Watermarking: the Devil's Work!

I left you last month 104 miles from Santa Fe, New Mexico, heading east on I-40 accompanied by a dog and two cats, with 1946 miles to go to reach Stereophile's new editorial home, New York City. To cut a long story short, I did arrive in New York (covered in dog and cat hair). After a nerve-wracking delay, so did our furniture. We will be living out of boxes for a while chez Atkinson, but that's a mere inconvenience compared with the Great Adventure of setting up a new listening room.

I had been in the Santa Fe room 12 years. It had settled around me like a comfortable old shoe. I knew every acoustic cranny, every acoustic nuance. No, it was not a perfect room, but drop a new pair of speakers or a new amplifier into that room, and I could quickly zoom in on what was different about the sound and describe it to you.

But now I have a new room.

How bad can that be? As I write these words, I flash on the angst Michael Fremer and Shannon Dickson recently went through in dealing with new acoustic environments, even when they were definite improvements over their old rooms.

While the house we bought in Brooklyn didn't have a suitable room ready and waiting, we had two basement rooms knocked together before we arrived to create a quite nicely proportioned space. I started modeling its acoustics on my PC, but then people started telling me that, with perhaps the exception of the RPG software, the various programs are not all they're cracked up to be. So, after getting this issue off to the printer, I'm going to put some carpet down, get all the books, LPs, and CDs out their boxes and into the waiting bookshelves, and set up my system. I'll keep you posted on how the process of settling into the new listening space develops.

Elsewhere in this issue you can read reviews of products that I think signal the different directions in which high-end audio can go. Long a two-channel kind of guy, Larry Greenhill explores surround-sound for music in his report on a five-channel Bryston amplifier. I report on my experiences with a "convergence" product, the CardDeluxe PC soundcard from Digital Audio Labs, which appears to sonically sacrifice nothing compared with high-end digital audio components. And Jonathan Scull listens to the third Super Audio CD player to hit these shores, the Marantz SA-1.

Both SACD and DVD-Audio—the latter finally reached the US in July—are exciting formats. CD was intended from the outset to be the replacement for the cassette, not the LP, and was specified, at least in my opinion, with a "good enough" mindset. With these new media, consumers at last will have access to a sound source that has not been compromised, that holds the promise of being transparent to all listeners with all kinds of music under all conditions. This has been a dream of mine since the 1960s.

But even as this dream is within reach, record-industry bean counters seem determined to take it away. As I retrieved my e-mail and browsed the news stories on www.stereophile.com at the end of each day of my transcontinental drive, I read about how, with DVD-Audio in particular, the music signal will be "watermarked" so that someone can extract the copyright information from the sound. An analogy would be those translucent logos that TV channels put in the corner of the screen—so you can be reminded, for example, that it is A&E that is bringing you Poirot.

As I explained in this space several months ago, because the watermarking code needs to be retrievable after the music data have been converted to analog, and perhaps transmitted over an AM radio link, or even compressed with MP3, the code has to be robust; ie, high in level. And if it is robust, it might be audible. Now, as David Leibowitz of Verance, the developers of the watermarking scheme adopted by the recording industry for DVD-Audio, pointed out in the August Stereophile's "Manufacturer's Comments," the watermarking is not as obvious as mixing Morse code with the music. The low-bit-rate (20bps) code is disguised to resemble the analog noise already present; in addition, it is spectrally shaped to get the maximum benefit from psychoacoustic masking. (In very simplistic terms, the code is loudest when the music is loudest in both the amplitude and frequency domains.) In addition, the watermarking doesn't have to be present all the time, nor is it mandatory for content producers to use it on DVD-As. (SACD uses a different scheme that apparently doesn't affect the music data.)

In his letter, Mr. Leibowitz reminded us that in 1999 the Verance code was subjected to listening tests in which pro-audio professionals, including Stereophile's Barry Willis, took part. These experienced listeners found it very hard to detect the watermarking under blind conditions (see "Industry Update," October 1999, pp.29-33), though at last fall's AES Convention in New York, at least one engineer who had taken part in the tests said that once the audible effect of the watermarking code had been pointed out, it was possible to hear it every time.

Earlier this year, the renowned English classical engineer, Tony Faulkner, called into question the results of the 1999 blind test ("Industry Update," August 2000, pp.20-22), pointing out that the test methodology was not strictly applicable to hi-rez classical recordings. As a result of Tony's industry-wide lobbying, further tests took place at London's Whitfield Street Studios at the beginning of July, under the auspices of the UK's equivalent of the RIAA, the IFPI.

Tony was one of the listeners. In an e-mail he subsequently circulated to interested parties, he wrote, "I have no doubts in my own mind now that the Verance watermark is clever enough and effectively unobtrusive enough for noncritical low- to mid-fi...ie, up to but excluding DVD-A, SACD, and high-quality CD. The bad news is that it was audible on poor-quality, bandwidth-limited, archive analog material to a 49-year-old engineer with a cold....For audiophiles paying extra money for a new player and for new discs, judging by what I heard yesterday, the watermark could reduce the perceived quality of DVD-A to somewhere between a good MiniDisc and a below-average CD."

As the watermarking is optional, I thought that this would represent a market opportunity for small record companies concerned about quality. If a multinational conglomerate decided that all its DVD-As would be watermarked whether or not the mark was audible, a small company could use the fact that it never watermarked its releases to gain market share.

But on my long drive east, I decided that this passive approach was not enough. As Barry Willis said to me in an e-mail, purists find the notion of watermarking far more offensive than the watermarking itself.

Yes, I do object to the idea of watermarking. If the record industry wants to treat their recordings as commodities, well then, by the shades of Adam Smith and Alan Blumlein, let consumers react accordingly. I call on Stereophile readers to boycott each and every company that releases watermarked DVD-As and CDs. That'll larn 'em to mess with our music.