Loudness: There's Confusion in Diversity


Loudness keeps coming up. After a declaration back in June that I wasn't gonna talk about loudness 'n levels anymore, I've since devoted myself to my half-Zen Garden as well as to a review of the fundamentals of compression. Then I attended the 127th AES Convention last month, and got sucked right back in. Dang!

This time it was a Broadcast Session titled "Loudness and Audio Processing for Broadcast," chaired by Dom Bordonaro of Cox Radio. Many of the usual suspects were on hand as panelists. Andrew Mason was there from the BBC; Jim Starzynski from NBC Universal; JJ Johnston, now with DTS Inc.; Ken Hunold from Dolby; Thomas Lund from TC; Tim Carroll from Linear Acoustic; Marvin Ceasar from Aphex and Frank Foti from Telos-Omnia-Axia.

They had plenty to say, and it was all good. The problem was, it was all diverse. I get the feeling I've heard most of it before; the problems haven't changed much (and certainly haven't been solved). We are still tangled up in the confusing and less-than-reliable relationship between levels and loudness, and we're still annoyed by commercials. The big news (for me, anyway—I've been dozing by my halfZen pool most of the summer and am kinda out of it), is that the Feds wanna get involved.

WHAT EVERYBODY SAID

I won't bore you with a blow-by-blow report of what was said, but:

  • • Mason discussed issues with commercials and program levels, the implications of various new meters, and the ITU-R BS 1770 algorithm for loudness and so-called "true-peak" levels.
  • • Starzynski talked about progress at NBC, and stated that "the industry is on the cusp of a solution." He talked about the Feds, taking the position that their interventions are unnecessary (more on that later). He suggested a reference level of –24 LKFS +/– 2 dB with no metadata, noting that we need buy-in from all players.
  • • Johnston suggested (a) that weighting filters can't get it right, and (b) the time parameter is also a problem. He asked if the problem was loudness or annoyance, and questioned our ability to measure annoyance and listener expectation.
  • • Hunold reviewed the problem as it exists for digital TV, suggested that there is a dB "comfort zone" and that metadata was still a viable way to maintain that comfort zone.
  • • Lund talked a lot about BS 1770 and TC's work on some of the underlying aspects of it, introducing the notion of a "Center of Gravity" of loudness and a Consistency measure. He briefly demonstrated TC's LM5D meter, which looked really cool to me. He also discussed inter-program level jumps, which are the big problems that I personally am most bothered by.
  • • Carroll thinks the proposed Federal legislation will pass, and that we may get dumbed-down audio as a result. He implored us, as industry players, to work hard to protect a format that will work for us. He thinks that we, in the United States, need to work toward BS 1770.
  • • Ceasar talked about production and the need for quality. He asked us to reduce listener fatigue and improve intelligibility. He advised us to "use our ears." I agree completely.
  • • Foti discussed some of the issues as they apply to radio broadcasting.


Whew! My apologies for any misrepresentations and incompleteness. I was scribblin' fast.

KEEPING CALM

The recent AES convention sposored a panel on "Loudness and Audio Processing for Broadcast." The panel included, back row (L-R) : David Bialik, Dom Bordonaro, Andrew Mason, James Johnston, Frank Foti, Tim Carroll, Marvin Caesar Front Row: Jim Starzynski, Ken Hunnold, Thomas Lund The big bulge sticking up in the center of the rug is the CALM Act (Commercial Advertisement Loudness Mitigation), legislation proposed in the House of Representatives to have the FCC regulate the relative level of commercials. In our less-than-entirely-successful struggles to manage the levels problem, we've managed to create a climate in which such a law seems to some like a good idea. Naturally, we have our doubts about the wisdom of such a thing, preferring to struggle on our own.

The bill itself, in its clarity and brevity, is more than a little scary. A definite crowd-pleaser. Excessively "strident" commercials will become illegal. Overly "noisy" commercials will become illegal. No definitions or mechanics were included. The FCC will simply take care of it all, within a year. What's not to like? (Don't answer that.)

WHAT NEEDS TO HAPPEN

Sometimes perfection becomes the enemy of "good enough." Sometimes a huge range of excellent ideas, improvements and fresh approaches just tangle us up and drag us down.

We are not talking rocket science here. We need to keep levels reasonably consistent, both during programs and between programs. That's all.

Starzynski described how the NBC Olympics production team was able to stabilize and maintain levels around the world, so that it became a non-problem for them, in spite of the production complexity.

However, we've made it harder for ourselves by insisting that we want it all to happen automatically, without human intervention. Even so, it's still not that hard.

Way back in 2003, I auditioned a nice piece of gear from TC that seemed to do the job very nicely (the P2 Level Pilot).

It can be done. We've got the technology to do it. We know how to do it. All we've got to do is do it.

As you may recall, I did my own informal study of commercial and interchannel level jumps early this year. I wrote at the time: "None of the commercials had outrageous audio. None of them were hypercompressed, or offensively bright or bass-heavy. They were noticeably louder than the newscast, but not offensively so. There may be problems for some set-top boxes with dynamic range control invoked, especially if dialnorm is not set correctly. Further, there is an obvious generic difference between the audio for commercials and the audio for programs, which is probably as it should be.

Some commercials are offensive for their hyped-up in-your-face quality. Such commercials are not usually encountered as part of 'the evening news,' [which is where I did my measurement]. I suspect that on the Speed Channel, or the kind of channel that specializes in WWF and related sports mayhem, we may encounter commercials that are a lot, ah, "perkier."

Because of that work, I'm a little concerned about Congress's fascination with noisy and strident commercials. They love to demonize stuff, but can be a little cavalier about test and measurement integrity.

Further, the relationship between loud and annoying is a very poorly defined one and should be approached very cautiously. It actually really might be better if we took care of our business ourselves.

With that, thanks for listening.

Dave Moulton is a pretty level-headed guy, and normally he is really dialed in. You can complain to him about anything at his Web sitewww.moultonlabs.com.

Dave Moulton