Compromise Elusive in Debate Over Blu-ray, HD DVD Form Factors

Undoubtedly, the most hotly contested issue is Blu- ray’s use of a 0.1-mm cover layer, below which data reside atop a 1.1-mm substrate. The transparent cover layer’s thickness affects the size of features that can be read. The thinner the cover layer, the smaller the pits can be, and the more data that can be stored. However, the thin layer creates production problems. At first, it was assumed new replication plants would be needed for Blu- ray. But supporters now insist modern DVD plants can be modified for Blu-ray.

A basic single-layer Blu-ray holds 25 GB of data; doubling capacity to 50 GB requires dual-layer construction. Blu-ray road maps predict as much as 200 GB of capacity with an 8-layer configuration. The feasibility of replicating Blu-ray discs -- particularly the 50 GB -- is the debate’s most divisive point.

Toshiba has said it won’t consider adopting the “0.1” form factor without credible, tangible evidence it can be produced in yields and at costs consistent with those of HD DVD production. While Blu-ray officially has remained mum on Toshiba’s doubts, Blu-ray sources say Toshiba has been given the critical data it sought. Additional proof of dual-layer’s feasibility, Blu-ray has said publicly in the past, is that for months Panasonic has been marketing 50-GB rewritable Blu-ray discs in Japan. All the evidence in the world won’t bring about 0.1 commercialization at Toshiba and Warner, which have enormous IP holdings in discs conforming to the existing DVD form factor, Blu-ray sources say. Not so, said Warner Home Video Exec. Vp Marsha King at last week’s Media-Tech Expo in Las Vegas. Her company would endorse 0.1 if its feasibility were proven, King told us.

HD DVD, like Blu-ray, uses a 405-nanometer blue laser but with double-sided discs made like a DVD, by bonding two 0.6-mm halves back-to-back. HD-DVD has been promoted with the promise of production on modified DVD presses. The HD-DVD laser has to see through almost 0.6 mm of plastics, amplifying tilt distortion. To compensate, HD- DVD uses a lens that doesn’t focus as tightly as that of Blu-ray, so widening margins for tilt error. But until recently, this has reduced HD DVD’s storage capacity.

Toshiba last week unveiled a triple-layer HD-DVD with 45 GB capacity (CED May 10 p1). That’s close to the 50 GB promised from a dual layer Blu-Ray disc and regarded by some industry sources as essential for HDTV movie releases. Toshiba also announced a hybrid disc with a standard 0.6-mm, 8.5 GB dual-layer DVD on one side, bonded back to back with a 0.6-mm dual-layer HD-DVD, with 30 GB capacity. By including SD and HD content on the same disc, double inventory at retail could be avoided during the transition from red- to blue laser technology. Theoretically, there eventually could be a double sided, dual or triple layer HD-DVD, with total capacity as high as 60 GB or 90 GB.

Likewise there seems no reason why the 0.1 form factor couldn’t be applied to double-sided Blu-ray discs, made by bonding two 0.6-mm substrates back-to-back, each with a 0.1-mm optical skin. This was mooted in 2002, when Philips proposed a miniature version of Blu-ray called the Small Form Factor Optical (SFFO) system. SFFO initially would store 1 GB on a 3cm disc, rising later to 4 GB from 1 GB per layer per side. Development work on SFFO, ultimately renamed Portable Blue, was frozen at the end of 2003, as miniature hard drive capacity and solid-state memory increased and cost decreased.

The fact that HD-DVD’s 0.6 form factor requires a 3- layer structure to deliver 45 GB, while the 0.1 form factor of Blu-ray can hold 50 GB in 2 layers, underlines a basic, irreconcilable difference between the 2 approaches. The 0.1-mm optical layer is easier to read through, so data pits can be smaller, but the disc initially might prove harder to make -- just as CDs at first were far harder to press than LPs, and DVDs tougher and costlier to make than CDs. By the same token, double- and triple- layer HD-DVDs may prove comparatively more difficult and expensive to make than single- or dual layer Blu-Ray discs -- just as dual-layer DVDs initially cost more to make than single-layer discs. The HD DVD camp didn’t reveal pricing last week, but a source told Consumer Electronics Daily producing the dual-sided, dual-layer hybrid would cost little more than today’s similarly configured DVD-18. Meanwhile, the single-sided, triple-layer 45-GB HD DVD would cost “dimes, not dollars more” than producing today’s dual-layer DVD-9 -- “maybe 3 dimes” the source said. For reference, a single-sided, single layer 15-GB HD DVD costs “pennies more” to produce compared with its SD-DVD sibling.

One scenario for Blu-ray/HD DVD unification harkens to 1995, when the “SD” camp’s 0.6 form factor was merged with MMCD’s EFMPlus modulation to create the unified DVD format. Engineers we polled about the viability of doing likewise -- mixing one camp’s form factor with the other’s modulation or error correction systems -- said unifying Blu-ray and HD DVD in that manner would prove a relatively tall order.

Ultimately, a merged spec’s pit density will be a function of cover layer thickness, said Kevin McDonnell, pres. of Eclipse Data Technologies. His firm discussed Blu-ray modulation at last week’s Media-Tech Expo, but McDonnell said Eclipse also has done HD DVD development work. Either Blu-ray’s “17PP” or HD DVD’s “ETM” modulation could be used in a unified format, but pit size will have the greatest impact on final disc capacity, he said.

Encoding efficiency is roughly equal between the 2 modulation systems, McDonnell said. In the case of 17PP, every 2 data bits are encoded into 3 channel bits, while in ETM every 8 data bits are encoded into 12 channel bits. To allow a read circuit (also known as a “slicer") to recover data from an optical disc, some management is required such that the total pit space roughly equals the total “land” space. In other words, if one were separately to add the lengths of all the pits and lands at any point in the disc, the totals should be nearly equal. Failure to do so causes slicer errors, and therefore read errors. Both 17PP and ETM manage this by providing for frequent “decision points” at which to adapt choices to better manage this balance.

With HD DVD, these decision points occur “worst case” every 1,092 channel bits, while in Blu-ray they occur every 67.5 channel bits. Theoretically, some data patterns can’t be represented in HD DVD (such problematic patterns do exist for CD and DVD), while the Blu-ray algorithm has no such limitation, McDonnell said. HD DVD comes up short in this area due to its emphasis on maintaining as many “DVD-like” attributes as possible, letting new hardware designs heavily use previously developed DVD solutions. Blu-ray started from a clean slate and solved this problem is a far cleaner way. However, the “cost” of the Blu-ray approach requires about 1% more channel bits.

File organization is among the few areas where the Blu-ray and HD DVD systems are “interchangeable,” McDonnell said. But sector organization is another source of disparity, he said. In both formats user data are organized into 2,048 byte sectors; 32 sectors then are joined together and protected by one large error- correction code (ECC) structure. But Blu-ray and HD DVD differ in their efficiency in this area, McDonnell said. A 32-sector ECC block requires 75,712 bytes for HD DVD vs. 76,880 bytes for a Blu-ray, he said. Blu-ray also requires about 1.5% more space to hold the same data. But the face value of this comparison is misleading, he said. The BD structure has a more robust ECC, with better protection against burst errors, so the extra overhead confers a benefit -- required to compensate for errors related to use of the 0.1 cover layer technology. If HD DVD’s ETM modulation ever were applied to the 0.1 form factor of Blu-ray to make for a unified format, ETM’s error-correction robustness would need to be ramped up considerably to compensate for the tighter tolerances, he said.

Were that or other unification scenarios to come to pass, proposed launch dates for either format likely would be delayed well into 2006. But for all the damage a format war could wreak, many believe it would be a delay that’s well worth enduring.