Dante Stella Home | Photography | Current Feature | Galleries | Technical | Comment


35mm color vs. the tyranny of the masses.

In 1986, our father announced that our buying a minivan completed the descent of our family into the middle class.

What Thom Hogan Predicts for 2004

Interesting. Hogan predicts huge declines in film and film camera sales, one big last gasp in film SLR production (including a Nikon F6 that sounds like a new version of the crazy 1948 Kodak Ektra camera - complete with interchangeable film and digital backs). Will a big decline in 35mm film sales hit? I think the answer is yes. The question is when.

A. The High End

Consumers pretty much follow any trend that looks like it is convenient. The first Kodak involved sending the still-loaded camera back for developing (sound familiar?) 120 (and 31 other types of rollfilm) was invented for amateurs in the 1890s. Then came 127. Then 126. Then Polaroid. Then 110. Then Disc. Then 35mm. Then APS. Then digital.

Thirty-five millimeter film was a flash in the pan for amateurs. Although 35mm film had been used in various "miniature" cameras aimed at amateurs since the 1930s (Leica, Contax, Retina, Zeiss Ikon), it came in big around the 1960s with the Nikon F and other mainstream, practical SLRs. It existed in tandem with 110 and 126 through the 1970s. There was a huge 35mm boom in the late 1970s and early 80s. First compact rangefinders. Then the Canon AE-1 - the first "point and shoot." The amateur slide of 35mm SLRs has probably been on a downslide for 20 years, because people moved first to point and shoots and then to disposables. Amateur preferences seem to be of the cheapest, lowest quality equipment that produces a passable image. But it gets worse:

-- Even if amateurs hadn't dumbed down, when they discovered Ebay and sold off all of those AE-1s (and other cameras), the market flooded with ultra-reliable and cheap used cameras.

-- Older cameras are much more heavily-built than todays and last for decades of use. If people want to buy a manual camera, they can buy a better old one and lenses for less than the price of a new one.

-- Manufacturers ran out of new features to add. Autofocus and ultrasonic motors came about almost 20 years ago. Then matrix metering. Then eye-controlled focus. What next? By the time the F5 came out, there was now a feature set in one camera that no pro would ever max out.

There is little question that consumers exert a huge pull.

B. The Low End

That covers the low end of the spectrum. The other end of the market is pros. The needs of professionals in image quality are strikingly like that of rank amateurs, but not because the photographers are not talented. It has to do with output.

The simple answer for them is that pros need digital because the world has pushed them into using it. Editors want things faster. Customers want things faster. What is interesting about the pro market for photography (at least what I've been able to observe) is that the technology produces results as poor as the customer will allow. In the case of a newspaper, you don't need huge resolution, because the output is low resolution. The web needs less. Magazines need more. The D1x has been a standard for quite a while, and numerous cameras (even a couple of amateur ones) have surpassed it in performance. Too much resolution increases editing time, media costs, and transmission times. It does not add anything to the finished product.

In the portrait world, the end user of the resulting prints is the same one who thinks that FunSaver® pictures look good. In late 2001, when DSLRs were maxing at 3-4MP, all these pros had Ebay listings stating that they were "selling my MF gear because I'm going digital." I always puzzled about those statements because there was no camera on the market with the same capabilities as a Hasselblad (there is one now that arguably does, but it costs $12,000). I was approaching the problem incorrectly. It doesn't need to be as good. Pro equipment only really needs to produce a result which is good enough for the audience. For the consumer of many professional portraits, 4MP is enough, and B/W work is an afterthought. Some pros have even told me that their customers couldn't tell the difference between color prints from 35mm and 6x6.

C. Monkey in the Middle

Caught in between these two groups are the advanced amateurs. Their outlook is a little bit different. They are perfectionists, they are impressed by numbers, and they are the consumers of their own work. This is part of why the most obssessive users of online forums about photography are advanced amateurs. The reason there are not more advanced amateurs is that the rest of the people with similar mindset are elsewhere building high-end model railroad sets, computer systems, and killer collections of golf clubs.

The advanced amateur probably clings to 35mm film because he knows that color film can deliver 19 megapixels of useful data on a scanner. He hesitates about buying a DSLR because he is convinced that it is not "good enough" unless it has 14MP or more . But then, he complains, blue channel noise is... or chromatic aberration is... you get the picture. It never dawns on him that more capacity means less speed and much more expensive storage. They will be unsatisfied with digital - forever. From what I have seen in testing, 14MP is the point where digital is, by most accounts, "good enough." When cameras hit this point, development will stop, because no one will be able to see a further improvement. That 5MP gap may never be bridged. This will leave the advanced amateur with the rationalization that digital just isn't good enough... yet.

Where the noose tightens is processing. Keep in mind that 35mm processing is toxic, exacting, and generally unpleasant. That means that the C-41 hobbyist is the odd exception. This has always been the case with color film. I am sure that I will get an email from some older gentleman who tells me that he does tetenal C-41 all the time. For $2-4 dollars a roll for negs in a completely controlled machine environment without formadehyde in the air I breathe, I'll pass on home processing.

But what about commercial processing? In my town, there used to be at least 10 pro labs, 10 CPIs (later Fox, and not a terrible place to get your stuff done), some Ritzes and some miscellaneous low-to-medium end outlets such as drug stores. The CPIs and Ritzes of the world had trained people who knew what good pictures looked like. If you complained, they would do them again. This is what you pay for with color film over digital - having someone competent deal with the basic post-exposure processing.

When the pro market started to become economically impossible (it is not just in the Detroit area but everywhere), most of the labs closed. Minilabs were oversaturated, and all of the Foxes closed. Ritz was ok, but it was (and is) very expensive. So the supply and convenience factor dwindled a lot, and it had nothing to do with digital. While there are a lot of places to drop film, the number of one-hour outlets is down a lot.

That's a minor inconvenience. Once you get to negs (and most machines do a good job with negs), you can always scan them. The problem is mass output. As wages plummet, labs close, and Costco/Wal-Mart (have they merge yet?!) have become the dominant processing venues (ironic that so many Leica users admit to having prints from a $3,000 camera done by McJob minimum-wage laborers (oh wait, has Wal-Mart started making its workers 'independent contractors' yet?)

The remaining lab techs (some of whom have no organized training in color printing, much less aesthetics) are becoming less and less competent with proof prints. In a world where contact sheets just don't happen anymore with color, proof prints are all you have to judge a negative by, other than a loupe and a long scan. Today's techs (if you can even call them that) are by and large incompetent. Most of the minilabs are Fuji or Fuji Frontier; many techs believe that if it is not Fuji [supermarket] film, it can't be printed. Bulloney (if that were true, someone should call the Antitrust Division of the Department of Justice). They don't understand channels, and that is a huge issue in color printing. It's not like they are going to sit there and do trial-and-error. This is not a fluke; I had three different labs tell me that. A fourth had Kodak channels provided by Fuji, which is how I know that the "no-channels-for Kodak" schtick is b.s.

Poor-quality labs will destroy 35mm long before it becomes intrinsically unpopular or obsolete.

My parting thought is: if my 35mm use is exclusively color film, and I get crappy prints, underexposed scans, and scratched negatives back from every lab around, I might as well hang it up. I myself am waiting for a cheaper 24x36 sensor - it doesn't even have to be 14MP. The only two things it needs to do are (1) render my 105/2 DC Nikkor with a 105mm field of view (and subject distance / depth of field control), not a 150mm and (2) be able to get past the meddling of the idiots at the lab. The film cameras can go to what Hogan describes as the "third worlds," which is apparently everywhere but the G-7 nations. I don't expect that digital power source issues will be resolved with today's technology.

D. The Great (Black and) White Hope

If I can make one prediction, it will be that 120 film black and white film will survive 35mm as a medium for many advanced amateurs and artists. This is ironic, since 120 film is 30 years older than 35mm and is the sole survivor of full-paper-backed roll films. Consider the following:

-- Black and white imaging is directly dependent on film or sensor size for tonality. More silver grains (like pixels) are needed to make smooth gradients between tones. Digital requires no less. With only 256 output levels, it takes a lot more pixels to convince the eye of smoothness. It is highly unlikely that a sensor is going to come about with high enough resolution to duplicate the tones in a 6x7 or 6x9 black and white image.

-- Color sensors are unsuitable for black and white work (or at least undesirable) because they require four pixels to make one (two green, a blue and a red) and then upsample/interpolate. Coupled with the antialiasing filter built into most cameras, and black and white becomes a balancing act between fuzziness and oversharpening. Kodak made the DCS760M with the ideal sensor - one that measures luminance only - but it was (and is) incredibly expensive. That type of sensor is unlikely to arise again due to the required production volume.

-- Film has a much higher dynamic range than digital - 15 stops on TMY, or a 32768:1 ratio. That means that you capture everything in the scene and can go back and adjust up or down at will later. This means that you don't have to mess around with histograms and don't have to worry about shouldering highlights.

-- Film can generate images, even with tonal range expansion (through extended development) which are completely continuous tone when scanned or printed. Digital does not have this capability yet, and if you start with a short scale image in Photoshop and relevel it, you can often observe the gaps in the curve. This happens very frequently with the auto levels function.

-- Medium format cameras (dumped by pros for relatively low-res DSLRs) are selling at all time low prices, making the price to performance ratio incredible.

I pointed out in my other discussion of digital that 8x10 photography has managed to survive alongside 35mm. If you have seen 8x10 cameras, they have no operational similarity to rollfilm cameras. Their film comes in sheets, not rolls. They require you to focus on a groundglass and subtitute film. They require you to look upside down and backwards at the world. They have bellows factors. They require tripods. The only things that they have in common with 35mm is that you set the aperture, the shutter speed, and the focus. The processing is not done in tanks. The pictures are not always printed on enlargers - and sometimes they are displayed at 1:1 magnification. The process is slow, deliberative, inconvenient and expensive You don't even have to argue that roll film is easier. But the results of 8x10 are indisputably better than anything else.

The few things that 8x10 has in common with 35mm, both have in common with digital. It is my surmise is that just as 8x10 managed to survive the rollfilm revolution, 120 b/w rollfilm (convenient enough to make it easy for people to do easily at home - with moderately sized enlargers) will continue to plow along in the same world where color digital cameras do.

Even if Big Yellow and Big Green stopped making film in the near future, you could probably get it for quite a while. Are Kodak and Fuji likely to quit soon? Probably not with Kodak sinking billions into a new film plant and Fuji making upgrades to its manufacturing base. Here's what you can probably expect in terms of steps:

-- continued film sales for at least several years accompanying by worse and worse processing

-- a plateau in sensor development for DSLRs at about 10-12MP.

-- a steadily declining selection of film sizes, probably hitting color 35mm first, then black and white 35mm first, then eventually larger b/w sizes.

-- reimportation of the massive amount of film that will still be sold to places where a digital camera costs as much as an average family income.

-- discontinuance of film by major brands when China finally succeeds in flooding the third world with passable digital cameras made at a labor cost of $1 each.

-- resumed production in Eastern (er... Central) Europe as a boutique item. By this time, most of the conventional photographers who are adults now will be retired.

There are a few things that point to film sticking around for as much as another decade.

First, as noted above, both Kodak and Fuji have both recently made heavy investments in film manufacturing in the United States.

Second, manufacturers know that film is a profitable (Kodak has even explicitly said as much). It is a mature product. Does it make any difference that Kodak is not doing any more R&D? No. Its black and white films are unsurpassed (and unsurpassable - since no one will ever put the same kind of money into development again). Its color films have reached a plateau. Ditto for Fuji's color materials. If you are not a saturation fiend, the struggle to improve color film was largely behind us.

Third, Wall Street believes that film is profitable too. This is what led to the flap over Kodak's new business model. If Kodak doesn't do a good job with film, someone will acquire the business (possibly by force) and operate it.

Finally, whatever the sales figures for digital, there is an installed base of millions (if not hundreds of millions of film cameras). Even if all demand for film disappeared tomorrow in the United States, Canada, Europe and Japan, which make up the bulk of the digital consumers, it would account for only about 1/6 of the world's population. Everyone else uses film and probably will until digital prices come down -- to the bargain basement. People fail to realize that in places like India and China a digital camera can cost almost as much money as a family makes in a year.

E. The Distance

Is this the malingering of a film user? If you think you see an anti-digital message here, please send me an email pointing out exactly where it is. And please don't react with a tirade to the effect that "film is dead," "digital is equal," or "everyone is going digital." Those still sound more like camera-store sales talk to me than reality. I don't dispute that digital color is fast becoming a viable reality (and sole option) for serious amateurs and that it will impact 35mm film. In today's processing environment, that may even be a good thing, even if at some theoretical level digital will have a hard time meeting 35mm on sheer performance numbers. I think it's a far different story for medium-format black-and-white, one that will take longer to tell.