« It's Expensive Because It Was Cheap! | Main | The Most Popular Large Format »

Thursday, 15 October 2020


Feed You can follow this conversation by subscribing to the comment feed for this post.

If you count the scan backs of the late 1990s, digital capture resolution eclipsed 35mm film earlier than David's timeline. The first digital camera I used professionally was a Phase One Powerphase in 1997. It was a scan back attached to a Hasselblad 503cx, and had a 7000x7000 pixel resolution (36MP). Within a few years it transformed into a $35,000 doorstop, but the Hasselblad stayed in service for a good while longer.

I remind a very related conversation I had with a friend, when I was doing my PhD, in the early 90s. We both were technology enthusiasts, and loved photography, but in that field he was much more informed than me. I remember discussing about digital photography, and how long would it take to catch analogue. At the time we both were convinced it would still take quite a few years, but probably would have been surprised to know that in 2020, 50 megapixels would not be something strictly restricted to professional cameras.

Another thing we would not imagine at the time is that "more megapixels" would become a promotional gimmick very early in digital photography, and that at some point people were rather careless about the actual quality of pictures, being often blinded by the "my camera has much more megapixels" moniker.

In my personal cases: I got a Fujifilm F30 pocket camera in 2006, and a Pentax K10 in 2007. I have been considering to buy new cameras sometimes, but in the end I decided to stick to both depending on the conditions. And actually, for my personal needs (printing pictures only occasionally), their 6 and 10 megapixels are more than enough. For instance, I have two large prints of flowers and trees in the wall of our living room, made with the K10, and they look fantastic.

What I miss in my K10 is a better focusing system and better response to very low light. But megapixels? Nah, with 10 I am more than satisfied.

I hung on to film as others were going digital but I relented and bought a Fuji F602 pro zoom in 2003 or there abouts. I liked the image quality but it was slow for anything but posed shots so I moved to a Canon 300D and then went though 10D, 20D (I had that over 7 years) and lastly 5D.

When MFT came along I was an early adopter with the GF1 but I couldn't live with back screen shooting so changed to a G1 which I loved and it came very close to dislodging my Canon 5D.

When the Sony A7 came out I preordered one and when it came I was sold as was all my DSLR kit and I also gave away my Nikon SLR.

I still have some film cameras including my first which I got for my 10th birthday, a Kodak Instamatic 36, but I never use them and even my MFT kit hardly gets used now and whenever I get the time to take pictures it's with my A7.

I have my issues with home printing and I have a love hate relationship with my Epson R2880 but I do feel that when I do print I'm in control of quality.

I'm now getting to the stage where "Letter to George" may apply, which I think is a sort of peace and happiness place.

The Fax machine still lives on. I have a SHARP machine and it's working like the day that I bought it. Slow as it may well be but I have heard - at least the last time I did - that those encrypted data from machine to machine are highly secure.

How timely your post is when two things happened to me today: (1) a friend of mine sent me this article while we were talking about the news Fujifilm is involved in making COVID-19 drugs.


(2) I am going to pre-order my iPhone 12 Pro, thinking to myself these days lots of the improvements in photography are in the software but the traditional camera manufacturers pretend not seeing the paradigm shift.

Our old friend Michael Reichman, when the Canon 30D (or was it D30?) came out with 3MP, for the first time for a decent price, claimed that it rivaled the image quality of medium format, and blew 35mm out of the water (his words).

I've not looked at many comparisons, but at least it seems clear to me that 35mm was surpassed about 12MP.


"The Coming World of Photography"
- Popular Photography, 1944

Still my favorite photography crystal ball. Some prescience, mostly myopia. But good fun.

I remember thinking that it would take something like 20MP to equal 35mm film. Then when I started printing from digital camera files I discovered that a 6 or 10 MP file was superior to a 24MP scan, it was actually much closer to medium format than I was expecting. It was quite a revelation.

After making hundreds of comparison prints, I found that a 10.2 MP Nikon D80 was clearly superior to a 5x enlargement of medium format and close enough to a contact print from a 10" negative that the difference was meaningless. Quite a shock.

In the 1990s I saw digital take over the news business, then by 2000 studio portrait photographers realized medium format film did not serve them as well as a 2MP digital cameras. And when the Nikon D70 came out at $999 in 2004.....

Now it seems like the choice is digital (and a phone is a viable choice) or wet-plate.

Astrophotography started going digital in the 1970s. I heard about it in the 1980s—and that immediately told me that digital would replace film in most photography.

I knew, by then, how rapidly digital electronics advanced compared to other fields (it was still new, low-hanging fruit was being harvested, plus a lot of resources were being put into basic research because it returned short-term profits); CCD sensors in fact advanced more slowly than general digital electronics because they needed to stay large rather than shrinking immensely, and much of the basic research was directed to shrinking things.

I still wouldn't have guessed, then, that digital would replace film as fast as it actually did.

The other thing David Vestal overlooked, probably because the information wasn't widely known when he projected his catch-up year for digital imaging, was that it takes fewer directly captured pixels to make an image that looks good than it does scanned-from-film pixels. The direct digital pixels are much cleaner than the second-generation scanned pixels.

I tend to agree with the earlier-made proposition that the driving force to digital could stem - broadly - from Internet exhibition of one's pictures. I think that the days of people assuming that photography was all about holding a print have, if not vanished, been pretty much pushed into the background.

I suppose that if that's true, then the attractions of print-making become ever more restricted to those who still make a connection between photography and art. Not everyone who owned a film camera was also making his own prints. For my part, I sympathise totally with those who believe that the print is the final expression of the photographer's ability. That being so, having prints made by third parties negates much of the photographer's assumed talent: even a minimal experience of darkroom work should be enough to confirm that a good printer can rescue a lousy shot and transform it into something far better than the camera operator has managed to create on film - or digital, for that matter...

Digital may have begun as a bit of a fad, in non-pro circles, but it quickly grew into being the only game in town. For younger folks just freshly into photography, digital has become the standard because those young souls never did experience anything else.

Not to turn this into another pointless film v. digital thing, but the rapidity of both the acceptance of digital and the resultant stranding of analogue photography on an exotic island has forced film into a kind of minor Veblen situation. I'd suggest that's why some successful "art" photographers have slid sideways into old processes: the more arcane the better to attract those with money and artistic pretensions into buying things considered ever more rare and, inevitably, more valuable, if only from that limited perspective of supply and demand.

It's a shame that analogue has become so expensive to do; however, that is inevitable when film and paper people have a massively reduced market: you lose the benefits of economy through scale. I guess that most are fighting their last battle.

I remember reading something ages ago in the pages of "Popular Photography" (mid 1990s?), by their technical editor at the time, to the effect that for most photographers, digital will be "as good as film as makes no difference" once the sensors hit 6 megapixels. I think the rationale for that number was that 90% of 35mm photos were viewed on a 4x6 inch print, and 6mp was plenty for that, with room to spare for 8x10 enlargements.

I don't recall if he gave any timeframe. But sure enough, film SLR sales died off like smelt once the first two sub-$1000 6mp SLRs (Canon Digital Rebel/300D and Nikon D70) hit the market in 2003/2004. So I thought that turned out to be a pretty sound prediction.

I hung onto film too long for my casual shots, and now I have a box full of 6x7 slides and negs I took on a month long trip in India, and a few weeks in Europe. They are waiting to be scanned, hopefully before I am gone.
I still do Pt/Pd in the dark room, although now with digital negs.

I've a photo of David via the jacket of his book "The Craft of Photography". Copyright is a thing so I'll hold off scanning it and sending it unless you think Fair Use is in play here. Rather low quality but a photo of him nonetheless.

Predictions are always interesting when looked at with hindsight.

As a kid in elementary school in the early 1960s, our science text book said that man may one day land on the moon by the year 2000.

They probably would have done it too if not for Y2K.

Interesting that you mention the how digital supplanted film use in the same article you mention a fax machine. In the early '80s my mother worked for Federal Express (still the whole name spelled out back then) "selling" Zap-Mail systems to their midwestern offices. In the days before fax machines became common in every office, and long before email was a common concept, Zap-Mail was Federal Express's instant communication product. It was basically a fax machine the size of a large office copier. But can you imagine getting a document from one side of the country to the other in just a matter of minutes? This was futuristic tech at the time and Federal Express made a lot of money with it when "it absolutely, positively has to get there overnight" still wasn't fast enough.

Just a few scant years later Zap-Mail was dead, completely supplanted by the humble office fax machine that did exactly the same thing for (almost) zero cost and it didn't even require a trip to the local Federal Express office to send. Convenience and low per-piece cost, once the equipment was purchased, won out. A lot like what digital did with film.

Once the quality gets good enough for their purposes, the public will go for convenience every time. To wit: Bell Labs worked for decades to make a wired phone system whose reliability was unsurpassed; that system was doomed once truly portable phones became practical. Consumers ignored the frequent dropped calls because of the dramatically improved convenience.

David Vestal's remarkable prediction of digital-camera resolution may be considered a variation of Moore's Law. In 1965, Gordon Moore observed that integrated-circuit (IC) component density was roughly doubling every 12 months. In 1975, he adjusted his observation to 2x every 24 months. (Others have split the difference to 18 months, but it's unofficial.) Note that Moore's Law, like Vestal's forecast, is an astute observation, not a literal scientific law.

Image sensors are ICs, so they could be expected to track the general progress of Moore's Law. Vestal's extrapolation predicts much slower progress: 2.88x every 7 years. However, the IC components counted in Moore's Law aren't only transistors or pixel sites. They include other electronic components, such as resistors and capacitors. When counting all the IC components on an image sensor, perhaps they indeed track Moore's Law. (It's difficult to find the total component count for these proprietary devices.)

Image sensors differ from other ICs, though, because there's a trend toward making them bigger on purpose. In other types of chip designs, the goal is to make them as small as possible to reduce costs.

Thus, Moore was talking about affordable ICs -- not the biggest, most expensive ICs that could be built at any given moment. Applying the same principle to Vestal's extrapolation would exclude outliers such as the 7000x7000-pixel Phase One back that cost $35,000 in 1997, which someone mentioned in a previous comment. Vestal's extrapolation is more accurate when applied to the relatively affordable digital cameras available at any point in time. I think a "pixels per dollar per year" trend line would be more faithful to Vestal's intention.

One of Mr. Vestal's comments stands out for me:
In our reluctance to change we are mistaken.

Wikipedia has a nice photo of Vestal here, with a credit to Cindy Flowers:


It's interesting, all this information about 3, 6 and 12 megapixel cameras being fine for double page spreads and overtaking 35mm film for resolution etc. Interesting because here we still are in the ongoing megapixel race, many of us interested in 100mp Fujis, 48mp Nikons or whatever. I never had complaints about resolution from anyone for whom I supplied 12mp D700 pictures yet still I feel twitchy that my 24mp Z6 "isn't enough". I wish I could stop this twitching.

The best news from the last 25 or so years is that photography has been liberated from photographers.

Producing a photograph from chemical photography was the equivalent of needing a printing press to write a postcard. I loved it, and spent many long days in makeshift darkrooms, before I knew better. That happened when digital photography became possible for me. These days I can produce whatever I want, when I want, for free, without building a private production plant to do the very most basic thing.

My first D'Cam was a 1.3mp pocket Kodak, which produced the loveliest, smoothest color images. It also produced crap in low light, or at goosed ISOs, but even at the time there were other, better, and far more costly options, so it was good.

I've never regretted leaving film, but if I ever wanted to do chemical photography again, I'd go in up to my armpits, making my own camera, mixing up my own emulsions and processing chemicals, and go at it as a true craft.

But I won't. Poop on it, all. I'll leave that to photographers.

Digital lets me create images -- pictures, not photographs. I don't care about photographs, just pictures that I can use, such as pictures in color, a type of expression that otherwise would be forever closed to me, being one of the 8% of males who has non-standard color vision. (Though I can at times see reddish-green or greenish-red, colors that humans can't.)

It's a good imaging world we live in now, and I enjoy the freedom to just do it.

There's a photo of David Vestal on Wikimedia. Here's the link: https://commons.wikimedia.org/wiki/File:DavidVestal.jpg. BTW, when looking for things, Google is a great place to start. All I did was type "photo of david vestal" into Google's search bar.

Mike, you should watch the iPhone 12 release video. I think a full two thirds of it is dedicated to photography (and the results are really, really good).

People keep asking why camera companies keep ignoring computational photography. I find it hard to believe that they're that blind, but you surely must have to be selling a huge number of devices at luxury margins to pay back the massive, massive up-front investment in hardware tooling and software (and now, complimentary technologies like LIDAR).

Yesterday I went directly from someone arguing that there's no room inside a digital Leica M to do IBIS, to a demonstration of how the iPhone 12 Max does sensor-shift IBIS.

I wonder who among the companies making cameras has the capital, culture, structure, transitional skill, supply chain and fabrication connections, courage and patience to make that generational leap. Maybe Sony? Fuji?

Patrick Dodds, I'm still one of those guys that feels that 12 megapixel is about it! I wish pro-based cameras would "roll-off" about 25 mp, and then use the technology/sensor-space to improve color, maybe 24 bit.

My first "pro-ish" camera was the Nikon D90 and I could do a nice double page spread with it. That's one of the things that made me focus on M4/3rds at 16-20 mp ; didn't need the mp's as much as I needed the multi-aspect ratios (and "carry-ability").

Funny I read this today: I was talking to a another photographer today who was doing a bunch of A/B/C tests among his equipment, and he said the rig he loved the most was his Canon 5D Mk1, with old film lenses from the Olympus OM series, using an adapter (12 mp). He's told me before there's something about the output from the Mk1 that seems better than the Mk2; and he proceeded to show me some exceptional looking pics from his results.

In fear of running afoul of the "shush" rule concerning used camera bargains, that no one knows is a bargain yet; I saw KEH had excellent 5D Mk1's with battery and charger for less than 400 bucks!

I sent you an e-mail message with a photo of David Vestal, and others, from a 2008 workshop. At the time, David was using a Canon digital Rebel, with 10 megapixels (based on the date), which probably already met or exceeded the real resolution and noise levels of 35 mm film. True to form, he was using a single fixed-focal-length lens. My memory says that it was the most basic, 50 mm f1.8, but maybe it was the 35 mm f2.

The comments to this entry are closed.



Blog powered by Typepad
Member since 06/2007