Camera Test: Sigma SD14

The long-awaited Sigma SD14 DSLR boasts a unique sensor and a whole lot of megapixels. But how well does it perform?

By Michael J. McNamara, Popular Photography & Imaging

If patience is a virtue, photographers shopping for a new Sigma DSLR are saints. After all, it’s been three years since the last Sigma digital (the SD10), and it was more than six months between the announcement and availability of the new Sigma SD14 ($1,600, street, body only).

Clearly, a truly unique camera takes time. And the SD14 is unique. It’s the first and only DSLR to use a second-generation Foveon X3 sensor, which has a 1.7X lens factor, boasts 14.1 megapixels, and is promoted as a color-accurate, detail-obsessed, low-noise alternative to the CMOS and CCD sensors used in other DSLRs....

READ ON

____________

Mike

Featured Comment by Craig Norris: I find the Pop Photo review somewhat lacking in imagination. The reviewer only compares what the SD-14 can do that a Nikon D200 can also do, under general consumer snapshooting conditions, and finds the Nikon D200 the better camera. Fine. If I was a general consumer, I might agree.

In my photography business, I often shoot executive portraits, mostly head and shoulder, with a Canon 5D at about 70mm focal length. I always shoot at ISO 100 in the studio for maximum quality, because I do see noise and other image degradation at any higher ISO, even with a Canon 5D. But I am plagued in almost every session by moire patterns in the texture of some people's clothing.

The Sigma SD-14 would solve my problem, because it is free of moire artifacts, an important fact not mentioned or tested by the Pop Photo reviewer. What the SD-14 lacks is a tethered shooting facility, and that's the only reason I'm not rushing out to buy the SD-14. The lack of high ISO range is not an issue for me. I'll use my Canon 5D when I need to shoot at high ISO settings. Use the right tool for the right job.

My point in making this post is that camera reviewers often do the general public a disservice. They sometimes publish a rather negative conclusion about a camera, without really understanding that there is no such thing as a perfect camera (that's why I have 16 cameras) but each camera has some great strength that suits it to a particular application. The reviewer should highlight what that great strength might be for each camera under test.

My other gripe is that I've never seen any of the usual camera reviewers test high ISO image quality under low levels of tungsten lighting—which is the most common scenario when a general user needs a high ISO. The published tests, such as the Pop Photo review under discussion, waste space by showing high ISO images of outdoor scenery, or high ISO performance under daylight colour studio lighting. What's the point? Who shoots at high ISOs in the bright outdoors? Who shoots at high ISO with daylight white balance? Shift to tungsten white balance where the blue channel gets amplified inside the camera significantly, and the results are somewhat different, and infinitely more relevant.

I'm prompted to write today because I'm becoming very jaded and dissapointed by the generally poor quality of today's camera reviews. The only good camera reviews I regularly see are in Japanese camera magazines. I'm fortunate in being able to read a little bit of the Japanese text, enough to gain an insight into the reviewer's reports and findings.

After reading Japanese camera reviews for several years now, I feel embarrassed at the poor standard of most camera reviews published in Western countries. (Perhaps I'll start a new business publishing translations of the Japanese magazine articles....)

Craig Norris
Hong Kong

Comments

[Hugh Crawford]

I'd have to agree with Craig Norris's comments above , however the sidebar about removing the IR filter and using the SD14 for infra-red photography is an interesting example of highlighting a camera's unique attribute.

I remember when Consumer Reports would list the full frame focusing screen of the Nikon F cameras as a flaw because it would not accurately reflect the cropping of machine prints or cardboard slide mounts.

English translations of the Japanese camera magazines sounds like an excellent business idea.

[Jim Kofron]

As a "saint" (an SD10 owner), I had an opportunity to shoot the SD14 on a trip to Death Valley around the time of PMA this past March. Let me echo some of what Craig points out.

The strengths of this camera lie in the quality of the images in good to OK light. Color aliasing is a thing of the past, although if Craig were to photograph houndstooth, he could spot luminance aliasing. In photographing fine details in color (imagine fabrics, flowers, etc), you'll find that a CFA camera will do strange things when the resolution of the detail approaches that of the sensor (like in a fashion shoot). The Sigma will handle that transition more photographically gracefully.

Enlarging prints from a Foveon/Sigma file is usually a joy. The pixels are "better"---there's less roll-off, and it's easier to go big. On my SD10 (a 3.4/10 MP camera), I never hesitate to print to 12 x 18", and the results are usually outstanding (if I've held the camera still...). Very little postprocessing work is needed, and sharpening workflows are very straightforward. This is in part due to the acutance of the image---the lack of an AA filter, coupled with the lack of spatial interpolation that is done with CFA conversions lends a sharpness coefficient that is really nice.

Where else does the Sigma shine? Landscape, portrait, macro. The SD14's shutter is a joy---very quiet. Viewfinder is nice. I actually love the ergonomics---it's a shooters camera. No real need to dive into menus for any settings (I noticed Pop Photo was dismayed that there was no "mountain", "flower", and "head" settings...) This camera is not an SLR for morons. You need to understand very complicated things like Shutter Speed, Aperture, Focus, Exposure. If you do understand those heady attributes---they're all easily accessible with the control wheel. Mirror lock-up? On the dial. The week spent with the SD14 really made me appreciate this all new design that Sigma did (and quite well). Autofocus is better, handling is great, ergonomics are on par to better than the SD10 (which I love). I'm getting one this month. Not for the extra resolution (which is nice, but I don't need to print that large)---but for the handling, the better low light capabilities, and the power management (Li-ion vs CR-V3s).

Downsides? As Craig mentioned, low light tungsten is an issue. Blue channel noise can crop up, and it tends to be blotchy and difficult to remove (but B/W does, and you get great results---it's funny to think of one of the advantages of a full color sensor is phenominal B/W, but it is). In camera JPG's aren't as good as RAW, but having shot exclusively RAW with the SD10, it's not an issue for me. The buffer is still slower/smaller than most other cameras out there (even entry level ones). This doesn't usually affect me---but if you wanted to shoot sports or needed more than 6 frames quickly, you're going to have to adjust.

I hope that Sigma will soon make SD14 raw files publicly available along with the next revision of their raw converter, so that photographers can work with the files and see for themselves.

[Ben Marks]

This line of cameras has always been interesting, but I have always been prevented from jumping onto the bandwagon by the proprietary mount and the lack of fast primes. Canon has a proprietary mount, of course, but my 5D can be used with all of my Nikon, Pentax screw-mount and Leica-R glass via adapters. Canon also has good, fast prime lenses. Sigh, Foveon, we never knew yeh.

Ben Marks

[Steve Muskie]

I appreciated Craig Norris' comment and agree with most of it. One comment I'd make to his question, "Who shoots at high ISOs in the bright outdoors?" I do and many nature or sports photographers do. When using a 500mm or 600mm lens with a 2x extender and stopping down one stop to improve image quality (f11 working aperture), I often need to up the ISO to 640 or 800 or higher late in the day just to maintain a shutter speed of 1/250 or better. So those high ISO daylight reviews are useful to at least a few photographers!

[Nick Dormor]

http://www.rytterfalk.com/2007/04/15/img-sd14-and-a-studio-sd14-vs-5d/#more-153 Carl Rutterfalk gives an enjoyable perspective on the camera. I'd like to see more 'personal' reviews such as Carl's as they give you a true hands on approach which is far more insightful than pages of resolution charts and noise data.

[erlik]

Well, I agree with Craig Norris on almost all points. Particularly on the quality of reviews. Many reviewers focus on things they can quantify - noise, difference in "real" ISO values between cameras, AF points, sensor size to resolution ratio, and so on. (Not really so much different than the creeping featuritis in compact cams.) And sometimes they simply don't get a camera.

The case in point would be PopPhoto's hands-on of Olympus E-410: "E-410 will appeal to entry-level DSLR shooters looking for a smaller, lighter, and more affordable camera body." Right. Like only entry-level users would need something that you can almost put in your pocket. And the reviewer apparently tried to hold the camera as a traditional SLR...

As to the particular review, "Our image quality tests came down in favor of RAW files over JPEGs."

D'oh! Really? Who'd'a thunk that...

[Robert Roaldi]

I have what may be a dumb question. Quite a few stock agency sites have resolution minima for submitted photos based on width x height pixel counts. Six mpix (3000 x 2000) and more are common. How do Sigma users handle this?

[dasmb]

I've been shooting film lately. I've been getting all my film back as 2 megapixel scans (higher resolutions are much more expensive from my developer). And I've been shocked at how, in some ways, these scans look better than the 10 MP shots taken with a 50 year newer camera and a much better lens.

This is one reason why the Foveon sensor intrigues me. On paper it should be more film-like and deliver some attractive results in a very different way from Bayer sensors.

However, this review falls into the same trap as most reviews of Sigma cameras I've seen: it considers image detail to be solely an issue of luminance at a pixel level. They make bold statements about detail but only to refute the claims of 14 MP resolution. As well they should -- it's a marketing trick that isn't necessary.

But the real question here is this: how does the overall FEEL of the image compare with other styles of image capture? We know that Bayer sensors capture greater detail on resolution charts. But at the same time, a pixel that captures all the information available to it discretely is inherently more accurate. It's a bit like comparing a black and white film with a color film. The B&W is going to be more detailed and contrasty...it's also got a third as much information on it.

(The same goes for high ISO. You know what else is noisy at 1600? Film!)

So how do you compare these very different types of image capture? Forget about the magnifying glass: make the same 16x20 print with images from a Bayer sensor, the SD-14, and a nice film camera, and compare the prints.

Until then, we've got to read in between the lines here. With the SD-14, we have a camera whose output is fairly detailed when shot RAW at low sensitivies. It ostensibly shoots more accurate colour and at the very least captures more color information per pixel. It has no moire patterns. It has programmed auto and priority modes and decent autofocus. And it can, by design, be modified to capture infra-red.

Sounds like a novel and potentially powerful tool that solves a lot of shooters' gripes. If it had a Canon mount, I'd order one in a heartbeat.

[Gary Mercer]

I question the validity of the resolution tests performed by Popular Phootgraphy magazine as while they boast a million dollar testing lab and their lense test results are superb, their camera tests results are questionable especially in regards to the recent SD14 tests in the July 2007 Issue. In the article, you'll notice that they make a note that images were converted to .tiff files prior to the resolution tests. Why? Because the software that they use only supports Bayer sensored cameras and and there is not raw support for the foveon .x3f file format. What does this mean? First of all is the question of whether Pop Photo used the Sigma SPP 3.0 software to convert the Sigma raw images to .tiff format and what settings were used. Secondly, if Adobe Raw converter 4.1 was used there is some controversy regarding this plug in to adobe photoshop, as some users have found that it actually seems to add a blur filter to raw files imported into the program forcing the users to have to sharpen the images significantly prior to importing into the Adobe raw converter. I'm hoping that Popular Photography will take a serious look at the the Adobe raw converter program and report their findings to readers.
So in lieu of our lack of knowledge as to what program was used to convert the the Sigma SD14's raw images to .tiff files, we are left scratching our heads wondering how accurate were the image quality test performed by Popular Photography?
The programs used for testing were designed for testing bayered sensored cameras, not the unique foveon sensor and if the raw images from the camera were not used for the tests, but converted to .tiff files--how accurate are the results?--while I have respect for Popular Photography's effort, the lower than expected test results for a camera billed as a 14 megapixal camera are suspect at the least and in my opinion do not tell the whole story. Its hard to believe that McNamara, one of the industry's brilliant minds and experts would stand by these test results and not question whether the testing methods used are 100 percent accurate and a true reflection of the SD14's optical performance.
Finally, in order to insure that the lense tests were as accurate as possible, Sigma Corporation sent two Sigma SD14s for the test and calibrated a Sigma 30mm ex lense and kit lense for each camera respectively. This would insure that the tests results were as accurate as possible and give the highest possible results. Popular Photography does not tell readers what lenses are used for testing resolution and longtime Sigma DSLR users know that the quality of the lense used has a big impact on the optical resolution achieved with their cameras. So with specially calibrated lenses specific to each camera, I would expect much better test results than were reported by Popular Photography. Where the lenses left on the cameras they were calibrated for, or where they interchanged between cameras, further putting the test results into question.
In conclusion, industry methods of testing cameras and explaining the results to consumers have always been a point of argument between manufacturers and reviewers.
Popular Photography further makes the statement that the only people this camera is for are current Sigma SLR film camera users who want to upgrade to digital. This is not a fair statement in regards to the SD14. Why are only current Sigma camera users the only consumers who would be interested in this camera?
Finally,Other reviewers, like the Editor of Australia's ProPhoto magazine have stated that the SD14 blows away the D200 in image quality and resolution. His review was fair and balance, noting the SD14's strengths and weaknesses and indicated performance of the SD14 in league with the highest megapixal Bayer sensored SLRs currently available, but for much less.
Popular Photography has a responsiblity to perform accurate and fair tests of cameras as it has a large readership and can easily deal a death blow to any camera's sales which is something that should never be taken lightly.
McNamara's claims that Foveon is hyping the foveon chip and misleading consumers is no different than Popular Photography's test results on the SD14.

[Carl Rytterfalk]

If anyone is interested in trying out some SD14 RAW files - you're all welcome to download them over at http://www.rytterfalk.com - click the RAW category or follow this link http://www.rytterfalk.com/?cat=31

I recommend using either SPP (Sigma Photo Pro 2.3 - 3.1) or Raw Developer 1.7.2 +

// Carl Rytterfalk

[Tom Pengilly]

I purchased the SD14 for infrared photography. I can shoot IR at ISO 50 and keep the shutter speed higher than I need for hand held shots. A typical image taken in daylight is at iso 50, lens at 35mm, 160th sec at 5.6 and with the auto focus on.

I do own a Canon 40D. The Canon is faster, a little better built and has much less noise at iso 800 and above. The SD14 images above 800 are not really useable.

Canon people will hate this but the Sigma SD14 takes a better quality image at iso 200 or less. The real problem you can have with the SD14 is when you use Photoshop camera raw to convert the images. The chepo Sigma software works better.
So the Sigma has it's limitations but at $525 for a SD14 body compared to the $1,200 I paid for my Canon I have to say the Sigma is a real bargain for a beginner or someone on a budget. Oh yeah the Sigma battery life is pretty bad.

[Jaroslaw]

Hi, I'm traing to use SD14 for IR photograpy but I find big problems with setting the W/B for these type of photos.
The only info I can find is very missleading
as no one I found on the net uses Sigma.
If You could put some light on that subject I would bee more then greatfoul

PS. I use 89b 007 filter.

Thanks a million,

Jaroslaw

[Terry Eldred]

It took a bit to get use to all the lil nuances, but each time I found something new about it was like Christmas all over again, this is truly a remarkable camera compared to it the SD9 and SD10 in the processing software that came with the camera or you can down load it from there web site. It allows you to double the file size, I tried this fetcher for the first time the other day, and wow truly amazing quality, the file went from a 6 meg file to 13.5 meg file.
In door shooting is much better than the SD9 and SD10 due to the fact you can with push of a button change the intensity of the flash, and change the brightness to get rid of background shadows.
I have found that none of the SIGMA SD series cameras did well at ISO 800 or greater, for example a blue sky will look like cottage cheese, this is very flustering, or trying to catch a soccer player as he is about to make the winning goal, or a race car heading down the track or 200 miles an hour. When f/stop just won’t get done, then you have to keep going up in the ISO just to get a good shutter, here in is the problem. Until they can overcome this noise problem at hi ISO’s they should just get rid of it, since you can’t use it. Sure you can take photos at ISO 800 or ISO 1600 only to delete them later
On to that which make this Sigma SD14 a wonderful technological advancement in the world of digital photography. When placed in the right hands and in the right light, is where the SD14 comes into its element, at ISO100 and a tripod, I challenge any other digital camera to bet the quality or true color representation of the SD14, of course a high quality of lens is a necessity an EX is a must.
Fixing miner color problems are easily taken care of in the RAW conversion soft ware, which comes with the camera. And now with the ability to double the size of the photo, will I can’t wait for my next time to head out doors.