Tikalon Header

Too Loud!

October 29, 2010

At last year's transition from analog to digital television, most of our television channels disappeared. Sure, some new channels appeared, but I don't speak Spanish, Korean or Japanese, so we were forced to get cable television. Our choice of cable was optical, not copper coaxial, and now we get our internet optically, also. One of the first things we found was that the variation in audio volume on the cable channels was far more dramatic than we experienced with broadcast stations. These changes were extreme, both between channels and as the content changed on individual channels. Although video signals on television are uniformly excellent, the quality of audio on television has been a neglected area. When stereo sound was first proposed for television, something which was very easy to do technically, the proposal was rejected by the FCC. The FCC claimed that television's impact was primarily visual, and stereo sound wasn't much of a benefit.

My solution to our television sound problem was to design and build an
automatic volume control circuit. Such circuits had been a hobby of mine since vacuum tube days, and I was interested in applying some of the newer integrated circuits to this task. My circuit worked well, but shortly after my building it, we bought a high definition flat panel TV with a built-in audio controller. Manufacturers had apparently identified this problem of wayward audio signals on cable and decided to incorporate a fix as a selling point.

Average audio signal level is easily controlled, but volume isn't what it's all about. An audio signal with the same average or peak amplitude level as another can still sound louder; and, in the case of a commercial, louder and more annoying. That's because
audio engineers can do some simple processing of the audio signal to increase its apparent loudness without exceeding modulation limits. The primary trick is compression, in which you increase the amplitude of the quiet passages. Not only that, but you can look at portions of the frequency spectrum of your signal and boost the amplitude of ranges of frequencies in a way that increases loudness.

Equal loudness curves via Wikipedia

Equal loudness curves per ISO 226:2003 (red). The blue lines specify the older ISO curve.

The phon is an accepted unit of loudness, and it's derived from auditory testing of a general population. As the figure above shows, the same sound pressure level produces a different loudness in phons at different frequencies. This was discovered in pioneering work by Harvey Fletcher and Wilden Munson at Bell Labs. As in most studies of this sort, they had their subjects compare one thing to another. Since 1 kHz is somewhat in the central range of human hearing, they asked people to adjust the level of a 1 kHz tone until its loudness was the same as tones at other frequencies. With today's computers and some free and open source software, a high school student could probably do the same thing for her science fair project. In all fairness to Fletcher and Munson, 1933, when their experiment was published, was barely into the vacuum tube era and an experiment like this was difficult to perform with the required accuracy.

It's easy to see from the figure that we can't hear anything above 20 kHz, and we have a hard time hearing low frequencies.
Audiophile amplifiers don't have volume controls. They have loudness controls to compensate for the fact that we need much more bass (low frequencies), and a little more treble (high frequencies), when the sound is turned down. This can be seen in the equal loudness curves figure, where we see that the dynamic range of the mid-frequencies is much greater than that at the low frequency end. This knowledge is summarized in another audio curve called "A-Weighting."

CBS Laboratories Stereo Loudness Controller, Model 711

CBS Laboratories Stereo Loudness Controller, Model 711.

In the late 1960s,
CBS Laboratories decided it would do something about the growing number of complaints about loud commercials by designing a loudness controller (see above figure). This controller divided audio into five bands that approximated the equal loudness curves, and the unit controlled the overall signal level based on a weighted average of the signals detected in each band. Possibly because broadcasters did not want to antagonize the hand that fed them, not many of these units were sold. Another reason might have been the price, $1,720 for a stereo controller (Model 711), which is nearly $10,000 in today's money. I had a brief career in broadcasting, and I saw many pieces of CBS Laboratories equipment, but I never saw a loudness controller.

Figure 1 from US Patent No. 3,582,964

Figure 1 from US Patent No. 3,582,964 for the CBS Laboratories Loudness Controller, from Ref. 2.

Nowadays, the same function as the CBS Laboratories Loudness Controller can be done much less expensively and more accurately with a digital signal processor (DSP). There may soon be a renewed interest in loudness control with the foreordained passage in November of a new law limiting loudness on broadcast and cable television. The US Senate unanimously passed Bill 2847, The Commercial Advertisement Loudness Mitigation (CALM) Act, on September 29, 2010.[3-8] A similar bill was passed by the US House of Representatives, and a committee will work out a compromise law in November. The FCC is already empowered to enforce loudness control via the "Communications Act of 1934, as amended," and this is a formal prod by the lawmakers to get some action from the commission to address television viewer complaints. The FCC studied the issue in 1984, but it didn't quite know how to define "loudness," so it never pursued the matter. Now, however, technology has improved and there is an established standard for loudness measurement and control. Senator Charles Schumer of New York, a consumer rights advocate and co-sponsor of the Senate bill, said
"TV viewers should be able to watch their favorite programs without fear of losing their hearing when the show goes to a commercial."

Under Senate Bill 2847, commercials could not be louder than the programs in which they're embedded. I guess that won't be a problem for shows like
WWF Wrestling. Television providers would have a year to tool-up to make this happen.

The broadcasters brought this onto themselves through their
Advanced Television Systems Committee. This committee, which established the standard for digital television in the US, also created a reasonable guideline in 2009 for measuring and controlling loudness. Under the anticipated law, the FCC will be required to enforce this loudness standard. As a recent article in a television trade publication indicates, there are a lot of equipment choices on the market that will accomplish the required loudness control.[9] The only problem that remains is that online content is not controlled by the FCC, so you can still expect wild loudness swings on your computer.

References:

  1. H. Fletcher and W.A. Munson, "Loudness, its definition, measurement and calculation," Journal of the Acoustic Society of America, vol. 5, no. 2 (1933), pp.82-108.
  2. Emil L. Torick, Richard G. Allen, Benjamin B. Baurer and Allan J. Rosenheck, "Automatic Loudness Controller," U.S. Patent No. 3,582,964, June, 1971.
  3. Senate votes to turn down volume on TV commercials, Yahoo News, September 30, 2010.
  4. Megan Gibson, "Senate Makes Itself Useful, Votes to Ban Loud TV Commercials," Time.com, October 1, 2010.
  5. Melissa Bell, "TV ads to lower their volume; ears rejoice nationwide," Washington Post, October 1, 2010.
  6. Bill to Regulate the Loudness of TV Ads Passes in Senate, Mediabuyerplanner.com, October 1, 2010.
  7. John C Abell, "U.S. Acts to Quiet Blaring TV Ads. Welcome to the 1960s," Wired, September 30, 2010.
  8. David Lazarus, "Congress considers shushing loud TV ads," LA Times, October 8, 2010.
  9. Steve Harvey, "Looking for Loudness Solutions," TVTechnology.com, October 13, 2010.