The VCR. I wasn’t an early adopter, or an extreme user. But the VCR indeed played a featured role in different stages of my life. It helped me keep my standing date with the TV show Thirtysomething back in the 1980s. In the 90s, it played Barney & Friends over and over when that was the only relief my then-two-year-old could get from the pain of the Coxsackie virus. And for at least a decade, it let my mother across the country see her grandbabies take their first steps, sing their ABCs, and reach all sorts of other milestones she would have otherwise missed.
I have professional memories of the VCR as well. I visited the source of the video recording, Ampex, just once, though I drove past its iconic sign for years. That visit was a true pilgramage: The VCR wasn’t the only thing that came out of that company; so did Ray Dolby, Nolan Bushnell, and a host of other engineers who changed the entertainment industry. Over the years, I also met with engineers at Sony and JVC, the companies that brought the videocassette recording to the mass market, and whose battle over consumer VCR formats is legendary. I told the story about the development of the VCR on the occasion of IEEE Spectrum’s 25th anniversary in 1988, recalling how the engineers at Sony and JVC made different design choices along the way to the same goal. You can read that article here.
I still have a VCR in my cabinet and a handful of favorite obscure movies that never made it to DVD sitting on a shelf. And so this month’s news—that the last-known company to continue to manufacturer VCR hardware is stopping production because it can no longer get parts—sent me gazing fondly at that now-irreplaceable gadget (my current model is a Go-Video VCR/DVD combo).
But my guilty-pleasure TV shows are available for streaming to my laptop (and watching with headphones) late at night. The kid videos were repurchased as DVDs, and now those are mostly obsolete, thanks to Netflix, Amazon Prime and other on-demand services. My mom left this world before the VCR did; I expect I’ll be livestreaming with my grandchildren.
So, goodbye VCR, and thank you. You served us well.
JVC and Sony transformed an ingenious concept pioneered by Ampex into a major industry
(The following article was published in IEEE Spectrum in a special anniversary issue in 1988)
Consumer electronics companies worldwide felt sure that the public would be interested in a machine that would tape their favorite television programs in their absence for replay at home at their leisure. But in 1971, there were no such products on the market for consumers, and there was still some debate over what exactly people wanted. Two companies determined to solve both problems were Sony Corp. of Tokyo and The Victor Co. of Japan, known as JVC Ltd. Yokohama.
Obviously, that product had to include the convenient cassette. In 1962, the cassette had won over the mass market to audio tape recording, which until then had interested only audiophiles. But “the video problem was 10 times as complex as the audio problem,” explained Joseph Roizen, a former Ampex Corp. engineer who is now a consultant for the television and video industries.
Video signals range up to 4.2 megahertz and contain far more information than audio signals, with their 20-kilohertz maximum. An audio tape is simply pulled past an immobile recording head; but most videocassette recordings use helical scanning, with the tracks running at a diagonal across the tape and with the tape typically spiraled around a rotating drum with two or more recording heads on it. Therefore, unlike audio tape, which is left in the cassette and simply moved past the recording head, videotape must be literally pulled out of the cassette and wrapped around the drum, without ever slipping out of position.
By 1971, several companies had already built videotape players that used some type of cassette and tried to sell them to consumers—but failed. Ampex, of Redwood City, Calif., had briefly attempted to develop a product called InstaVideo that used tape cartridges. (A cartridge has only one reel, the supply reel, the take-up reel being built into the player, whereas cassettes have both supply and take-up reels built in.)
The InstaVideo (also called InstaVision) project died soon after it was brought to market. One of its problems was the cartridge, which was less reliable than a cassette and sometimes frustrated users. The group also could not get the cost down to a reasonable consumer price. Another problem, explained Roizen, was that Ampex had earned its reputation in the professional video realm, so that the sales force never seriously marketed the InstaVideo product, nor did distributors and retailers perceive it as a supplier of consumer video products.
A consortium of New York City businessmen with no experience in consumer electronics formed a company called Cartridge Television Inc. to launch a cartridge-based consumer video recorder—Cartrivision. The group spent huge sums on marketing and advertising but went bankrupt when tape problems necessitated a short recall. (For several years afterward, enterprising engineers were buying the unpackaged guts of the units for less than $100, packaging them, and reselling them.)
CBS Inc., in New York City, tried a different approach: a film cassette for home viewing of theatrically released movies, called EVR. This format could not record, however, and consumers were not interested. (Many of these failed formats are displayed at the Ampex Museum of Magnetic Recording, Redwood City.)
Sony meanwhile had developed the U-format or “U-matic,” a cassette-based recorder—in collaboration with JVC and Matsushita Electric Industrial Co., Osaka, and with licenses from Ampex—and had introduced it as a standard for VCRs in 1971. But the $2000 recorders and the $30 cassettes (in 1988 dollars about $6000 and $90) were big and heavy. The VCR unit measured 24.25 by 8.125 by 18.125 inches (61.6 by 20.6 by 46.4 centimeters) and weighed in at 59.5 pounds (27 kilograms). Consumers were again unimpressed, and the companies quickly retargeted the product to the educational and industrial training markets, where the U-format proved popular.
Smaller and cheaper
But as consumer product companies, neither Sony nor JVC was satisfied with the limited educational and industrial markets. They knew that to appeal to consumers they had to develop a VCR that was both smaller and cheaper than the U-format.
The companies hoped to work together to establish a standard for helical-scanning videocassette recorders using tape that was half an inch (12.5 millimeters) wide, which, said Roizen, “they were gong to flood the world with.” They easily agreed that the tape width should be reduced to a half inch, rather than the three-fourths of an inch. specified in the U-matic design. Then the trouble started.
Masaru Ibuka, the founder of Sony, who in the early 1950s had decreed that his engineers were to design a transistor radio the size of a man’s shirt pocket, came into the Sony offices one day, tossed the company’s employee handbook onto a table, and told his employees that the target for their VCR project was to be a videocassettes smaller than that handbook. The size of a standard American paperback (150 by 100 by 15 mm), it was to hold at least one hour of color video, he said.
Meanwhile, the then general manager of JVC’s Video Products Division, Shizuo Takano, decided that it was time for JVC to come up with a worldwide standard for home video. To get things going, the general manager of JVC’s Research and Development Division, Yuma Shirashi, drew up a matrix of requirements that was not quite as simple as the size of a paperback.
One key requirement of the system was a “more-than-two-hour recording capacity” because he noticed that movies and sporting events typically lasted two hours.
Sony showed a prototype of its proposed Betamax format VCR to Matsushita and a few other Japanese companies in 1974. According to Japanese trade paper, the chairmen of Sony and Matsushita met in secret late at night on the subway, with the Matsushita side arguing that it had found a way to get two hours of playing time on a cassette only a third bigger than a paperback book, with the Sony side unyielding on size and unwilling to go to a lower playing speed, which would make high picture quality harder to achieve.
Both Sony and JVC claim that their original VCR models had offered 240 lines of horizontal resolution and a signal-to-noise ratio of about 45 decibels. Frank Barr, who tests video products for Advanced Product Evaluation Laboratories in Bethel, Conn., said that at the top of the line, the early Betamax models by Sony had a slightly better resolution and signal-to-noise ratio than JVC’s early VHS models. One reason for this slight difference lies in the selection of carrier frequencies‑the VHS carrier signals fall between 3.4 and 4.4 megahertz, the Betamax signals between 4.4 and 5.g MHz, the greater bandwidth allowing higher resolution. Though this difference was almost indiscernible, it led videophiles to recommend Betamax as the ultimate format, Barr said.
After discussing the matter for about a year, the companies still would not compromise their primary design goals‑paperback size versus two hours playing time‑so they decided to go their separate ways. (A Sony spokesman told IEEE Spectrum that, to this day, “Quite frankly, it is our believe that the VHS format was realized only after a thorough study of the Betamax system.” JVC, on the other hand, said that VHS was an independent design effort based on the matrix of goals drafted in 1971, and that when the company saw the Betamax and what JVC viewed as its fatally short recording time, its own product was only about 18 months from going into production.)
Whatever the real story may have been, Roizen said, “The monolithic Japan Inc. was split.”
In addition to tape width, the companies were agreed on the use of helical-scanning technology. In audio tape recoding, the recording head stays put and lays a longitudinal track on the moving tape. In early professional video recording, four heads on a rotating drum laid tracks directly across the width of the tape.
With the quad format, as it was called, information could be more densely packed then with the longitudinal format; also, because the tracks were so short, problems with tape stretching were reduced. On the other hand, one track could not hold all the picture information in a frame, which was therefore separated into 16 tracks, with each track read by one of the four heads on the drum. Differences in head quality and alignment led to banding on the screen or “venetian blind” effects.
Helical scanning, which warps the tape around the drum at an angle, like a candy cane’s stripes, has the advantage of quad recording—reducing problems caused by tape stretching—but not its drawback—each slanted stripe can carry a full frame.
Going to a ½-inch tape in a reasonably small cartridge required a number of technological advances that, working together, reduced tape consumption from approximately 8 square meters per hour for the U-format to approximately 2 m2/h for the VHS and Betamax units (the writing speed of VHS is slightly lower than that for Betamax: 5.80 meters per second versus 7.00 m/s). For one thing, advances in IC technology made by Sony and other companies allowed VCRs to produce a better picture with less noise (the signal-to-noise ratio in the U-matic was 40 dB as against the 45dB claimed for the first Betamax and VHS recorders).
Moreover, improvements in video heads reduced their gap size by about a factor of 10, to 0.3 micrometer, allowing the tracks they wrote and read to be smaller and thereby increasing recording density. Also, advances in magnetic tape (specifically, the use of a cobalt alloy for the magnetic coating) increased its sensitivity and made it possible to pick up very short wavelengths.
Changing the guard
Besides the industrywide advances in IC, head, and tape technology, Sony and JVC found means, albeit slightly different, of adapting to their products another recording technique that increased information density.
A technique called azimuth recording had been used in black and white videotape recording since the late 1960s. In azimuth recording, the video heads are mounted at angles—tilting one to the left and one to the right—from the perpendicular to the run of the tape. Because the tracks recorded by these heads are not parallel to each other, the heads are less likely to pick up the wrong signal.
Sony tried to apply this technique to color video recording in the U-matic, but it did not work. The color signals, which use lower frequencies than black and white signals, interfered with each other, and Sony had to leave blank spaces of tape as guard bands between the video tracks.
A researcher at Sony, Nobutoshi Kihara, continued to work on this problem even after the U-matic went into production. He developed a phase-inversion system, recording the color signals on adjacent tracks 180 degrees out of phase with each other, to eliminate interference between the signals.
JVC meanwhile came up with its own solution—a phase-shift system, recording each color signal 90 degrees out of phase with adjacent tracks. Both solutions let the companies eliminate the tape-wasting guard bands, and both were patented, Sony’s in 1977 and JVC’s in 1979.
M for manufacturability
While Sony was content to duplicate in its Betamax the U-loading mechanism developed for the U-matic, JVC instead used a system it called M-loading. JVC says that M-loading made the machine easier to manufacture, more compact, and more reliable, because the tape guide poles did not contain moving parts. Sony argued that M-loading was not superior and that U-loading only looked more complicated, whereas in reality it was a simple mechanical apparatus and indeed better than M-loading because it reduced stress on the tape (an M-loaded tape wound around two sharp turns, a U-loaded tape wrapped around one pole only).
Others say that both U- and M-loading solved the same design problem, and neither had a major advantage.
With U-loading, a single arm reaches into the cassette, pulls out the tape, and wraps it around the head. With M-loading, two arms, on either side of the recording head, grab the tape and pull it against the head, the arms traveling a much shorter path than the U-loading arm.
M-loading allowed JVC’s machine to be more compact than Sony’s—so much so that the unit was half the volume and still left more room between components than the Sony design. U-loading made it easy for Sony to add a picture-search function (fast-forward while still viewing the image) to its design, while JVC had a slightly harder time adding picture-search to its machines. (M-loading as initially designed put so much stress on the tape that the tape could not be allowed to run at high speeds without first being drawn back into the cassette, away from the head. JVC solved this problem by changing the stationary guide poles to rotating guide poles.)
To record for a longer time than Sony, JVC used a cassette tape 30 percent larger in volume and, as already noted, a lower writing speed (5.8 m/s versus Sony’s 7.—m/s). Other things being equal, reducing the writing speed reduces the signal-to-noise ratio. JVC said it overcame this disadvantage by giving the signal a greater pre-emphasis boost, increasing the magnitude of some frequency components more than others to reduce noise.
Increasing the signal in this manner, however, leads to bleeding in white areas of the picture. Accordingly, in the JVC design the signal is also first sent through a high-pass filter to eliminate low frequencies, next has its high frequencies amplified and then clipped to stop the bleeding, and finally has the high frequencies recombined with the low frequencies and clipped again.
Sony offered the Betamax to Matsushita and other Japanese companies as a standard. Toshiba Corp. and Sanyo Electric Co. eventually took them up on this offer. JVC persuaded several other Japanese firms to join it in producing VHS machines. In the United Sates, Zenith Corp. initially joined the Betamax group, while RCA Crop. Went with VHS.
Those consumers that marketers call “early adapters”—the technically literate videophiles with money to burn—quickly committed themselves to Sony’s Betamax because of reports that its resolution and signal-to-noise ratio were better. But since few of them—and hardly any consumers in the mass market—could tell a difference in quality between the two formats, the convenience of longer playing time won out, and today the VHS format is clearly the consumers’ choice, particularly in the United States.
The first models were introduced in 1975 and 1976—Sony’s Betamax SL6300 at 229,800 en ($820 at 1975 rates) before JVC’s HR3300 at 256,000 ($915). Then the two formats began converging. Sony responded to JVC’s built-in clock (for unattended recording) with a plug-in timer module for its original units and with built-in timers in its later models. Sony also introduced Beta II, with two hours of playing time, and JVC responded with JVC long-play, a six-hour format.
Both companies steadily worked to improve their picture through better signal processing, magnetic heads, and recording tape, and both added features such as the ability to program the VCRs for weeks at a time. Today both formats boast five to eight hours of recording time, depending on the type of tape used, and horizontal resolutions of between 400 and 500 lines. (These top-of-the-line models, known as the S-VHS and ED-Beta, are not downwardly compatible with earlier units.)
Fumio Kohno, Sony’s managing director, told IEEE Spectrum: “Competition between the Beta and VHS formats has contributed greatly to the improvement of both. It has also stimulated progress in home VCR technology, such as 8 mm video, and in digital audio tape.”
--Tekla S. Perry
The author wishes to acknowledge the help of Joseph Roizen of Telegen
Magic Leap, the stealthy Florida-based company that has so impressed investors with its augmented reality demos that they’ve ponied up $1.4 billion to date, is beefing up its efforts to create content. The company has established a development lab on the Lucasfilm campus in San Francisco, located, Venture Beat reported, close to the Yoda fountain.
Last week, Magic Leap founder and CEO Rony Abovitz announced a partnership with Lucasfilm to develop Star Wars-related apps for Magic Leap. They’ll be using what the company calls its “Mixed Reality Lightfield” technology, the details of which are still under wraps. That project will clearly be happening in the San Francisco Lab. But the company isn’t betting that Star Wars games alone will be enough to make the technology take off. It plans to draw all sorts of developers into its San Francisco lab to work on applications for the new augmented reality technology.
Magic Leap may have some serious competition for augmented reality developers; with Pokemon Go capturing the imagination of gamers all over the world, it turns out that the AR explosion didn’t wait for Magic Leap to reveal its technology. AR games can be engaging even if you have to hold a phone up in front of you to peek into an AR world. Pokemon Go could be great news for Magic Leap, in the sense that it’s getting players used to AR technology and whetting their appetites for an AR interface that they don’t have to hold out in front of them. Or it could be a challenge: Can Magic Leap create compelling enough content to justify purchasing an expensive new device.
To support the developers it attracts, Magic Leap is in the process of hiring a team to work with those developers. Its job listings on LinkedIn currently include a number of posts that will be part of its “advanced content research group in San Francisco.” These include software engineers, infrastructure engineer, interaction engineer, and technical director. Some of these positions also are listed on the company’s internal hiring site.
There’s good news on the technology job search front, according to preliminary results of a survey by consulting firm KPMG. The firm queried 138 U.S. tech CEOs, and 95 percent expect to increase the size of their workforces over the next three years. Some 55 percent expect to grow at least 6 percent. (The full report will be released in August.)
This growth in tech jobs for humans might have been even larger, however, but some positions will be filled by automation and machine learning systems: about three quarters of the tech CEOs expect automation and machine learning tools to replace at least 5 percent of their sales, marketing, technology, and manufacturing workforces.
Or, as the report from KPMG puts it:
“The majority of technology companies plan to increase their human workforce at least 6 percent over the next three years while adding cognitive systems to create a new class of digital labor that can enhance human skills and expertise.”
Bob Melk, president of job search firm Dice, told me that software engineers in particularl should benefit from this trend towards workforce automation and machine learning. Dice sees, he says, “a high demand for software engineers and software developers, representing 7.5 percent of all jobs posted on Dice.”
Does this mean tech companies are going to start issuing reports on their digital labor force?
Last Wednesday, I found myself hiking around in the hills above Silicon Valley, tramping down dried grasses and keeping an eye out for rattlesnakes, as warned. I was trailing Adam Wolf and Nona Chiariello as they hunted for a good spot in the Jasper Ridge Biological Preserve, 1200 acres of chaparral, redwood forest, and grassland owned by Stanford University, where Chiariello serves as a staff scientist. Their mission: to find the perfect location in which to pound in a metal post and mount the latest gadget to join the Internet of Things: Wolf’s Pulsepod.
The Pulsepod is a sleek puck, about the diameter of a dessert plate, packed with sensors and communications gear, intended to watch plants grow (literally) as it simultaneously collects data about the environment around them. Wolf and his startup, Arable, intend to market the gadget as a $500 replacement for $10,000 weather stations with $5000 net radiometers. They expect the first users will be agricultural researchers and specialty crop farmers eager to monitor microclimates and plant growth in order to predict both long term effects of the environment on plants and to make short term decisions, like when to water and when to harvest.
Google has been back in the news in the past week about its diversity—or rather, its lack of diversity. Last Thursday, Google released its 2015 demographics, updating diversity numbers made public two years ago in response to public pressure. (That pressure was possibly kicked off by a blog post by then-Pinterest software engineer Tracy Chou.) The numbers weren’t good then, and they haven’t gotten much better: Women now make up 19 percent of the technical workforce, up one percent from 2014 and two from 2013; Hispanics make up 3 percent and African Americans 2 percent—those numbers haven’t budged.
The stats released gave no age demographics; separately last week, a motion to begin the process of turning a pending age discrimination suit against Google into a class action lawsuit was filed in a federal court. The lawsuit pegs the median age of a Googler at 29, and, if certified, the class action suit would be open to anyone over 40 who interviewed in person for an engineering job since 13 August 2010.
“About two years ago, I attended a Women Who Code meetup in Washington, D.C.,” says Veni Kunche, founder of Code With Veni and a senior software developer with the U.S. Geological Survey. “I was blown away by the support that this community was providing to women who were just getting started in their journey in tech. I wish I had such support when I started my career as a software developer in 2003.”
The career path of a woman in tech, Kunche thinks, can be lonely. “Sometimes being in a room full of men whether it’s at work or a in a classroom can make you feel like the odd one out, as though you don’t belong there,” she says. “It is important for women to know that they are not alone.”
There’s an ethos in Silicon Valley—people who have started a company give back by helping the next round of people who want to know how to start a business. The experienced entrepreneurs answer emails from newbies, take their phone calls, and give what advice they can.
For the past nine months, the Nasdaq Entrepreneurial Center has been trying to make it easier for the newbies to learn how to start a business, and the veterans to pass on their knowledge. It’s a San Francisco-based nonprofit started by the Nasdaq Educational Foundation and supported by Nasdaq, which owns the eponymous electronic stock exchange in the United States and a number of other exchanges around the world. The Nasdaq Entrepreneurial Center is offering a steady stream of group classes in things like startup law, crowd funding, diversity, PR training, founders equity, employment law, pitch prep, and mediation, along with individual mentoring, to just about any would-be entrepreneur who wants it. The main requirement? That you are working on your startup full-time, that is, that you are going all-out to turn your idea into a business.
Where are all the women engineers? That’s a question that engineering educators and recruiters have been asking themselves for years now. Twenty percent of engineering graduates are women—but only 13 percent of the engineering workforce is female. It’s not a pipeline issue; engineering schools have been graduating women for a long time now. It’s been easy to blame women leaving the engineering workforce to balance the demands of family, but is that really it? Many have said there’s a culture problem, but what exactly does that mean?
It turns out, according to a recent study, that at least a big part of it happens when women are in mixed groups that need to divide chores. Typically, the division pushes the routine or boring work at the women, the challenging or interesting work at the men.
Researchers Caroll Serron at the University of California at Irvine, Susan S. Silbey at the Massachusetts Institute of Technology, Erin Cech (who performed the research at Rice University but is now at the University of Michigan), and Brian Rubineau at McGill University conducted the study to try to get a better idea of just why women who made it through years of STEM education start migrating out of tech. The effort, funded by the U.S. National Science Foundation, looked at 40 undergraduate engineering students, male and female, spread among the four schools. The students were asked to write in diaries twice a month, from freshman orientation through graduation, about anything of interest in their lives. The researchers also conducted interviews with 25 students at each school during years one and four.
In their paper, published in the May issue of the journal Work and Occupations, the researchers say the problem starts early—the first time engineering students are asked to work in teams. About the women, they write, “Their first encounter with collaboration is to be treated in gender-stereotypical ways.”
Here’s one example they gave from student comments:
There was this one case where, in our design class, two girls in a group had been working on the robot we were building in that class for hours, and the guys in their group came in and within minutes had sentenced them to doing menial tasks while the guys went and had all the fun in the machine shop.
Once the women get out of the classroom, the researchers say, moving into internships and summer jobs, the problem continues. The women are assigned routine tasks while their male peers are offered more challenging opportunities. One example from a diary:
So I’m two weeks into my research position and for the first time in my “working career” I’m really enjoying what I’m doing. The last two summers I’ve been working in an engineering internship position at X, the military defense government contractor . . .. The environment was creepy, with older weirdo man engineers hitting on me all the time and a sexist infrastructure was in place that kept female interns shuffling papers while their oftentimes less experienced male counterparts had legitimate “engineering” assignments.
Ironically, the researchers noted, recent curriculum changes in engineering schools may be making the problem worse, not better.
“Many engineering programs have introduced a greater emphasis on design and team-based learning in the classroom, in essence mimicking and modeling the worksite, not only because it is arguably more creative and effective work practice but also because it is assumed that this will complement women’s social talents and enhance their opportunities for persistence in the field,” they wrote. “We find, however, that a gender differential in students’ professional role attachment tends to be produced in exactly those collaborative encounters in team-based design projects.”
Can this be fixed? The researchers suggest that engineering schools should consider “directed internship seminars” as one possible tool for ensuring that student internship experiences “are dissected to help people learn from the problems women face.”
In the meantime, is there anything women can do to fight the problem themselves? Valerie Coffman, chief technology officer at on-demand 3-D printing and prototyping firm Xometry, says there is. After considering the study results, Coffman told me that, while this kind of thing likely happens all over, she is indeed willing to believe it happens more often in tech. “In situations like this,” she suggests, “you need to be your own best advocate and aggressively seek out the most interesting and challenging projects. If your male peers try to task you with menial work, you can tell them 'no'. They're your peers, not your bosses.”
On Wednesday, in the former San Francisco church that now serves as the headquarters of the Internet Archive, pioneers of the Internet and the World Wide Web joined together to call for a new kind of Web—a decentralized Web. It was a call for change, a call for action, and a call to develop technology that would “lock the Web open.”
And in the audience were the developers and entrepreneurs and thinkers who are going to try to answer that call. These men and women (because the next Web will have mothers as well as fathers), many sporting dreadlocks or tattoos, grew up with the Internet and love the Web, but believe it can be better and are determined to make it so.
This meeting, the Decentralized Web Summit, was part of a 3-day event organized by Brewster Kahle, founder of the Internet Archive, and sponsored by the Internet Archive, the Ford Foundation, Google, Mozilla, and others. It was as much a revival meeting as a tech conference, a feeling enhanced by the rows of pews that made up the seating. There was a lot of fan-boying and fan-girling going on, as the tech leaders of tomorrow buzzed about how they might get this or that luminary to sign their laptops. (Had there been printed programs—there were not—I’m guessing the rush for autographs would have been intense.)