Cable Ratchets Down the Volume
The voices telling the cable industry to pay attention to its audio level problems are growing louder. For years, operators have known that there this is a problem area. There actually are three distinct challenges, all related to inconsistencies between diff...
For years, operators have known that there this is a problem area. There actually are three distinct challenges, all related to inconsistencies between different programming elements. The key, as the election season approaches, is that one of these issues - the volume of commercials compared to the programming surrounding it -- actually is getting Congress close to acting.
The other two challenges, which can be addressed with the same technical fixes as commercial volume, are inequalities between non-commercial programming on the same channel and differences in volume between channels.
While this is a trio of old issues, they have grown worse in the past decade. Miranda Technologies' [www.miranda.com] product manager Guy Marquis said that the transition to digital transmission has heightened attention to audio disparities because the extremes are far greater than they were in the analog world.
In the older scenarios, Marquis said, the spread between the loudest and softest audio was only 40 to 50 decibels (dB). Now, he said, the range between an exploding bomb in a war movie and a whispered conversation in a romance can reach 90 to 100 dB. That big difference is exacerbated by program creators, who tend to accentuate the volume because of the better overall quality of digital audio, which is particularly apparent in high definition action programming. "You jump out of your seat because it is really loud," Marquis said.
Politicians have taken on the issue in a head-on manner. It seems odd that politicians would spend time on this issue in the context of the vastly more important work that they have before them. But subscriber feelings are strong - one of the most common complaints received by the FCC is that commercials are too loud - and politicians are always eager to score points on issues that clearly are easier to deal with than dicier matters, such as the environment or the war in Afghanistan. This is doubly true in an election year.
One politician who apparently sees it that way is Anna Eshoo (D- CA). She introduced The Commercial Advertisement Loudness Mitigation (CALM) Act. (H.R. 6209) in June of 2008. Those who don't like the thousands of pages that comprise a typical Congressional bill should be fans of CALM: The entire bill is about two pages long. It gathered 90 co-sponsors and passed last December.
Senator Sheldon Whitehouse (D.-RI) has introduced companion legislation in the Senate. The bill has passed the Commerce Committee, but is awaiting action by the full Senate. There are minor differences between the two. Thus, if it does move forward and passes, the two must be reconciled before being sent to President Obama. This can be done via a conference committee to work out the differences or a move by one body to adopt the exact language of the other.
Thus, new laws could take effect in the very near future. To an extent, the very brevity of the bill had a galvanizing effect on the industry. The fear was that the mandate to the FCC was so simply stated - essentially, "there's a problem, go fix it" - that the industry feared that solutions would be imposed upon it.
Eric Conley, the Vice President of Video Network Monitoring for Tektronix [www.tektronix.com], said that Comcast was the driving force behind a working group aimed at figuring out a solution that satisfied the legislation - and, by extension, the FCC - in a way that was operationally sound and financially feasible for cable operators.
"If [the FCC was] to come up with something, it may be extremely expensive and impractical," Conley said. "The goal of Comcast and the rest of the working group was to put together a solution that worked, and propose it." Conley had been the CEO of Mixed Signals, a cable test and measurement firm that was acquired by Tektronix in May.
The one piece of guidance in the CALM act was that within one year of passage the FCC would use the A/85 best practice recommendation developed by the Advanced Television Systems Committee (ATSC), an organization that develops voluntary standards for digital television.
To an outsider, it might seem that keeping an explosion in a war movie from shattering subscribers’ ear drums and otherwise controlling audio levels is a relatively easy job – at least compared to turning a cable network into a phone system, delivering 3D television and some of the more exotic challenges that cable operators have surmounted on a routine basis during the past few decades.
That’s true – and misleading. The actual job of maintaining audio levels on a single or limited number of adjacent channels is not too difficult. The complexity comes when the process has to be replicated on a 24/7 basis for hundreds of channels.
The approach to controlling audio levels starts with a standard called BS.1770-1, which is a product of the Radiocommunications sector of the International Telecommunication Union (ITU-R). Part of BS.1770 is an equipment monitoring standard called LKFS, said ATSC President Mark Richer. The goal is to create a consistent way to translate loudness and other audio readings and metrics into a true measure of volume as perceived by listeners.
Finding the LKFS enables monitoring gear to identify audio that has an unacceptable deviation from what is known as the dialnorm, which is a Dolby Laboratory measure at which audio encoders generally are set.
In a way, the process is similar to the mean opinion score (MOS) used to judge telephone network voice quality. “You have to model human experience,” Richer said. “We’ve made recommendations on loudness measurement which includes target loudness levels and peak values. We made recommendations on audio monitoring and set up and methodology to control interstitial loudness, which is the loudness [differences] between programs and commercials.”
Finding the LKFS and the comparing it to the dialnorm across hundred of channels is the challenge – and the vehicle by which the industry’s long standing volume problem most likely will be solved.
Solutions from Tektronix and Miranda
Tektronix and Miranda both offer solutions to the challenge. Conley said that its Sentry audio and video monitoring line keeps tabs on all programming from a wide variety of sources, including IPTV.
Its volume level control functions – which in many cases can be added to units that don’t have it via software upgrade – includes tracking of overall volume, dialnorm and channel history, and sends alerts if volume is outside of parameters. When that happens, he said, programming can be re-encoded or other steps taken. It is not uncommon, of course, for producers of commercials to record them in a way that is most likely to make them louder than the other programming. Sentry offers a way to squelch that little trick.
Miranda’s Marquis said that the company has developed automatic loudness control circuitry in association with Linear Acoustic [www.linearacoustic.com] and Jünger [www.junger-audio.com], a German firm. The technology is a standard feature on its XVP-3901 family of audio and video converters. The primary goal of the family of products, Marquis said, is to do such things as convert content from digital to high definition.
Mitch Askenas, Miranda’s Director of Business Development, says that the EVP addresses loudness within a channel by decoding the programming stream, correcting loudness discrepancies and re-encoding. Currently, Marquis said, a discreet XVP-3901 is needed for each channel. Marquis said that the company is working on a higher density approach, which Askenas says should be introduced early next year.
A separate Miranda product, the IRD family, essentially keeps program suppliers honest. The dialnorm part of the metadata – information about the programming stream – that accompanies the programming. In some cases, the dialnorm number in the metadata leads misleading and in essence tricks the STB into boosting the volume beyond the surrounding programming. The IRD makes sure that the dialnorm numbers are accurate. Since no decoding and re-encoding is necessary, one IRD card can support six programming streams, Askenas said.
The bottom line is that the solutions to the volume issue don’t seem to be out of reach. What is necessary is motivation to develop and harness them. This clearly is emerging as competition increases – and subscribers are less willing to put up with annoyances from their service providers – and legislative pressure builds.
“This is a problem across all industry segments,” Richer said. “It is really important that those in the industry settle this problem, regardless of the [fate of the] CALM Act. It really does bother some customers.”