I’m thankful for…

If you don’t live in the States, yesterday was a holiday on which we overeat and count our blessings. I’m going to count a few in this post, concentrating on photography.

Affordable high-quality ink jet printers. We’ve had these for so long that it’s easy to lose sight of what a difference they make and how far we’ve come. In the chemical photography days, printing even what are today modest-sized prints – say, 16x20s – and doing it in only black and white, was a undertaking that required great skill, a sizable investment, and some remodeling of your dwelling. Add color and increase the size to what we routinely pull from even 24-inch printers, and you’ve just taken almost everybody out of the list of people who could do that at home. If your digital experience doesn’t go very far back, you won’t remember when people paid well over $100K for inkjet printers that produced evanescent output that looked terrible by today’s standards. My first printer with pretensions to photographic quality was a Dupont 4Cast dye sublimation printer that cost $40K and whose output faded in a matter of weeks. Or you could have a 4×5 negative made on a film recorder – don’t even think about buying one of those for your house – and print it in the darkroom.

The early home inkjet photo printers were not up to chemical photography standards, not by a long shot. In the late 90’s Fujifilm produced a gel-chemical light-exposed photo printer, the Pictrography, that could finally fit in a typical house and, at about $20K, be afforded by a photographer with deep pockets and produce stable output. I sprung for one. But it had serious limitations when it came to print size: a bit over 8 ½ by 11 inches.

By the turn of the millennium, Epson, hp, and Canon had started to get their inkjet act together. I bought my first wide-carriage printer, the Epson 9000, and soon ran into issues with permanence. I bought a set of third-party cartridges that were supposed to fix that, but the cure proved worse than the disease when my nozzles developed terminal arteriosclerosis. After a few years, Epson upgraded the printer in my house to 9500 standards, which fixed the evanescence, but wouldn’t work with lots of papers that I was using. They finally got it all together with the 7600/9600, and I still have a 9800 that is running like a train despite its advanced age.

In the last decade, inkjet printer technology has gotten to the point where bounding leaps of advancement have given way to slowly creeping improvements. We’ve gotten more inks: both more dilutions of the same colorants (“light light black”, anyone?) and more colors than the traditional CMYK. We needed the dilutions because the 1 picoliter drop size that was the goal in the early 90s has proven elusive. Resolution is holding steady, but bronzing is almost completely fixed. While the pro-level printers have gotten even more finicky about standing idle – they work great when used hard every day – the consumer ones have gotten more clog-resistant. While I would like to be able to print the equivalent of an 11×14 silver gelatin contact print, what we have now is good enough for almost everybody, and I don’t see anything earth-shaking on the horizon. Indeed, the coming of flat, large, low-pitch, big-gamut displays threatens to make paper prints quaint artifacts at some point in the future.

CMOS sensors. I’ve designed with CCD memories. I still remember burning my finger on a CCD clock driver IC that I was using to store the connection table for the first Rolm CBX. I used that technology because DRAMs weren’t quite there yet, but I always mistrusted it, knowing the number of things that had to go perfectly for it to work. It seemed to me that sensors were a better application for CCDs, since small errors would go unseen. But – tip of the hat to Eric Fossum – we have something that is not only much better now, it looks like it has runway ahead of it. The integration of ADCs on the sensor chip was a big step forward, and we are seeing the number of those circuits proliferating. Stacked sensor technology is allowing us to get the data off the sensor faster, and will soon allow the ADCs to move from the periphery of the sensor and make global shutter practical for large sensors. Switchable conversion gain and other advances have driven down read noise – especially high-ISO read noise – to levels unthinkable in the bad old CCD days. We can get decent photographs in near-darkness, and in, bright light, ones that, depending on the conditions, rival or exceed the best 4×5 film cameras could do in their heyday.

Sensors will continue to improve, marginally in QE and RN, greatly in resolution and readout speed. But we’re already at a point where there is hardly any photographic project that can’t be successfully undertaken with the sensors we have now.

Think about that last sentence. It’s not quite a “let’s close the Patent Office” equivalent, but it’s pretty amazing.

Foba camera stands and Arca Swiss Cubes. I use both every day, and I can’t say enough about how they allow me to be more precise and in the flow, what a tactile joy they are to use, and what a satisfying experience that usage brings. Neither are cheap, but they won’t become obsolete any time soon, either.