Wednesday, 15 February 2017

The Building Inspector

He showed me an illustration of a teacup and asked me what I thought it was. "It's a penis-holder", I told him. "Or a penis-cover. Men put it on their penises to keep them warm. Penis, singular."

It was a teacup. I wasn't in the mood for this kind of rubbish. Over the years the council had sent several inspectors to our premises to make sure we were complying with their rules and I was starting to lose my rag. "It's a teacup."

The inspector seemed surprisingly chipper. He was a very intense man. "Well that's the point. For you - for you - this is a penis-holder; for most people it's a teacup, but how do we know that? What makes it a teacup?" to which I replied that it was obviously a teacup. The design was made to hold hot liquid such as tea, but it was too fancy for e.g. engine oil or hydraulic fluid, ergo if we’re being generous it was a hot beverage holder, and as this was Britain it was a teacup. It couldn't be a penis-holder because it would fall off, unless it was held on with straps, which were not apparent in the illustration thus suggesting that they didn't exist and therefore it was a teacup.

After some minutes of this banter he cut to the chase. The council had been going through their records. They had been asked by the government to audit their property portfolio, but there was a problem. The architectural firm that designed our office building had long since gone bust, and the council’s records had been mislaid. Our building no longer had meaning, so the council had appointed a building inspector to determine its nature. And here he was.

"Have you ever heard of hermeneutics?", the inspector asked me. "Hermeneutics is the scientific art", and he stressed those words, "the scientific art of determining the meaning of a text through careful consideration of the historical context. Not just texts. All things. When I say text I mean everything that has meaning, or rather latent meaning, which means everything. The clothes on your back have meaning beyond their physical form. This chair." There was a pause. "The arrangement of the chairs, even." He leaned forward. "We are all wafts of dust in the cosmic wind."

He explained that he had been asked to perform a hermeneutical analysis of our building in order to determine its true function. "For you and your employees," he said, "it is an office building, with office workers and photocopiers. But you are not professionals", to which I replied that in fact I was a professional, I had a degree. He apologised. "But you are, at least on a philosophical level, not trained in the art of contextual analysis". Given that our company paid the rent I felt that we had a jolly good right to define the context of our working environment, but nonetheless he produced documentation from the council which, to my dismay, showed that they would accept his recommendations in the event no suitable counter-argument was presented at a board meeting scheduled for next month.

And so the building inspector embarked on a series of tests. He measured each of the rooms. He showed me his laptop, on which he had created a three-dimensional model of the building complete with little models of all the employees. "There's even a little model of me", he said, pointing to a stick figure standing at the apex of an imaginary pyramid one hundred feet above the roof of the building. "The pyramid represents my field of view", he explained. "I am simultaneously an observer of, and a participant in... and a commentator..." he said, as he fiddled with the laptop's trackpad, "simultaneously a chronicler of the ongoing construction of the ongoing construction of the dominant critical framework." The laptop had illuminated buttons that could make smiley faces appear on the screen. Every fibre of my being urged me to press one of the buttons, but I suspected he would accuse me of assault, or obstruction, so I refrained.

Mid-way through the month he had refined his model. The building was now a series of separate rooms linked together with double-ended arrows. To complicate things one of the receptionists had got married and moved away, so we had moved the reception desk around, which upset the building inspector. He forbade us from moving any more furniture and so for two weeks I took great pleasure in shifting the bins slightly before I left for the night just to mess with his mind. On my own initiative I produced a small file of press clippings which demonstrated that the building was a former technical college that had been built in 1965 but then sold off when the college merged with a local university in 1983, at which point our firm had moved in. He explained that he had not been hired to analyse press clippings; that was not his field. It appeared that the council hadn't bothered trying to contact the architects, even though two of the partners were still alive, because their intentions were irrelevant. To all further objections the inspector accused me of being originalist - "one of them" as he put it - which was apparently a bad thing.

One month later we met in the council boardroom. The inspector presented his findings, but I had done some preparation of my own. It seemed that the inspector was not a popular figure. He had attempted to apply hermeneutical analysis to the council staff, arguing that although the chairman thought of himself as an experienced executive, and had indeed worked for twenty years as a council chairman, he was instead cut out to be an Alpine farmer, driving cattle to and from mountain tops. The chairman had taken this as a jab at his weight and they had not seen eye-to-eye since then.

Furthermore the inspector's qualifications were suspect. He had a degree, but it was in podiatry. He had spent several years treating foot diseases, but after staring at hundreds of infected toenails he had had an epiphany which led to his abrupt career change. Perhaps some kind of foot fungus had rotted his brain. Nonetheless the council were bound by law to at least conduct the inspection, albeit that they were not compelled to abide by his findings, and therein lay a glimmer of hope.

During the presentation the inspector revealed that he had had yet another insight. Buildings were not homogeneous, he explained, they were instead vessels that housed a spectrum of meaning that pivoted through time. Thus our building was not a creation of bricks and mortar, it was a contextual gradient. He declared that the lower floors were in future to be room for prayer, and that the upper floors were to gradually merge from office space into a public swimming area, which would over the course of the next twelve years become a smelting furnace. The lower floors would, over a longer period of time, transition from prayer room to a place where expectant mothers could gather and swap tales of the moon. He suggested that the pool could be lined with carbon fibre-impregnated ceramic so that it would not melt. Seventy-five years hence the library would become a historical record of a library, with imaginary books and a staff of actors paid to act out the roles of library assistants.

This was of course not on. It sounded very dangerous, besides which there was no longer any space for our office. Luckily I had spent several years working for one of those internet companies that had been fashionable a decade or so ago, so I knew a thing or two about convincing rubbish. I presented my defence.

"I put it to the board", I said, "that although your inspector's analysis has merit, it is excessively skeuomorphic and fails to account for the impact of the internet of things", which was true on both counts. "The examples he gives, of the swimming pool and the grain silo - the suspension bridge as well - they are backwards-thinking legacy artefacts. Within a few years the building will be thoroughly embedded in the cloud, at which point its physical form will be meaningless. It will instead be a hot-swappable processing node in a giant hive mind." I was warming to this.

"In my view the traditional roles envisaged by your inspector are architecturally heteronormative. It is not the place of la humanidad to metaphorically assign genders to the structures with which we cohabit this planet. To unpack frankly, it is up to the structures themselves to define their meaning". As I said all this I pulled up some slides that had diagrams of buildings linked together ephemerally with wi-fi rays. They were also linked to passing cars, an airliner, and some little stick figures holding mobile phones. The headline read DEVICE MASHING.

"This slide illustrates what I'm talking about. In the very near future, entities - and that includes buildings, and people, but I'm talking about buildings - will occupy a metasynchronous space where cloud-enabled devices can define their own context. Cloudizens, I call them. Citizens of the Cloud, and in my view citizenship of this new realm is not the sole preserve of the white man, or of men, or even people". I glanced at the building inspector, who was at this point fiddling with a pen. He had a grim expression on his face; I imagined terrible visions of rotting toes dancing in his mind.

"However our office has not yet reached critical mass. In order to achieve singularity we need to scrum the power of big data into a cloudy singularity, and we won't be able to do that if half of the building is a bloody swimming pool. To sum up, although the inspector proposes a transformational framework that ostensibly reflects the hermeneutical exactitude of the building in question, he is mistakenly imposing an objectivist framework rather than harnessing the latent emergent power of the building's consciousness. He is, not to put too fine a point on it, not sufficiently woke".

The board convened to consider their findings, and that was the last I heard of them, and the building inspector.

Sunday, 5 February 2017

Canon EF 50mm f/1.4: "Good-sized, heavily-ribbed, and rubberized"

What is there to say about the Canon EF 50mm f/1.4? It's the EOS system's fast standard lens, and has been since 1993, when it was launched. It's almost a quarter of a century old and has been reviewed by all of the major photography websites. Back in 1993 Popular Photography, a print magazine, concluded that it was naff wide open but splendid when stopped down.

They used that very word, splendid. They used some other words as well, see for yourself:

Nothing has changed since then. Ducks still cannot count. The 50mm f/1.4 remains on sale and is just as it ever was; lots of photographers who use Canon gear have a 50mm f/1.4 lurking in a cupboard for a rainy day. They never use it, but it's there.

When digital SLRs became affordable the 50mm took on a second wind as a portrait lens, because on an APS-C camera it becomes an 80mm. For the images in this post I stuck it onto an EOS 50e, an old film SLR which has eye-controlled focus.

On a physical level the 50mm f/1.4 is "second generation EOS". It uses the same design language as the 28mm f/1.8, 85mm f/1.8, and 100mm f/2, which were all released within a few years of each other.

The star begins with black tapes, and early closing.

What was eye-controlled focus? It was a Canon thing from the 1990s. A handful of film SLRs used it, but it was abandoned in the digital era and has never been brought back.

The idea is that you pick an autofocus point in the viewfinder by looking at it. The viewfinder has an infrared sensor that tracks the movement of your cornea. It sounds slightly worrying, but if you don't mind having the camera shine infrared beams at your eyes the system works surprisingly well, albeit that the EOS 50e only has three autofocus points. When I glanced left, the left-hand focus point lit up; when I glanced right, etc.

The problem is that sometimes you want to look at the whole frame, not just the little part that you are focusing on. In the following image the focus is wrong because I stopped looking at the correct spot:

Canon's 50mm EOS lens range has always had a schizophrenic quality. The low-end 50mm has always been Canon's cheapest prime; the high-end 50mm has been among the most expensive, with the 50mm f/1.4 floating in the huge gulf that separates them.

Back in 1987 the 50mm f/1.8 was part of the first batch of EOS lenses. It was bundled with the very first EOS cameras as a kit but was also available separately for $70-80 or so. Two years later Canon launched the 50mm f/1.0, which remains to this day the fastest autofocus lens ever made. At a price just shy of $3,000 it was one of the most expensive EOS lenses, beaten only by some of the exotic telephotos.

In 1990 Canon replaced the 50mm f/1.8 with the cheaper, plastic-bodied 50mm f/1.8 MkII, which is still on sale today.

At some point in the 1990s or early 2000s the 50mm f/1.0 was quietly discontinued. From that point onwards Canon didn't have a high-end 50mm until 2007, when the 50mm f/1.2 came out. In the mid-2010s Canon also launched the pancake-sized, not-actually-50mm 40mm f/2.8 STM, and the normal-sized, actually-really-50mm 50mm f/1.8 STM, which both use a silent stepless focusing system optimised for video, but there has been no direct replacement for the 50mm f/1.4 and so it soldiers on much as it was in 1993.

Canon also sells a 50mm t/1.3 cinema lens, which is apparently similar to the 50mm f/1.2 but in a tougher body, with external teeth for manual focusing gear. Of note Yongnuo, a Chinese company more famous for its cheap, decent-quality flash units, makes a clone of the 50mm f/1.8, which apparently isn't quite as good but is slightly cheaper.

Design-wise the 50mm f/1.4 has aged well. Technologically however it's behind the curve. It uses a variation of Canon's ultrasonic focus system called micro-USM, which is apparently slower and more fragile than the ring-type USM used in Canon's more expensive lenses. Compared to my 70-200mm f/2.8 the 50mm f/1.4's focusing "feel" is less smooth, but not offensively so. The manual focus ring feels like plastic moving on plastic. Unusually for a Micro-USM lens you don't have to flick the AF/MF switch to enable manual focus, you can just twist the manual focus ring.

The 50mm f/1.4 poses something of a conundrum. The f/1.8 is very cheap, so masses of Canon fans have one. They probably bought it as their second lens. The f/1.4's image quality is better, and it's slightly faster, but it's not overwhelmingly superior and the speed difference is negligible.

Meanwhile the 50mm f/1.2 is very expensive, and the image quality is optimised for f/1.2, which is to say that it's no better than the f/1.4 stopped down. The 85mm f/1.2 is much better, so why not buy one of them instead? They're roughly the same price. The 40mm f/2.8 aims at a slightly different market. The end result is that if you want a good fast 50mm for the EOS system and you don't want to pay a fortune the 50mm f/1.4 is essentially the only option and yet it's disappointing, because the image quality is just good, but not great.

50mm is a classic focal length. Manufacturers use it to showcase their expertise, because it's relatively simple to make a fast 50mm lens, and people like fast lenses. Over the last few years there has been a wave of 50mm lenses from other manufacturers, notably Carl Zeiss and Sigma. They're generally better-made and optically superior to the 50mm f/1.4, but much more expensive.

For all its slightly-underwhelming reputation the 50mm f/1.4 is a good performer, but as Popular Photography pointed out a quarter of a century ago accurate focusing at f/1.4 is difficult. Here are a trio of shots - they're 100% crops from the centre of a 21mp image, taken at f/1.4, f/1.8, and f/8. I focused on the 11 in the clock face:

It was taken near the close focus distance, at which range the depth of field was minuscule. Wide open there's quite obvious vignetting, which seems to affect autoexposure slightly; there's also purple fringing on highlights, and furthermore if you look at the background it's slightly green, which is a consequence of bokeh fringing. I conclude that sharpness in the middle at f/1.4 is fine if the focusing is correct. At f/8 it is, like every other 50mm lens, terrific.

And that's the 50mm f/1.4. You probably already have one, or you don't!