Who Is DVD's One True Master? - digital video disk, includes related articles - Statistical Data Included

Emedia Professional, August, 1999 by Mark Fritz

For a long time now, the question of which media is the best media for DVD mastering has been a no-brainer. Until recently, Digital Linear Tape (DLT) was the only affordable digital media that offered a decent data throughput speed and enough capacity to hold a full 4.7GB disc image.

According to John Town, vice president of Research and development at Nimbus CD International (A Technicolor Company), all 1,500 DVD projects replicated by Nimbus so far have been mastered from DLT. Likewise, Cinram's Kenny Aldridge, product engineering manager, pegs the percentage of projects submitted to his company on DLT at "99.9 percent." Indeed, DLT has become so entrenched as the mastering media for DVD that it is nearly impossible to find a replicator who can directly master from anything else. For most veteran DVD developers (who already own DLT drives), this is an acceptable status quo. But some new upstart developers--especially those working with DVD-ROM rather than DVD-Video--have begun to grumble.

Among the new breed of developers who don't have DLT drives, complaints about the cost of DVD mastering run rampant. After all, they've already shelled out the big bucks for a DVD-Recorder, so being told they also need to convert their project to DLT in order to replicate it is understandably discouraging. Today, with Pioneer's DVD-R drives configured to hold up to 4.7GB of data, DLT is no longer the only game in town, capacity-wise. And with DVD-Recorder prices down to $5,400, it seems likely that more and more developers will be using DVD-R, just as their CD-developing antecedents migrated en masse to CD-R. It also seems likely we'll hear more and more complaints about inflexible replicators who remain committed to DLT.

So why are replicators so stuck on DLT? Replicators answer that question with double talk and buck passing, but also with some good logic. The most truthful answer, says Andy Parsons, vice president of product development and technical support for Pioneer New Media Technologies, is that DLT is used because that's the way it's always been and there's been no compelling reason to change. "DLT is a tried and true, proven media. Replicators want to use something they're familiar with and know is dependable," he says.

CONVENTIONS: THE WAY WE WERE

Although the DVD industry seems barely old enough to have forged ironclad conventions, the use of DLT is as close to a convention as any. The standard procedure for developers has consisted of three steps: (1) author a disc image on a big hard drive; (2) transfer it to DLT: and (3) send the tape to the replicator. At the replication facility, a DLT drive (and only a DLT drive) is connected directly to an expensive (six-figure) Laser Beam Recorder (LBR). Most of these machines use a photoresist plate (akin to a film negative) to create the image that gets exposed to the glass master and is developed in a chemical bath for use in stamping. The large majority of the mastering systems that drive this process come from Doug Carson & Associates (DCA), which got an early start in the business and has since created several de facto standards, including the Disc Description Protocol (DDP) file that provides the LBRs with disc identification information.

To replicators, the step from source media to glass master is very important. "This is a big issue because we're making an expensive glass master from this input media--a critical step," says Nimbus' Town. "Everything else we do downstream depends on that glass master."

With the introduction of first-generation DVD-R, DVD production itself changed only slightly. Many of today's developers agree that DVD-R is the best testing media currently available, saying that other media (a hard drive, for example) just don't come close enough to the final product (a DVD) to serve as a reliable test media. Testing a DVD project on anything other than DVD, they argue, is just approximating. So despite the still relatively high cost of DVD-Recorders, many developers have insisted on using DVD-R for testing. They do most of their authoring on their hard drive, burn the disc image on DVD-R for testing, and then make final adjustments in the next step--the step from DVD-R to finished master on DLT.

But is this last step necessary? Some people think so, some don't, and some are content to follow the prevailing wind.

As Pioneer's (and specifically, DVD-R's) most visible spokesperson, Andy Parsons is probably the last person you'd expect to say nice things about DLT. Yet even he concedes there are logical reasons for using DLT. "Optical media--not just DVD-R, but any optical media--is a sensitive media than tape. If there's a big thumb print on that DVD-R disc and you don't find it until after you've made an expensive glass master, you've basically wasted a lot of time and money," he reasons.

Parsons isn't sure replicators will ever accept any optical media as the primary source for DVD mastering and he doesn't blame them. In fact, he suggests that even if DVD-R media were faultless, it might be a good idea to stick with DLT. "For replicators, going to DLT provides an extra intermediate check, an extra last-minute QA step. And going to DLT provides an extra level of QA for developers too."