MPEG-1 Samples Are Shipping
C-Cube says single-chip decoder is ‘100 percent bug-free’
Hopeful that it can put its reputation for quality problems to rest, and that the quality of MPEG-1 will be sufficient to start the acceptance curve of digital video in the consumer market, C-Cube Microsystems of San Jose, CA, says its MPEG-1 full-motion video decoder chip will be ready for volume shipments in July.
The decoder chip, called CL450, is based on C-Cube’s proprietary RISC (reduced instruction set computer) processor optimized for video, and samples are being shipped now for $250 each. Mauro Bonomi, director of marketing for C-Cube, says that once in production, companies ordering quantities of 100,000 or more units per year will pay $45 per chip.
IS VHS QUALITY GOOD ENOUGH?
It’s widely thought that MPEG-1, a digital compression standard, produced an image that’s approximately VHS quality. Some say that MPEG-2, already significantly better than its predecessor, will be powerful enough to yield much better than broadcast quality video—broadcast quality meaning that it is comparable to NTSC, the U.S. broadcast standard. (Some people define “broadcast quality” as what is the minimum acceptable level for broadcast over mostly substandard receivers, although NTSC is actually much better than most of our TV sets are capable of displaying.)
But Bonomi says part of C-Cube’s preparation for the market with the CL450 has been around the issue of image quality. He says Bell Atlantic solicited an independent test of digital video decoded by the CL450, shown side-by-side with VHS, to between 2,000 and 3,000 people.
Discerning consumers. Bonomi says 32 percent selected the VHS picture as better quality than MPEG, but slightly more than a third selected MPEG-1 as better and about the same number perceived no difference.
Let’s hope they’re right. C-Cube has much to lose by MPEG-1 being perceived as insufficient, and to date many professionals do perceive it is. They say consumers are much more discerning than they are usually given credit for. (Then again, consumers are the ones whose eyes trick them into believing that video is actually better quality if its got a high-quality sound track, so it is probably safe to give C-Cube the benefit of the doubt for the time being.)
IT’S ‘BUG FREE’
Samples have been shipped to Philips and a number of consumer electronics companies in Japan, says Bonomi. Four beta sites have been testing the chip since late April, less than a month. Even with such a short field-testing time, Bonomi is willing to say the CL450 is “100 percent bug-free.”
He’d better hope that claim stands the test of time, too; C-Cube lost a great deal of credibility in the industry by shipping buggy versions of its JPEG chip to NeXT Inc., credibility that it is laboring long and hard to regain today. One story most often connected with the company is about how Steve Jobs purportedly yanked the JPEG chip off the NeXT workstation’s motherboard because he couldn’t get a sufficiently stable version of the chip from C-Cube.
OPTIMIZED FOR CONSUMERS
The CL450 is the first single-chip MPEG-1 video decoder, and it’s been optimized for consumer applications. It provides either NTSC or PAL video output, can decompress MPEG-1 video at 1.2 megabits per second (near CD-ROM data rates), and will be capable of 3 megabits per second as CD-ROM data rates improve. Eventually, Bonomi believes, incremental improvements will yield digital video the quality of SuperVHS.
A custom version, the CL450i, will deliver video to Philips’ CD-I boxes as part of an add-on video module (see Digital Media, Vol. 1, No. 2, p. 3). Though it was not the CL450 that was used at Philips’ awesome digital video demonstration at the March CD-ROM conference in San Francisco, Bonomi says the decoder technologies are identical. He claims CD-I’s video module will produce video of exactly the same quality.
KARAOKE, THE BREAKTHROUGH PRODUCT?
Products incorporating digital video, especially interactive multimedia, are having a hard time winning market acceptance because they lack familiarity.
The traditional analogy is analog versus digital audio: the reason why CD technology rocketed to success was quite simply because music sounded better on CDs than it did on albums.
There hasn’t been a matching scenario so far for digital video, but Bonomi says he believes C-Cube is on the way to creating one via the explosive market for karaoke products in Japan. In fact, he says, karaoke sing-along systems will be “the major source of revenues for the immediate future” for the CL450 chip.
Karaoke is enormously popular in Japan, and less in the U.S. and other markets. Karaoke bars are stocked with equipment that provides both music and generic video that includes song lyrics, onscreen footage of, say, a couple walking on the beach, to lend more of a “music video” feel to the scene.
Right now, the clip video and lyrics are stored on laserdics in giant jukeboxes. All professional karaoke systems installed in bars use the laserdisc jukeboxes, which both take up an enormous amount of space —they store 2,000 to 3,000 songs each —and are very expensive in relation to CDs.
But C-Cube believes — and Bonomi says he has the support of Japanese consumer electronics companies to prove the point — that CD-ROM is much more advantageous technology for karaoke. Not only does it have a much smaller footprint, a major advantage in densely-populated Japan, but it allows true random access and uses a much lower-cost engine.
Bonomi sees a new analogy, and a great market opportunity, in creating systems that make it easy to see the benefits of transferring karaoke from laserdisc to CD technology. “It was easy to create CD titles for the audio market—you just copied them over from analog,” says Bonomi. “The karaoke market presents a similar phenomenon. There are already lots of songs and video on laserdisc. Now they can use our technology to put them on CD-ROM.”
Denise Caruso