30 second length. My guess is that you and me will be jumping for the remote. The commercial will be too loud.
Just like old times.