There is no longer any doubt that digital cameras are supreme picture-taking instruments whenever the traditional interchangeable-lens single-lens reflex type of camera is called for. With imaging chips ranging from 4 megapixels (MP) to almost 14, they’re capable of the image quality and output options that film cameras are. They provide the greatest on-the-spot control over the picture of any cameras…ever. Fitted with multi-gigabyte memory cards, they can shoot almost endlessly without stopping to reload. And to top it all off, several are selling for under $2,000 now, and even under $1,000, in the case of the latest Canon model.
You may recall in past issues where we described D-SLR models from Canon, Contax, Fujifilm, Kodak, Nikon, Olympus, Pentax, Sigma and others – within their respective levels, all wowsers – and since then the low-priced Canon Digital Rebel and Nikon’s high-speed D2H have joined the field. Looking at this remarkable array of equipment, with in-camera histograms, rapid framing rates, on-site review of the pictures and direct access to printers, we are contemplating a new, and perhaps golden age, of photography.
But wasn’t it all kind of sudden? Ten years ago – heck, seven – the only digicams most people could afford were barely reaching XGA size (1024x768 pixels), were slow on the draw, lethargic once working and were patterned after snapshot cameras. They made great jewelry to adorn high-tech groupies, but were hardly tools for the working pro. That was 1996, and now it’s only 2004. It’s great being where we are today, but, um, how’d we get here?
In fact, considering the stumbles and stutters of its origins, it’s sometimes a wonder the D-SLR developed at all. For a very long time, it embodied more wishful thinking than practical utility. We couldn’t exactly say its beginnings were shady, but they certainly were dubious.
Before the Beginning
It’s not as if engineering an electronic camera was a novelty by the 1980s. There’d been electronic television cameras fifty years before, providing the basic genome for today’s D-SLR. The CCD replaced the TV picture tube during the 80s, taking much less space and making compact designs more attainable – just what you’d want for a still-image field camera. Both Canon and Sony leapt upon the potentials, building high-end, 35mm-style SLRs around CCD-based imaging systems. They were on the market by mid-decade.
But they were off the market by the mid-1990s. Building cameras is one thing. Generating, storing and distributing their pictures is another.
Nowadays we don’t give a second thought to downloading flash-memory cards to PCs and sending the result via Internet to designated storage, display or printing facilities. But in 1986 there were no flash-memory cards, few PCs (none with the power to process photos) and no Internet as we know it.
There were, however, dye-sub printers, offered by Canon and Sony, Fuji and Kodak, among others. Those manufacturers liked the idea of cameras that fed printers that took consumables. Kodak didn’t market a 35mm-style electronic still camera at the time, but they and others supported the alleged advantage of the electronic image – its immediate transmissibility to anyplace telephone lines lead – by selling an assortment of e-photo playback and transmission devices.
The very need for those devices in the first place, however, was the Achilles Heel of the proposed electronic revolution. Those gizmos cost plenty, and the electronic SLRs were already selling in the vicinity of $5,000 (in mid-80s dollars). This was not something to be taken-up by Joe and Jane Foto, those All-Americans who snapshoot their way through their days. For what they’d spend on a complete electronic still-photo system, they could buy a car.
Of course there were the newspapers – who has a greater need for on-the-spot immediacy? USA Today and other leading publishers received demo systems and training from the camera companies, their results being brought to broad attention.
However, the method of making electronic pictures used the broadcast television standard (in the U.S., NTSC), and it had problems of its own. Most confounding, the TV picture was meant to be seen in motion, not as a still-frame. To this end, each frame was composed of two fields, interlaced. At a shutter speed of about 1/60-sec., it was easy for a fast-moving figure to register in two different locations in the two fields. This would be invisible when each frame flashed past in a 30-per-second movie. But in a still-frame, if someone was gesturing, they could easily end up with two right hands.
Besides not being made for the still image, the NTSC picture was not designed to be duplicated. Why should it be? When it was established, all TV was broadcast. You saw it once, and that was that. So in order to broadcast the picture signal without consuming too much bandwidth, the colors were "composited" into a single channel. While true component-color TV, with RGB traveling through three wires, was already the higher-quality darling of in-house video production and distribution, the electronic still systems of the 80s conformed to the more widely-installed broadcast standard.
The electronic still cameras of that epoch wrote their pictures to videotape, or at least a videotape substance shaped into a disk. Resembling the 3.5-inch computer floppy then coming into vogue, the two-inch plastic-encased disk suggested a compatibility between the camera and early PCs. There was much speculation about what the future would bring, once a blazing clockspeed of 10MHz was available at a "pricing sweetpoint" of about $1,500.
Beating the Drum
During a previous era of in-house AV, it had been "AV specialists" who most frequently decided which systems to use for in-house communications. But SV was pitched more to the corporate consumer of AV than to the producer of it – the account executive or catalog buyer who knew more about Asian Fabrics than key-to-fill lighting ratios.
Shown photos made under just-so conditions – flat, almost shadowless lighting, for example, to match the relatively limited contrast range of the NTSC picture – the corporate exec could see pictures that looked almost as good as film. Then as now, close-ups of faces, which portraitists rarely light with great contrast, were handed out like candy to prospective purchasers of the new system.
Similarly, promoters of SV were liberal dispensers of sports photos, taken under stadium lighting already calculated to satisfy broadcast cameras. The promoters were quite correct to say, and the purchasers to believe, that SV could produce decent pictures. It took the AV specialist to say it simply was not the medium best suited to long-shots of an automobile-parts inventory, or any other subject with lots of small details.
The SV-SLR cameras themselves, however, did plenty to reinforce the impression of solidity and endurance of the medium. They were, in every way, handsomely crafted instruments with a very solid feel. The Canon RC-701 and RC-760 (essentially the same camera, the later having the "high-band" signal that migrated from videotape formats like hi-8) to this day have the satisfying feel of a well-made instrument, while the focus and zoom rings of the special-built lenses are as smooth as silk. In materials and finish, the Canon SLR seemed every bit the equal of Canon’s best 35mm. Novel for their possession of a flat top plate – Canon installed their viewfinder prisms in an inverted position – the cameras bristled with innovation and futurism.
Sony’s diverse line of ProMavica models were a bit clunkier, but seemed all the more solid for it. Given Sony’s well-deserved reputation for ruggedness in industrial and broadcast TV cameras, their SV machines cast an aura of seriousness at first glance. One three-CCD model produced component-color within the camera – a feature that would improve picture quality somewhat, though would be reduced to composite color for most output devices. Still, to the target buyer, the term "three chip" sounded a lot like "more horsepower." Sales are built upon sexy one-liners.
A lot of people bought the act. With its bold front, SV convinced a large chunk of the American press that its proliferation was inevitable. Based on such optimism, the press gave glowing appraisals of electronic photography and glossed over the occasional catastrophes that sometimes ensued.
The worst case was a late-80s rollout by the Ford Motor Co., introducing their new models to dealers and the press. The AV producers were told to produce the whole show in SV. Nobody knows the origins of that dictum, so we can only speculate on how it arrived on the management level.
Automotive photography is an art unto itself, however, with crispness and brightness and subtlety of shading a basic expectation. An NTSC picture simply can’t reproduce it. Even multi-monitor videowalls, very "in" at the time, couldn’t save the SV image. The show was a flop, and a lot of people got fired.
Whether SV was a marginal success or a colossal failure depends on your point of view. Arguably, it stimulated further investment and research into the technological fields that eventually spawned today’s wondercams. Arguably, that would have happened anyway, once PCs were powerful enough. Equally arguable, SV’s inability to become "the future" turned people off to electronic cameras and delayed real progress. Arguably, the point SV made/proved was that just because you build it, they won’t necessarily come.
The Digital Dawn
The SV-SLR cameras were beauties, but they took lousy pictures. Their problem had nothing to do with their construction, but with their method of recording. Although presented as futuresque, the NTSC color standard harkened back to the late 1960s, and 20 years later was one of the oldest imaging standards in use.
Besides the high-end SLRs, plenty of consumer-level SV cameras showed up too. Sony’s original Mavicas used the two-inch analog disk, and Canon produced at least four point-and-shoot models that did the same. These weren’t always cheapies – the priciest Canon SV-P&S cost a good $3,000 – but at least one model sold for under $1,000. The best-known of the Canon lot was the RC-250 (also known as the "Xap Shot"; originally the "Zap Shot," and sometimes the Ion), but it was the later RC-570 that could display its own pictures without need for an external player. This obvious but heretofore overlooked improvement extended the camera’s potential, as did a vast system of accessories Canon provided just for their SVs.
Other manufacturers dabbled in SV, but without Canon’s bravado. Kodak offered their SV transmission equipment almost begrudgingly – "if there’s a market," one Kodak rep told us, "Kodak will be in it." Minolta modified the 35mm Maxxum camera with an electronic back and disk drive and produced a prototype SV player, but neither went on sale in the U.S. Olympus produced at least two SV prototypes, one requiring a separate player, one not; neither went on sale. Nikon showed a prototype at the 1987 PMA show, but described it in guarded terms – and did not bring it back the next year. Casio produced a consumer-like model, which looked ready for the market – but we’ve found no evidence that it actually went on sale
The SV system had been introduced in 1982, with the original Mavica camera. While even Kodak executives, in exclusive interviews, admitted being scared to death by it, ten years later SV was an endangered species. Canon kept the faith, as the PC was growing up. Intel machines were reaching the 386 range and 33MHz clockspeeds, and the Macintosh was arriving on a 32-bit bus. Modems were still poking around at 300 baud, but the potential was there.
Canon added a digitizing card for installing in a computer slot, finally turning the SV picture digital. The RC-570, simple as it was, became the first camera priced under $10,000 that could download directly to a computer (which by then could have had an astounding 20MB hard drive). With the under-$1,000 VGA-quality (640x480 pixels) consumer digicam not due till 1996, Canon had much of a burgeoning market to itself.
The first true digital SLR, the Kodak DCS 100, had been shown behind closed doors in 1990, having been produced previously as the Hawkeye for unspecified agencies. It was on the open market by 1991 – a Nikon F3 with a CCD system fit within. It wrote to an external hard drive mounted in a case slung from the shoulder. It was slow and cumbersome, heavy for some, but its CCD contained an astounding one million pixels. For that you could pay upwards of $25,000.
Although widely described as a film-dedicated company dreading electronic imaging, Kodak was actually a pioneer and leader in professional-level digicams. Their DCS 200 was on the market by 1993, its main contribution being an onboard memory-card system (using a PCMCIA Type II-size disk drive), replacing the shoulder pack. By the mid-90s, Kodak was offering three different sensors – including a whopping 6MP model – in both Nikon and Canon EOS bodies.
Minolta joined the fray by 1994 with their RD-175 camera, a Maxxum with a digital back containing three CCDs. Unlike the RGB chips for NTSC component color, the Minolta contained two chips for green (luminance) and one for red and blue. Theirs was also the first in its field to interpolate the image to a larger size for output. For all its individuality, the camera was quite pleasant to work with. It produced pictures that were sharp, snappy and richly saturated, even if having a slight affinity for the purples. We look back upon it as one of the most appealing of that first wave of D-SLRs.
Like Kodak, Fuji was a film-based company that had a history rich in CCDs. Also like Kodak, they questioned the wisdom of building their own camera body, when Nikon was already making such good ones. But where Kodak modified 35mm Nikons for digital use, Fuji put their chip in a body Nikon made special for digital. The camera, whose picture-taking features approximated the F4, bore both companies’ names. It also produced a very creditable 1MP picture.
All these early digicams produced sharp pictures, at least at smaller print sizes. But many of them suffered from a phenomenon called color aliasing. This occurs when some part of the shot – usually a small detail, such as the pattern in tweeds, automobile or air-conditioning grilles, the tips of waves on water, etc. – intersected one or two color sensors on the CCD, but not all three. An odd maelstrom of colored dots could result, intruding on the "photorealism" of the picture. The problem persists to this day in all digital cameras (except the Sigma with its three-color pixels and any three-chip RGB designs that might come along) but goes unseen, after in-camera processing removes the "incorrect" dabs of color.
By 1998, the digital consumercam had arrived while another wave of pro-level cameras made its debut. Fujifilm introduced the DS-300 (followed by a very close cousin, the DS-330). The camera had a non-SLR viewfinder and permanently mounted zoom lens, but it also had a 1MP chip – still professional caliber by the standards of the time. Bearing the look and feel of a professional film camera, the camera made great pictures and was priced under $2,500.
At Photokina 1998, Minolta showed the prototype of their new two-chip professional SLR, the RD-3000 – flat-topped, like the Canon RC-701 – the first interchangeable-lens D-SLR to use lenses other than those built for 35mm. In this case, the optics were borrowed from Minolta’s Vectis APS camera – very sharp lenses designed for a film format closer to the size of CCDs than 35mm frames. Nikon, Olympus, Pentax and most recently Canon resumed the use of smaller interchangeable lenses designed specifically for digital cameras in mid-2003.
Kodak’s efforts in D-SLR design took a giant step forward in 1998, with their DCS-520. Not only did it abolish the appearance of color aliasing, it had a contrast range that was positively film-like – easily nine stops from shadow to highlight areas, with detail visible in both. It still was priced in the $15,000 vicinity, but it was a heck of a picture-maker for those with the money.
The year 2000 was a turning-point in many peoples’ minds, as Nikon brought forth the original D1. As rugged as an F5, it had a robust contrast range in a camera designed from the get-go for digital usage. With a price point around $5,000 (less lenses), it authored in the concept taking wing today: an excellent performer could take digital pictures, at a price a much broader range of professionals could afford.
