I’m currently listening to Brian Wynn in the How to Think About Science series. He tells a story reminiscent of the physicists wagering the apocalypticness of the first atomic bomb explosion. Wikipedia provides a slightly drier, but more complete recounting of the decision to use water to put out the reactor fire at the Windscale Pile 1, Oct 11 1957.

I’m also reminded of this photo I took at the Los Alamos science history museum a few years ago.

There’s nothing more reassuring than interesting old people. I have no idea who Helmut Knewton was outside of this short interview with him (a photographer who did a lot of nudes apparently), but based solely on its content, I may have to rename another plant.

Q: When you arrived in Singapore you had five dollars to your name, which you immediately spent in a brothel.

A: My sound financial sense told me there was no difference between having five dollars and being completely broke.

Q: You never really talk about the Holocaust.

A: I have no animosity against the Germans. I will never forget or forgive but I find the Germans are the only ones who are seriously confonting their past. When I was offered the Great Federal Order of Merit, June said “You can’t possible accept it!” So I asked Billy Wilder and he said “You’ve got to take it!” I preferred listening to Billy.

And so on.

The following rant is inspired by the ending of the ’08 Photographic Marketing Association convention down in Vegas, and by all the photographic poking around I did with my recent trip photos. It will be a long time before the reviews are in on the new equipment from the PMA, but frankly it looks like all the same trends taken one year further. Digital SLRs keep getting better (looks like Canon may have taken back the edge from Nikon in the budget SLR category, but it’s an embarrassment of riches all round). Compact cameras get cranked out faster and cheaper, and with more stupid megapixels. Possible exceptions: Panasonic is having another go at the TZ line which could be great if they really have cut down the noise, a number of the ridiculous-ultra-zooms are starting to open up their wide ends (perhaps the curvature of the earth has begun to limit the value of extending the tele end of these zooms much further) and then there’s this bizarre but endearing beast, which if it actually gets out the door this year will be my first impractical purchase after my Carlos Slim heist comes off. You hear that Sigma? I’m waiting on you. With these and a few other exceptions, the increasingly obvious and boringly old problems with the modern compact camera market are not going away. So on to the rant.

Not to blame my tools, but my old camera sure is getting old. The on/off slider is so busted up that just touching the front of the camera while it’s turned on turns it off, and I think the lens is going soft, especially around the edges. Frankly, it’s incredible that it has survived as long as it has, through as much extreme abuse as it has. It’s a testament to the integrity of the Canon design squad that it works at all. It’s a 6 year old model, which is several lifetimes in digital camera years. The sad fact is that if I wanted to replace it today, I would an awfully hard time, despite 6 years of new designs. Compact cameras have become huge sellers in the intervening years, and are the victim of a proportionally grotesque marketing bubble. Certainly, there are lots of new featurey-things that have emerged. Some of them are downright useful (real, physical image stabilization is a giant step forward in hand-held photography, big old LCD screens are a lovely thing). Most of the new features are useless or near-useless, and have just sucked up R&D energy, physical weight and space in camera designs, and consumer’s money. Face recognition? Touch-screen LCD screens? Fake-out “digital image stabilization” modes?

More than feature-creep, the biggest downfall of the compact camera market has been the arms race between megapixels and noise. When I bought my 4 megapixel unit, the prevailing wisdom was that 4 megapixels was okay for most uses and 6 megapixels was plenty for anything but truly professional applications. Today the prevailing wisdom, if anyone bothers to think about it, is that 4 megapixels is okay for most uses and 6 megapixels is plenty for anything but truly professional applications. Yet compact cameras are routinely jamming 10 or more megapixels onto sensors that are physically no bigger than the one in my scratched up silver brick. Because, I guess, people shopping for cameras still respond to “megapixels” as some kind of feature, and it’s one that can be easily interpreted and communicated by a salesman or a sign. You know, more “resolution”, whatever that is. On compact cameras, sensors are tiny, typically thumb-nail size or smaller. Compare that to digital SLRs, which have 10 times or more the sensor area of a compact, and feature roughly the same range of megapixel counts.

The problem is that each pixel on the sensor actually is a physical little object, a sensor in its own right. Sometimes the pixel-sensors are called “photosites” to distinguish them from “sensors”, in the sense of the the entire sensor chip that sits behind the lens and collects the light to produce the image. If you cram twice as many photosites onto a sensor, that halves (quarters? my geometry is a bit weak) the size of each photosite. So given the same amount of time that the shutter opens (and shutter speeds aren’t going up, obviously), that means each photosite is going to get struck by half (or a quarter?) as many photons. From those fewer photons the pixel sensor has to decide what it was pointed at.

Camera engineers have gotten better at making each photosite a little more sensitive, pulling more info from each photon. But not much better. Not enough to keep up with the fewer and fewer photons each smaller and smaller photosite receives as megapixel counts totter higher and higher. Consequently, new cameras have to over-amplify the signal that the sensors record. Imagine turning up the volume on a fuzzy radio station to hear the music better. Sensors naturally produce some electronic noise, and are somewhat inaccurate in how they record the photons they receive. Amplifying the recorded signal means amplifying that noise and inaccuracy along with it. So modern compacts have for years been suffering from an increasing noise problem. That “grain” in your photos can be artsy-artistic sometimes, but usually it just mars detail. And that mottled red-green-blue “colour noise” is just irretrievably lame. Both noise flavours are the direct result of megapixel obsession. In terms of raw image quality at least, compact camera sensors are not getting better. They’re getting gradually worse.

That’s a strange scenario for a category of electronic gadgets, especially one as heavily invested in R&D as digital cameras. Marketing has screwed the cameraperson. There are a few exceptions to the rule. The Canon G7 and G9 come to mind, and of course the Fuji F30 and it’s descendants. But notice that even in those camera lines which have a reputation for low noise, each new model comes with a promise of better image processing to handle noise, and more megapixels, and a net wash or even loss in the noise department.

Resolution is not pixel count. No matter what the guy at the Best-Buy tells you. If your camera has a bijillion pixels, but your lens is too soft and smears the image out on them, or if they produce images which are so noisy they have to be smeared by “noise reduction” processing after the image has been recorded — and despite manufacturer’s claims, there is no noise reduction processing that doesn’t smear details at least some — or if your pixels were just kind of lame to begin with, then you don’t have a bijillion resolution. When I bought my S45, the 6 megapixel S50 replacement model had already been out for a while. Long enough that users had established that the lens on the S45 was sharper, and that the gain in real resolution from the new megapixels in the S50 was lost to the lens quality. So I bought the lower pixel-count camera, and never regretted it. Frankly, it’s been a great little camera.

So why didn’t I buy a G9? Well, my other beef with modern cameras is lens angle. For some reason, almost all compacts these days start at a 35mm “wide end”. I’m not satisfied with the 28mm on my S45, I can’t imagine regressing. Of course, they all have longer and longer zooms, but the wide angle is more useful than the tele for most shots (at least it is for me, and I suspect that’s true for the average photographer, if they thought about it). But again, 6x vs 3x zoom is an easier number to sell on then “24mm versus 37mm wide angle in 35mm equivalent viewing angle”. On the plus side, real wide angle actually does seem to be starting to show up in more compacts. On the other hand, my other other beef with modern compacts is raw image format. Back when they designed the S45, it took 4 seconds to nudge a raw image through the internal chips in the camera, a fair chunk of the memory card to store it, and a long time even for your computer to “develop” it once it came off the camera. Now camera chips are smarter/faster, memory cards are much bigger and computers are that much more capable, but raw has almost disappeared from even “pro-sumer” compacts. I’m looking at you, G9. wtf?

So why don’t I buy a digital SLR? Well, portability and the likelihood that I’ll actually have it when I want to take a photo, obviously. But I think I might just. Unlike compacts, SLRs have been getting better, faster, lighter, cheaper and more feature-crammed every single year. Makes you wonder what they could have done with the compact if they had really tried, doesn’t it?

I’ve added a few photos from my christmas-break trip to New Orleans and Texas (particularly Enchanted Rocks State Natural Area near Austin) and Joshua Tree Nat’l Park in California.

It was a fairly random trip, the sort of backpacking bounce-around I mostly associate with ‘overseas’ travel. Turns out that can be a lot of fun to do in the US but (surprise!) kind of expensive too. One thing all the main locations had in common was enormous potential for photography. So it’s a shame I didn’t take many photos, and fewer good ones. Photographs or no, I have lots of great material stored away in my head where it can mulch up and spontaneously generate romantically distorted “memories” of what were romantically distorted events even at the time. I’m looking forward to it.

What I learned from photoshopping my photos from New Orleans: that city has very few right angles.

Hiking out of Joshua Trees

Strange to see the Albertan oil boom showing up on Kottke.org as a novel topic. If this is news to people who haven’t been personally associated with the oil patch, then it’s far past time this news gets out. This should be and needs to be an major regional national and international issue.

Note that the Edward Burtynsky photographs seem to be of the smaller of the two major oil sands projects.

In the comments for this flickr photo of the Beer Despot’s lovely old neon sign is some historical commentary on how the joint used to be a drive-in, why it isn’t now, and why the sign is regulatorily doomed to entropy.

I keep thinking about this interview I read a while ago with Michael Crook. Crook is a urban tunneler. He explores the fringes of public space. The interview is mostly about sewers. He lives and explores in Toronto. The interview makes Toronto sing with life the way Michael Ondaatje’s books have. I doubt it’s coincidence that one of the named drainage systems he refers to is “the Skin of the Lion”.

Drains of Canada: An Interview with Michael Cook

I won’t try and pull any particular quotes out, the whole thing is amazing. And then there’s the photographs.

I haven’t spent much time on Michael’s website, Vanishing Point, but superficially it appears to be one of the best websites I’ve seen.