This is the second part of an article about the Kodak DCS 460, an ancient six megapixel digital SLR from 1995. Part one is over here. The photographs that accompany this section were all taken with a DCS 460, in Bath, England, using a Sigma 15-30mm lens. The DCS 460 has a larger sensor than most digital SLRs, with a focal length multiplication of only 1.3x rather than the 1.5x of most modern Nikon bodies. At 15mm, the 15-30mm therefore emulates a 20mm lens. The camera is unusually sensitive to infrared, and without some way to prevent infrared radiation from reaching the sensor, the DCS 460's images tend to have wonky colour which is extremely hard if not impossible to correct.Back in 1995 the photographer was expected to combat infrared with the use of an infrared reflecting "hot mirror" filter on the lens. Unfortunately the Sigma 15-30mm can't use conventional filters, because it has a protruding front element, and so there is no way to capture images without running into the infrared problem.
I've tried attaching a small infrared filter to the rear of the lens, with some success, but this is fiddly and runs the risk of the filter falling off and getting caught in the DCS 460's shutter, with disastrous results. As a consequence of this I have transformed the images into black and white by retaining the red and part of the green colour channels, and disposing of the blue channel. Given that the blue channel tends to be noisier than the others, this is perhaps for the best. The end result is a set of images that emulate a weaker version of the infrared look, except for this shot, which was taken after the sun had gone down, and is still in colour:
There is a lot of information on the internet about modifying digital SLRs so that they become more sensitive to infrared, thus allowing digital photographers to recreate the "black skies / glowing foliage" look of infrared film. In contrast there is a very little on the internet about reducing a digital SLR's sensitivity to infrared. Digital Photography for What It's Worth has an extremely useful although dated page on the general topic of infrared digital photography here, which dates from a time when most people could not afford digital SLRs, and instead used high-end 2mp compact cameras. Diglloyd has an article about the infrared contamination issues that affect the Leica M8 digital rangefinder, particularly when shooting synthetic fabrics. Apart from that there is almost nothing. The DCS 460 is, incidentally, a poor choice as a proper infrared camera. It isn't sensitive enough to produce true infrared images with short exposures, and it can't record exposures of longer than half a second without introducing noise and a graduated purple cast that becomes more extreme towards the bottom of the image. Kodak made dedicated colour and monochrome infrared versions of the 1.5mp DCS 420, and possibly the DCS 460, although these variations were made in tiny quantities and were not sold to the general public.
Modern digital SLRs are not nearly as sensitive to infrared as the old digital models. The one notable exception is the Leica M8 digital rangefinder, which does not have space within the body for a suitably thick infrared filter. The M8 suffers from the same kind of infrared problem that affects the DCS 460, although it is much less pronounced in day-to-day shooting. The M8's sensor is made by Kodak, and so perhaps infrared sensitivity is a Kodak thing. The odd thing is that the Kodak DCS 14n, which did not have a Kodak sensor, also seemed to suffer from what looks suspiciously like infrared contamination, judging by this article by Rob Galbraith (I will add that he was reviewing an early version of the camera). Perhaps Kodak liked the look, dunno. No doubt many hundreds of thousands of images have been shot with Leica M8s and Kodak's full-frame DCS cameras without complaint from the photographers or clients, and I might be hypercritical or just wrong in this respect, but it seems to me that the men of Kodak had a thing for purple.A very few modern cameras are deliberately designed to be sensitive to infrared. Fuji's IS-1, IS Pro, and S3 UVIR (based respectively on the S9100 compact, and the S5 and S3 digital SLRs) were all aimed at the law enforcement and scientific markets, although as far as I can tell they were sold to the public as well. Sigma's SD10 and SD14 were not dedicated infrared cameras, but had a Kodak DCS-style removable infrared filter that could easily be unscrewed without special tools. The SD-series' combination of removable infrared filter, no anti-aliasing filter, and lack of a Bayer matrix means that they are, for black and white work, technically very similar to Kodak's monochrome DCS models, and much cheaper, although with a lower resolution.
It isn't possible, or at least it's not practical to use software to correct a shot that has been contaminated with infrared. If the infrared colour cast was consistent across the whole image it would be easy to apply a custom colour balance, but in the real world different objects reflect different amounts of infrared. The chlorophyll in grass and foliage reflects much more infrared than rocks and soil, or for that matter a wall painted the same shade of green as the grass. The situation is complicated by the fact that different types of foliage reflect different amounts of infrared - the healthier the plant, the more it glows - and that if I boost the saturation of the colours globally, objects that are captured with natural colour become hyper-saturated. The only solution is to apply local colour corrections based on the infrared reflectivity of the material in the photograph, which would probably not be any quicker than colourising a black and white image by hand.Here's an example of the kind of washed-out, wrong-looking colour that the DCS 460 produces without the use of an infrared filter:
~The DCS 460 was launched in 1995. From what I have read during the course of researching into the camera's historical mileu, the world was unimaginably different back then. In Britain there was a small but real chance that the next Prime Minister would be the young and handsome Michael Portillo of the Conservative Party. Doctor Who had been off the air for six years, and there was no chance that it would ever return. James Bond had been out of the cinema for just as long, although Goldeneye seemed pretty decent. In 1995 most people in Britain did not own computers or have access to the internet or own mobile phones. In fact, from what I remember, people who owned mobile phones were often mocked for owning mobile phones, because mobile phones were synonymous with yuppies, who were loathed.
Still, I'm digressing. A quick shufty on the internet reveals that precisely no famous fashion or editorial photographers used the DCS 460; or, if they did, they did not shout about it. No famous images were taken with a DCS 460. Do you remember the famous shot of the Afghan Girl that was was on the cover of National Geographic? That wasn't taken with a DCS 460. None of the images you remember were taken with a DCS 460. I have to assume that the big photographic magazines of the period for whom money was no object - National Geographic, Homes & Gardens, Playboy - would have evaluated the camera, assuming they had heard about it. Kodak must have sent out a press release at some point. The camera sold for £25,000, they must have wanted to boast about that, at least.I'm not joking about Playboy magazine by the way. Digital photography is ideal for pornography, although it is really more suited for internet pornography, which was a very small field in 1995. From what I remember, Playboy's centrefolds up until the late 2000s were shot with large format film cameras that could capture every strand of a model's pubic hair, although by the late 2000s this was less important because the models no longer had pubic hair. The irony is that Playboy's modern-day centrefolds are shot with expensive high resolution digital medium format backs, but are so heavily retouched and smoothed and airbrushed that the resulting files only contain a few megabytes worth of data.
National Geographic had been an early adopter of digital photo editing, famously and notoriously for its February 1982 cover, which had been edited so as to move some pyramids closer together. I have no idea which software the company used in 1982, or indeed if the photo really had been edited with a computer at all rather than with scissors and an airbrush; a lot of websites pass on this information, but they are vague on the details. Still, it seems to me that the magazine would have easily had the technology in 1995 to go digital. The venerable and all-knowing Eamon Hickey wrote a fascinating article about the magazine's use of digital photography for an aviation feature shot in 2002. At that time the photographer used a 5.8mp Nikon D1x and a 6mp Nikon D100. I can surmise that, if the magazine was willing to run a series of 5.8mp images in 2005, it would have been willing to use a 6mp image in 1995.
But resolution is not all. The DCS 460 has a host of problems that would have made it a poor choice for the magazine. The fact of its digital ISO speed being fixed at 80 might not have been the limitation it first seems, because National Geographic photographers were used to working with slow-speed Kodachrome 25 and 64, and Ektachrome 100 slide film. A photographer used to working with the limited latitude of slide film would have not been shocked by the DCS 460's similarly narrow exposure parameters. The DCS 460 is a noisy camera by modern standards, but then again film has grain, and even in 1995 it was relatively easy to filter out both noise and grain.
Presumably the biggest problem would have been the camera's odd colour reproduction. When I think of the National Geographic I think of photographs taken at dawn and dusk; I think of vivid sunsets, bright reds and greens and deep blacks. The DCS 460's shadow noise would have made this kind of photography problematic. I am sure that with the expertise and technique of a National Geographic photographer anything would have been possible, and if he had been handed a DCS 460 I would now be looking at a set of technically and artistically brilliant DCS 460 images of Mono Lake in California (for example). Perhaps the magazine did use the camera after all, but kept quiet about it.
NASA used the camera, as I have mentioned before. In fact they seem to have modified it to shoot at ISO 100 and 200, with an increased RAM buffer.In other respects the DCS 460 has a few practical problems. Back in 1995 the standard 35mm-format professional SLR was the Nikon F4, which could chunter through a roll of 35mm film at a rate of 5.7 frames per second, and could do so in wind and rain, in the midst of shellfire, for a very long time, assuming that the photographer had brought lots of batteries. The DCS 460 is not much bulkier than a Nikon F4, but it is not weatherproofed, and I would be wary of using it in sandy or dusty conditions. The camera half is held onto the digital half with a single screw, and there is no watertight seal between the digital back and the camera body. The PCMCIA card slot opens directly into the camera's insides, and does not have rubber seals, or indeed a catch. It is held shut with a spring. In desert conditions the camera would quickly fill up with sand and dust. Still, at least one man took a DCS 420 out into the wild, a chap called Marc Bryan-Brown who used one to photograph parts of Arctic Siberia, according to this mirror of Kodak's old website. The same website has a few brief mentions of the DCS 460, with some samples, and this page about a man called Tom Hopkins has possibly the only photograph on the entire internet of a man holding a DCS 460.
The lack of a review screen would not have been a major issue in 1995. None of the other digital SLRs on the market had a review screen either, and it was expected that digital photographers would review their images on a laptop (Hopkins, mentioned above, appears to have done just this). Given the speed of computers and memory cards in 1995 that would have been ponderously slow in practice, although faster than having to develop film. With my modern 2.4ghz Core Duo desktop with two gigabytes of memory it is a breeze to page through the DCS 460's files, but on a a 33mhz Powerbook 540c and 16mb it must have been a very slow process. In absolute terms it would surely have been much quicker than developing film, but the quality of the wait is different when a computer is involved. An hour spent waiting for film to be developed is an hour in which to have a cup of coffee, or fill out some forms. Five minutes spent watching a computer's progress bar fill up is five minutes of mental torture.
A single-user DCS 460 setup circa 1995 would have not left much change from $100,000, totting up the cost of two bodies, a stack of memory cards, replacement batteries, hot mirror filters, and a powerful laptop computer and all the equipment required to transmit images electronically. In terms of image quality the DCS 460 really has no advantage over slide film assuming that the photographer has access to a good-quality slide scanner for free, and is much less versatile than negative film, and so its selling points would have been purely practical; the camera could produce high resolution colour images ready for almost immediate publication, at a negligible cost-per-shot once the sundries had been paid for. The ability to edit white balance after the fact, without having to mess around with warming filters during the shoot, must have been a handy bonus as well. I don't know how long it would have taken to transmit one point two gigabytes worth of image files via telephone in 1995. A long time. It might have been quicker and more reliable to courier the memory cards, in which case why not courier film cartridges instead? A typical PCMCIA memory card of the period had a capacity of 160mb, at some no doubt astronomical cost. A 160mb card would hold just 26 of the DCS 460's 6mb raw files. In practice the DCS 460 was really aimed at photographers who would use the camera tethered via SCSI to a computer, storing images on the computer's hard drive.
Ultimately it seems to me that the DCS 460 was a technical triumph that, in practical terms, fell between too many different stools. With the two megapixel DCS 420 and NC2000e it was practical with 1995 technology to review and quickly transmit a series of killer shots of football matches so that they could be published in the late edition. Perhaps, with a speedy workflow and without taking too many spurious shots, a photographer could have used the DCS 460 for the same purpose, but the extra resolution would have been useless for newsprint. Conversely, for arty portrait shots the DCS 460 did not quite have the versatility and resolution of 35mm film. At the very least it did not have the right balance of practicality, image quality, and resolution.Ultimately the DCS 460's low fixed ISO, high noise, and lack a screen made it seem like a relic of a bygone age by the late 1990s. By the early 2000s Kodak's digital SLRs no longer had the edge over the competition in terms of resolution, and there was no secondary role for them to fulfill, and so Kodak gave up on the photojournalism business. The company briefly concentrated on the wedding, protrait, and product market, without sufficient success, and then left the professional SLR marketplace entirely, to concentrate instead on producing and developing sensors, which they will presumably continue to do until the company goes bankrupt, at which point I imagine Kodak will sell off everything it has via eBay = lots of cheap film.
This is the end of part II. In the third part I might talk about interfacing the DCS 460 with a contemporary Apple Macintosh laptop. Or I might go on holiday to Italy and take lots of photographs with Kodachrome slide film. I might just do that.
