Monday, November 18, 2013

The Smartphone as a Hardware Platform for Photography

People use their smartphones a lot for photography.  Ooh, sorry, I probably should have asked you to sit down before I dropped that bomb on you.  The smartphone has been a software platform for photography for a long time, thanks to semi-decent hardware, and a lot of creative developers.  There's the biggies for sharing, such as Instagram, or editing, such as Snapseed, and these apps have helped me take some camera phone pictures that I actually like.

My favorite apps, however, have always been the somewhat more quirky, esoteric apps such as Rando or ThrowBack.  And of course there's apps for shooting time-lapse, apps which turn your camera into a light meter, apps that help you record metadata for your film photos, and even apps that aren't strictly photography applications, but which have been enthusiastically adopted by photographers, such as apps that help you figure out where the sun will be.

So that's cool.  But even more interesting, there seems to be a growing movement to use the smartphone as a hardware platform for photography.  It started with add-on lenses and adaptors, which range from, "hmm, interesting" to "please, don't insult me", and has gone seriously over the top with the Sony QX10 and QX100, which by all appearances are entire add-on cameras.  Conceptually, they remind me of the Ricoh GXR, except they might actually be successful because they leverage something you already own.

That's the key to the smartphone hardware platform movement.  I can't be bothered to buy a film scanner, but I'll rave about a demonstrably inferior product because at its heart is a phone I already own.  These hardware add-ons come from everywhere, but in particular crowdfunding sites such as Kickstarter and Indiegogo seem to spawn smartphone photography add-ons at a dangerous rate.  By which I mean, dangerous to my wallet.  In addition to the Lomography scanner, I have funded and am eagerly awaiting the Snapzoom, Poppy, and Nova.  I even funded the Cookoo watch, mostly because I wanted a remote trigger for the shutter on my phone (which is lucky, I guess, since that's about the only thing the watch does that actually works).  And it was only through an act of sheer will that I avoided pulling the trigger on a $450 Enfojer.  By the way, the Enfojer video has the best opening line of any Kickstarter/Indiegogo pitch video ever: "Hello!  And enough about me…"

There's something undeniably appealing about being able to essentially buy peripherals for your camera.  Not a lens, or a flash, or a grip, such as you might buy for a conventional camera, and which simply helps you do more of what the camera already does, and which may not even work with the next camera you get.  Rather, many smartphone camera peripherals are radical enough that they turn the camera into something that it's otherwise not.  And for the most part, they're generalized enough that they'll work with your next smartphone.  And that's key.  The ultimately disposable nature of the smartphone actually increases the value of the peripherals, since you know you'll rarely have to wait longer than 24 months before you've revitalized the camera at their core.

I feel that somewhere in all this is a profound insight waiting to be made, but I can't quite articulate it.  Since I seem unable to end this post in a blaze of pithy wisdom, I'll just say that I think it's interesting that, for whatever reason, most of these devices originate on crowdfunding sites.  It's almost like Kickstarter and Indiegogo have become the app store of smartphone hardware add-ons.

Monday, September 16, 2013

The World's Greatest Film Scanner

I shoot film. Not a lot, but it happens. When I shoot film it's usually with a wonky camera that does something cool that my digital camera can't do. Such as the awesome Lomography Spinner 360. Who could say no to a camera with a rip cord? So anyway, I shoot film. Then I take it to one of the three places in the world that will process it (no prints, and please don't cut the negative). And then it sits in a drawer. That's pretty much it.

What I need, of course is a film scanner. I've never bought one because, damn. Another piece of crap that does one extremely limited thing that I hardly use, and which takes up space in my life. Of course I could have the photo lab scan the film, but they can't seem to manage unusual negatives such as those from the Spinner 360. So, yeah, into the drawer.

Until the Lomography Smartphone Film Scanner came along.

I originally funded the scanner on Kickstarter, but it can now be purchased directly from Lomography's site. The LSPFS is not really a scanner at all. Instead, it's a device that holds your smartphone, and some film, and points the one at the other. It backlights the film, and it is adjustable to hold most any smartphone. Once you've got your phone clamped in place and your film inserted, you simply snap a picture of the film with the phone's camera. Of course you can do that with any camera software on your phone, but the LomoScanner software provides built-in functionality to scan and automatically assemble motion pictures from your Lomokino, or panoramas from your Spinner 360. And of course it's got one-button sharing to the social photography network de jour.

Did I almost forget to mention that the scan quality is crap? You're no fool, you probably figured that out already, though it may be even worse than you imagine. Don't think that you'll be getting scans at the camera's full resolution, because the film frame occupies only a small portion of the camera's field of view. For example, my iPhone 4S has an eight megapixel camera, but the scans I get with the LomoScanner are closer to two megapixels. That's hardly the only problem, of course, since the dynamic range of your phone's camera is tiny compared to that of film, and so on. Even so, this is still the world's greatest film scanner, for the following reasons: unlike any other film scanner, it is small enough, cheap enough, and easy enough that I actually bought it and use it. I don't have to buy a dedicated piece of delicate electronics that will eventually become obsolete, or break. The LSPFS doesn't seem likely to break, its electronics consist of a single lightbulb. An LED probably, but I haven't taken it apart to see.

And while relying on a smartphone camera to scan your film might guarantee mediocre quality, the quality will improve every time you buy a new phone. (And hey, if you use your phone's built-in HDR mode, you can eke a bit more detail out of the scans.) The film is coming out of the drawer. And best of all, when I'm done scanning, the LomoScanner is so cute and tiny that it can go into the drawer.

Tuesday, August 27, 2013

Doin' it Wrong, Doin' it Right

Camera manufactures just keep on making cameras.  I don't know if you've noticed.  The latest and greatest will come out and then a few months later, boom, there's a new sherif in town.  Good as today's cameras may be, we expect better versions tomorrow, in the same way that we expect the sun to rise.  And yet, as with the dawn, there are occasional cloudy days.

The Sigma DP1 Merrill

I am one of the three people who bought the original Sigma DP1.  I loved that camera right up until the day it spontaneously stopped working, a not uncommon experience judging by the internet chatter.  The DP1 was a flawed masterpiece.  It was a quirky, limited camera that was slower than dirt in every respect.  It took 5 seconds to boot.  The auto focus would carefully explore all of its options at a leisurely pace.  You could get in a quick nap while the camera buffer cleared.  The lens was an f/4.

But the camera produce these gorgeous, lush, perfect photos.  I wrung more beauty out of that camera than any other camera I've ever used.  Of course that was due in no small part to the Foveon X3 sensor, but that ridiculous f/4 lens was a thing of beauty as well.

None of the above was ever a secret.  Every review written said essentially what I've said here.  OK, here we go: you work at Sigma.  Your job is to design and produced the next version of the Sigma DP1.  What are you going to do to make the camera "New!" and "Improved!"?  This isn't a trick question.  The answer is as obvious as it seems.  When the latest and greatest DP1 finally did emerge, what did we get?  Did Sigma deliver the obvious?

When you read reviews of the DP1 Merrill, you notice a funny thing: they sound exactly like reviews of the original DP1.  "Slow, limited camera, beautiful photos."  But here's the thing: these reviewers are comparing the DP1 Merrill to other cameras.  They're not comparing it to the original DP1 which is, of course, what I did.  As a member of the cult of DP1 it couldn't be helped.  Even before I picked up the DP1 Merrill, I was excited by the fact that the lens was now an f/2.8.  And when I first started using the camera, I was very pleased.  The performance could hardly be called "class leading" or even "middle of the pack" but it had certainly moved up from egregious to acceptable.  Sigma, it seemed, had made a reasonably successful effort to address the shortcomings of the original camera.  I was ready to buy.

And then I looked at the photos.  And they looked… OK.  No one else seems to have reached this conclusion, so maybe my expectations were just too high.  Read the reviews at Imaging Resource or Luminous Landscape, these guys will tell you how excellent the image quality is. And they're not idiots.  It's hard to argue with the comparison shots that Luminous Landscape posted: the Sigma does indeed seem to be bitch slapping the competition.  But as I said before, I wasn't comparing the Merrill to a Sony NEX-7 or a Fuji X-100.  I was comparing it to the original DP1.  It's possible that an objective comparison of the Merrill to the original DP1 would reveal that the Merrill's pictures best the original's in some fashion as well.  But I wasn't making an objective comparison.  I was shooting pictures, looking at them, and reacting, and the photos from the Merrill just didn't do it for me.  It's not something that I can explain with color swatches or a resolution chart.  Here's the best I can do: more than any other camera I've used, the DP1 was transparent.  By which I mean, when I looked at its photos, all I saw were the pictures, I didn't see the technology that produced them.  The Merrill is not transparent.  When I look at its photos, I can see the fingerprint of the camera that produced them.

As such, the Merrill just isn't worth it.  I gladly suffered the idiosyncrasies of the DP1 because it produced photos like nothing else.  But the Merrill has lost that edge.  It may still stack up in an image comparison, but for me the magic is gone, and there are a zillion cameras out there that are better in every other respect.  If Sigma had simply made a DP1 that was fast, I would have bought it. As it is, I've moved on.

The Fuji X-20

In March of 2012 I took the original Fuji X-10 on a weekend trip to New York.  I loved that little camera, it was such a joy to use, with its retro style and its manual controls.  My favorite aspect was the zoom lens.  It was a manual zoom, which is essentially unheard of in a point and shoot.  And there were nice little touches, like the zoom ring doubling as the power switch, and the optical viewfinder zooming in sync with the lens.  I really enjoyed shooting with it, and I felt like I got some nice shots.

There were only two things that bothered me about the camera.  First, I wished that the image quality was a bit better.  It was fine, but given the cost, and the otherwise excellent nature of the camera, I was expecting better than fine.  In particular, noise performance, even at low ISO, could really have been better.  Second, I loved the optical viewfinder conceptually, but I found that I didn't actually use it much, for two reasons: I had no idea where the camera was focusing, and I had no idea what exposure settings the camera was picking.  I needed to look at the screen to see that info, and so I always looked at the screen, and the cool little viewfinder went to waste.

Introducing the Fuji X-20.  It is identical to the X-10 except for two things.  Do you want to guess what they are?  Image quality, in particular noise, has been marginally improved. And focus indicators and settings readouts have been added to the optical viewfinder.  Awesome.  The X-20 is, in essence, a perfect version of the X-10, and it seems to me that should be the goal of any successor.

Look, I'm not breaking any ground here, but I'll spell it out just the same, if only for Sigma's benefit: if you want to improve something, you need to correctly identify the shortcomings of the original, and then fix them, and only them.  Simple.