Monday, February 27, 2012

Forty-One Megapixels... In a Phone

Nokia has always been something of an engineering firm first and a software firm second. Actually, they're more like a software company third of fourth. Which is of course the reason why they had their ass handed to them when Apple released the iPhone with last-generation hardware powering next-generation software.


Still, you can't argue that Nokia hasn't produced some badass handsets in its time. The Nokia N-Series phones were, even during the iPhone's early days, the undisputed masters of cool tech.

Now with their Windows Phone partnership well under way, Nokia appears to have found some of its mojo and has gone and done something absolute wacky: they have released a phone, a phone, with a 41-megapixel sensor in it.

The Nokia 808 PureView is packing both the largest and highest resolution sensor ever in a phone. Nokia is treating this as a pretty big product launch and is making sure that everyone understands their logic, which isn't bad logic. Basically, combined with great optics, those forty-one million pixels are sampled at various rates to achieve different resolutions based on what you want.

That said, whenever you see a company instituting a "trick," it usually turns out to be a complete gimmick. To Nokia's great credit, this does not appear to be the case. The full-resolution images of 38Mp are impressive because they reveal optics that are only slightly behind the resolving power of the sensor. Obviously, contrast is a bit low, and edges are soft, but what were you expecting from a lens the size of Polly Pocket's dinner plates?

Truly impressive is the amount of noise at full-resolution. Namely, there's very little. When downsized to something more reasonable, such as 12Mp, the noise fades significantly and you were left with what are the absolute best color transitions ever from a camera phone.


At the link I provided earlier, Nokia has posted three demo shots. Obviously, as with all demo shots, these are staged for maximum impact, but even with that taken into consideration, these are amazing photos. Colors are rich, reds and blues are deep and powerful, noise levels are very low, and the detail from such a small lens is impressive. Nokia should be very proud of their engineers.

The comparison that I posted above considers one of their demo shots. The full-res photo is soft and a little washed out, but it handles the bright light of the sun very well. But when downsized to 12Mp, the impression of it being a camera phone absolutely disappears. It now looks like a high-end P&S camera. The dark blue sky, which I am hoping was not specifically touched up in Photoshop or any shenanigans like that, looks great. This is where the tiny sensor would devastate with noise levels, and the phone passes with flying colors.

I was a huge Nokia fan, and in many ways still am, and would love to own this phone. Unfortunately, they put their lot in with Microsoft, and I am in an absolutely monogamous relationship with Android. Even worse, this phone doesn't even use Windows Phone! It uses the now-officially-dead Symbian OS. So while this camera is impressive, it is a dead-end product, an orphan, a tech-demo, released simply because they can.

Sunday, February 19, 2012

DxOMark Starts Reviewing 4/3 Lenses

DxOMark, which has been woefully lacking in 4/3 and Micro 4/3 lens reviews, has decided to fill that hole in their line-up. I'm assuming that the problem wasn't them, but instead was Panasonic and Olympus not sending them production samples. If that was indeed the case, I chalk it up as symptomatic of those two companies' complete disconnect from the photographic world and their resultant belief that only point-&-shoot buyers are interested in their products. As such, there is no need to send samples to a website that deals almost exclusively with the enthusiast/pro segment of the market.

They have only done three reviews thus far, but they are all of very interesting lenses: The Panasonic X 14-42mm, the Olympus 12mm f/2.0, and the Olympus 45mm f/1.8. All three lenses perform very well as regards resolution.

The Panasonic is the real surprise. If it was faster, it would be a significant winner, but alas, it's not. It's still a great lens, just slow. I can't wait to see the tests of the Panasonic 14-45mm and 20mm f/1.7, both of which are gems that haven't left my side for two years. The 14-45mm is the best kit zoom that I have ever used and, based on my own tests, is noticeably better than the 14-42mm.

The two Olympus lenses put in the best performance that could reasonably be expected for lenses of their type. The 12mm does better than I expected seeing as I wasn't terribly impressed with its cost/performance ration when I first used it. I enjoy the speed, but when compared to the similarly priced and much wider Panasonic 7-14mm, its sheen dulls a bit.

The 45mm shines again, but their test doesn't tell the whole story. The contrast of the 45mm was where I was truly impressed. I like sharpness, but high-contrast at edges is what truly makes an image pop at most resolutions. The Canon 85mm f/1.8 is one of the best lenses on the planet, and the Oly stands up well. If you are invested in Micro 4/3 and plan on staying in it, the 45mm should absolutely be in your kit.

These reviews remind me why I like 4/3 and why I think the format has potential. If only the cameras could catch up to the lenses, the system would be truly great.

Wednesday, February 15, 2012

Canon G1X RAW Images Posted

DPReview has posted RAW images of the new Canon G1X and they are impressive. Again, the sensor impresses, the lens, notsomuch. It gets very soft at the edges at longer lengths. At less extreme distances, the lens appears to be quite sharp, but all things considered, it disappoints a bit for the price. I don't expect razor-sharp images for the price, but I do expect something usable.

The sensor is very impressive. The JPEG images posted by Focus Numerique line up nicely with the results on DPReview. Basically, the sensor in the Canon outperforms the Panasonic GX1 and GH2 at all ISO's. From ISO800 and up, the difference is nearly a full stop. Lower ISO's reveal that Canon appears to cooking its RAW's, but with even that taken into account, this camera nearly dances toe-to-toe with full APS-C cameras.

DxOMark shows that some of the best elements of the RAW files, such as extended color depth and dynamic range, even with the camera's good noise performance, are where the smaller sensor falls down. The disappointing DxO performance puts this at the top of the 4/3 world, but still behind the larger SLR and NEX pack.

I'm feeling that I can make a decent determination of the camera, and while it's not terrible, it's not good. All of the benefits of owning a Canon are obviously there, workflow will be good, interface and controls will be excellent, and available accessories are unmatched. But even considering that, the price is just way too high. I'm sure that it will sell well. Hell, the Nikon V1/J1 sold well, and it's a ridiculously over-priced camera.

If the lens was faster, the $800 price point would be more reasonable. Combined with some pretty significant softness at the far end of the lens, I just can't see through to such a price premium. If the camera was $100 cheaper, it would be much better positioned. But, again, that's from the perspective of a camera freak, and as Nikon proved with the V1/J1, an overpriced camera, well-marketed, can still sell like gangbusters.

Monday, February 13, 2012

Panasonic GX1 Gets Real Review

DPReview, the undisputed juggernaut of online camera reviews, has posted its review of the new Panasonic GX1. The sensor was a pretty big disappointment when DxOMark ran it through its test, with a score that barely defeated the now-old 12Mp sensor found in most earlier Micro 4/3 cameras.

DPReview's test holds few surprises. Panasonic has refined all of the practical elements of the camera and delivered the first camera that can be considered a successor to the amazing GF1. They mention a few things that are disappointing to lose, like the drive lever, but ignore the ridiculous placement of an iA button that you cannot reassign to a function that, I dunno', someone might use.

Overall, though, it is apparent that Panasonic did a good job of moving the GF1 philosophy forward. Unfortunately, they needed a lot more than simply moving forward. They needed a leap. The market is very different than the market that greeted the GF1. Sony has really outdone themselves with the NEX-5n and NEX-7, and while the lenses aren't quite as compact, they are of impressive quality. The mirrorless market has moved on quickly, and Panasonic has yet to bring its A-game.

Wednesday, February 8, 2012

Sigma SD1 Drops Price By More Than Two Thirds

In a move that absolutely no one expected, the hilariously overpriced Sigma SD1 has had its first price drop.

The camera started off life at $8,000, dropped to around $7,000 before launch, and has, I'm sure, sold no more than five across the entire planet. Sigma's CEO tries to save face by saying that manufacturing issues forced them to price the camera as high as they did. This is, of course, total nonsense.

If your company has a product that exceeds manufacturing expectations to the degree that this camera supposedly did, you don't increase the price to the point where no sane person would buy it, you either eat the loss or you don't make the product. Those are the only two choices that make good business sense.

Eating the loss allows you to ramp up production until costs drop, you then make up the loss on later profits. Not making the product at all allows you to continue developing the product in preparation for the day with it can be made. Pricing the product into the stratosphere maximizes your loss.

No, Sigma's pricing decisions was based on wishful thinking. They hoped that people would buy the marketing and actually think that this camera was the equivalent of a medium format camera. I'd call it arrogance if it hadn't been so foolish.

At its new price, the camera is a much more feasible tool. Obviously, enthusiast photographers who will want to do a wide variety of things with their camera need not apply. The lack of video and bottom-of-the-class ISO performance limits the camera's versatility. But for two groups of people, the camera is not un-tempting.

Studio photographers have been dealing with moire ever since the inception of digital cameras. It's an unavoidable consequence of the Bayer array. The Foveon sensor is immune to this effect. 100% immune. As digital cameras, especially those aimed at pros, are increasingly including light anti-aliasing filters, or no filter at all, the moire effect is maximized.

For most people, it's not much of an issue, but for anyone who has ever done shots in a studio with lots of fabric, the pain of moire is well-known. For them, the SD1, especially at its current cost, could make a lot of sense.

The other group that may want to look at this camera are landscape photogs. At base ISO, the performance of the Foveon sensor is very good. It falls off a cliff after ISO640, but that's irrelevant. Colors are rich and varied, and, most importantly, transitions into shadow are exquisite. As long as you aren't a resolution whore, the new Foveon sensor is a good choice. Sigma claims a resolution equivalent of 30MP. I've found it closer to 24-25MP, but I still think that is enough for most landscape applications.

The SD1 is not the camera for me, but it can be the camera for others. Its new price stops me from laughing in their face and instead makes me take their product quite seriously indeed. I think that it still needs work and refinement. It needs details that other, cheaper cameras have. It also needs much better workflow. That is something that companies who have never dealt with the pro market don't realize is important. Workflow can literally make or break a system.

I wish Sigma the best of luck, and I hope that the inevitable SD2 fixes all of the issues with this camera and truly brings Foveon into the mainstream.

Monday, February 6, 2012

The Olympus OM-D Is Not Enough

The shape of Olympus' line-up for the next year is pretty much solidified and after more hype than it likely deserved, Olympus has delivered more than it has in the past, but not nearly enough to make up for its past failures.

First, the good. The autofocus on the new camera has been tweaked to play nice with Olympus' Zuiko lenses, which means that autofocus on darlings like the 12-60mm and 50-200mm will finally be better than total crap. Autofocus on new lenses should be excellent, which is an oft-overlooked element of camera discussion. Granted, that's because camera companies frequently tout "improved autofocus" as code for "we actually haven't done anything to the camera."

Olympus at least showed that they weren't lying when the E-P3 released with autofocus that actually was better than any mirrorless camera on the market. That gives them a little more credibility. But the rest of the specs are woefully inadequate for the premium price that they are asking.

As with autofocus, a camera can be made in the details. But aside from the autofocus and weather-sealing, the OM-D is lacking specs that cameras that cost half the price have. Cameras in the same price range have details that devastate the OM-D. Details like a high-resolution viewfinder as in the Sony NEX-7, excellent ISO performance like the Sony NEX-5n and Samsung NX200, goodies out the ass like the Nikon D7000 and Sony A77, or world-beating video performance like the Panasonic GH2.

In both details and performance, the OM-D is handily beaten by almost every competitor on the market. If only I could bring myself to hate it, I'd feel much better.

Unfortunately, that weather-sealing is absolutely unique. No other camera in the price range has it. Truly, I can't think of many cameras below $2,000 that have it aside from the Nikon D300s and the Canon EOS 7D. That one feature goes a long way towards making me want this camera. I've been using a Panasonic GF1 since its release, and the lack of weather-sealing has been an issue more than once. But am I willing expend over $1,000 to get a camera that will function identically save for the sealing? No!

I am already invested in the system, and I need much more to get me excited. This camera isn't a complete disappointment, but it's not exciting either.

Another Apple


From fō-tō-gră-fē Photographs

Wednesday, February 1, 2012

Everyone Can Calm Down. Olympus Is Still Incompetent.

Some of the key specs of the upcoming Olympus OM-D have been leaked. First, the camera's official name is more in-line with Oly's current naming scheme, the E-M5. Second, we now know what sensor the camera will be using.

As for my criticisms, even the name of the camera is stupid. We already have the E-PM, and now the E-M. We have the GX1, and Canon has the G1X. The alphabet soup of the modern camera world is so jaw-droppingly, eye-poppingly, soul-crushingly inane that it makes me want to abandon it all and just learn how to paint really quickly.

But that is neither here nor there. Or hither or yonder. We are here to talk about the camera, and the heart of any camera is the sensor. That's why camera geeks always wait with bated, saliva-soaked breath for the release of the RAW files for any new camera. Because if you already have a camera in the system, if the sensor doesn't offer a big increase in quality, you have no reason to upgrade your camera. That's why the recent generation of Sony sensors generated a solid eighteen months of excitement.

So it was with similar excitement that the micro 4/3 world waited for the new Olympus camera. No one was able to confirm that it was just another repurposed Panasonic sensor, so we all hoped that it was Sony who had made it. With Sony's sensors being easily the best on the market, and Micro 4/3 hurting the hardest in color and dynamic range, this was exciting.

As Olympus has become adept at doing, they have disappointed us. The sensor is exactly the same sensor as in the GX1 and G3, itself something of a disappointment when it premiered a year ago. That means that the RAW files will be the same. That means that the image quality between this camera and the similarly priced Sony NEX-7 will be utterly enormous.

It's not all bad news. The camera looks great and is weather-sealed, which is an oft-neglected feature that all good photos appreciate. The supporting hardware is good, with very fast shooting speeds. The buffer is disappointing, but 10 shots is still adequate. And if the autofocus of the camera has been worked such that it finally plays nice with Olympus' pro Zuiko glass, that's a huge plus.

That said, while a camera is certainly more than its sensor, the sensor is still its heart. It doesn't matter if the camera takes ten shots per second if those shots are all sub-standard. Truly, the fact that the rest of the camera appears good only makes the state of the sensor that much more disappointing.

All of our hope now lies with the GH3. If we don't see something significant in it, I know that I will be lost to Sony or Fuji, and I suspect that many more enthusiasts will follow suit. It's put-up-or-shut-up time for the 4/3 group.