Sony & Philips push two-layer Super Audio CD
The press were shown Super Audio CD at London's Abbey Road Studios in late February. First we listened to a comparison of DSD with three other data formats (all from hard disk): 96kHz/24-bit; normal CD (the same data truncated to 16 bits); and SBM-processed CD. Although the reproduction chain was not optimal, DSD won, hands down. Most striking was how DSD differed from the 16-bit CD, which sounded colored in comparison, with reduced dynamics, bass definition, and even high-frequency content. We also heard the same music on a prototype Philips Super Audio CD player.
In their briefing, Sony and Philips announced plans to license the Super Audio CD to other companies. As of March 1998, licensees of the current CD format would have the option of expanding their agreements to include Super Audio CD at the same royalty levels as are currently being paid for CD. Hardware and software products are planned to be introduced as early as Spring 1999. Other companies that have already shown serious interest in the Super Audio CD include Marantz (owned by Philips), Accuphase, and Sharp.
The licensing strategy is particularly smart, as it smoothes the transition from CD to Super Audio CD for all parties involved. As for marketing, the backward compatibility with normal CD is also a very strong point: new titles need be brought out in one format only. You can buy a new Super Audio CD and play it on your existing equipment, knowing that when you've collected enough of them to make it worth buying a Super Audio CD player, you'll be able to enjoy the higher audio quality (and surround possibilities, if that is your drift) hidden in the second layer. You won't have thrown your money away; you'll have invested.
By contrast, it appears that DVD WG-4 has still to come up with a proposal that includes backwards compatibility with CD. Should they comply, royalties would have to be paid to Sony and Philips for using the "Red Book" CD format; these would come on top of the royalties already being paid to use the DVD patents.
As yet hardly discussed in public is the so-called "watermarking" proposed by Philips and Sony aimed at reassuring the record industry, whose main preoccupation with a new-generation medium appears to be combating piracy. It seems that there are plans for three levels of watermarking, two of which look transparent to sound quality.
The first type of watermark would be a visible pattern generated by modulating the width of the pits representing the 1s and the 0s. The width-modulation technology necessary to create this pattern will not be made available to CD bootlegging plants, so a "plain" disc would not show the pattern---a bootlegged copy would be obvious to the buyer.
The second type of watermark consists of packets of code hidden among the program data but disregarded when the music signal is reconstructed for playback---like conventional PQ codes. Unlike PQ codes, the watermark packets are inserted following a key scheme known only to the manufacturer and therefore impossible to locate by third parties. The code can also be used to block replay on unauthorized equipment.
The third and suspect type of watermarking tries to hide a set of tones or narrow noise bands under the masking curve of a musical signal. Supposedly this cannot be detected by the ear, but special equipment would be able to filter these out and recognize the code that the set of tones or noise bands stands for. The amplitude of these tones/noise bands is so high that they will easily survive compression, analog taping, or even broadcasting. This would enable the record company to automatically log broadcasting of all new titles and register royalties due. The audibility of this third scheme is being examined by TNO, a scientific semi-governmental body in the Netherlands.
Time and time again we have discovered that the human ear can hear things we previously thought it couldn't. Copycode has been stopped. So should this third type of watermarking, especially with such a promising, very-high-quality audio format as DSD.