Sigma SD15 Preview

The Sigma SD15 is the successor to the Sigma SD14. Both are characterized by their use of a Foveon sensor.
Unlike the sensors in most DSLRs, which use separate pixels to detect red, green and blue light, in a Foveon type sensor each pixel generates a red, green and blue signal. The Foveon pixels take advantage of the fact that light of different colors can penetrate to different depths in silicon. By measuring the signal generated at different depths, the color of the light can be determined.

So far the only DSLRs to use Foveon sensors have been made by Sigma and the sensor used has 4.65 million photosites. However, the pixel count is multiplied by three because each site generates signals for red, green and blue. Sigma and Foveon call this a 14.1MP sensor. Though you can’t really directly compare it with a conventional 14MP sensor (using a Bayer color matrix). Tests show that a 4.65MP Foveon sensor generates higher image quality than a 4.65MP conventional sensor, but it’s not really equal to a 14MP conventional sensor. Perhaps it’s more closely equivalent to a 10MP Bayer matrix sensor. The comparison is
difficult because it depends if you look at B&W resolution or color resolution, and it depends on the quality software used to eliminate moire patterns, which tend to be much stronger with conventional Bayer matrix sensors.

It’s a bit of a mixed bag. In some respects the 4.65MP (or 14.1MP) Foveon sensor may be better than a 10 MP conventional sensor, in some respects it may be worse. It’s certainly interesting technology, though it really hasn’t caught on with anyone but Sigma.

The Sigma SD15 is quite similar to the earlier Sigma SD14. They use the same sensor, but the SD15 has a more advanced processing engine (the “True II Processor”) and the LCD is increased in size from 2.5" to 3". In all other respects the two cameras appear to have identical specifications and features.

The Sigma SD15 is compatible with Sigma’s own line of lenses. Though the mount is somewhat similar to the Canon EOS mount, Canon lenses cannot be used on Sigma DSLRs.

h2. Official Sigma SD15 Press Release

Sigma announces the development of the SD15, a 14 megapixel (2,652×1,768×3 layers) Digital SLR camera.

The Sigma Corporation (COO: Kazuto Yamaki) is pleased to announce the new SIGMA SD15 digital SLR camera.

This camera is the latest model in Sigma’s digital SLR camera SD series, powered by the 14 megapixel Foveon X3 direct-image-sensor it can capture all primary RGB colors at each and every pixel location arranged in three layers.

The developing SD15 incorporates the “TRUE II” new image processing engine. It provides high resolution power and reproduces high definition images rich in gradation and impressive three-dimensional detail. Incorporation of the large 3.0 inch LCD monitor and improved processing speed provide ease of operation of the camera.

h3. Development of the SD15

Since October 2002, Sigma has introduced three digital SLR cameras, the SD9, SD10 and SD14 to the market. In March 2008 Sigma also introduced a high-end compact digital camera, the DP1 which uses the same large image sensor as is featured in Sigma’s digital SLR cameras. They have established a strong following from a wide range of photographers, both amateur and professional. Photographers expressed a desire to incorporate the image processing engine “TRUE”, which is used in the DP1, into a digital SLR camera. In order to meet this demand, the SD15, with high resolution direct image sensor, has been designed around the new “TRUE II” image processing engine. This combination delivers superior image quality as well as improved processing speed, operation and performance.

h2. Brief Specifications of the Sigma SD15

Sensor Foveon, 2640 × 1760 pixels (x3), 20.7 × 13.8 mm
Lens Mount Sigma SA Bayonet
ISO settings 100, 200, 400, 800, 1600 (extended mode)
Shutter speeds 1/4000 sec to 30 sec + Bulb
Exposure compensation -3 to +3 EV in 0.3 EV steps
Metering 8-segment, center, cw average
Continuous Drive 3 fps
Storage Compact Flash Type I or II
LCD 3.0 ", 150,000 pixels
Weight (inc. batteries) 750 g (26.5 oz)
Dimensions 144 × 107 × 81 mm (5.7 × 4.2 × 3.2 in)

More

Original text ©2008 Bob Atkins.

Sign in or Sign up to post response

    • FOVEON SENSOR IMAGE QUALITY ,

      The Foveon sensor It's not a Bayer pattern sensor . It has its own Paradigm breaking architecture , that do not deserve comparisons . Therefore Resolution , ISO and Noise expectations, based on comparisons to a Bayer sensor ,have no meaning when applied to a Foveon sensor.

      A Foveon sensor behaves similar to 100 ISO Transparency film. More specifically like Kodachrome .

      Foveon sensors has proved to be exceptional at controlled artistic photography, and it is the only sensor that , because of its absence of Anti Alias filtering (Be it Hardware or Software based ) is really Lens Limited , rather than Sensor limited. That is , it is the lens Resolving Power that has the most influence on the Overall Camera Resolution .

      Hundreds of people that have replaced the original Sigma SA mount for a Leica-R or Nikon-F can attest to the Unbounded Image Quality that the camera can produce , when liberated from the restrictions of the original , low quality lenses , that effectively act as a Blurring filter themselves. Leica-R lenses and Zeiss ZF lenses ,exhibit a remarkable Synergy with the Foveon sensor , since all of them are optimized for Maximum Sharpness and Resolution , at Full Aperture.

      The incorporation of a Parallel Processing pipeline , in the upcoming SD15 model, promise to increase the overall Performance as well as the Image Quality , even further. You can make your own informed judgements , regarding liberated ,converted cameras , by visiting our galleries and our Forum:

      SIGMA CUM LAUDE

      http://www.photo.net/photos/Luis-A-Guevara
      http://www.pbase.com
      http://gallery.leica-users.org/v/LUIS+A+GUEVARA/
      http://luis-a-guevara.deviantart.com/
      http://www.summiluxart.com/
      http://www.sigmacumlaude.com/

    • To post a reply Sign In
    • Pics with the Foveon sensor look no different or better than anything from Nikon or Canon, and even saying 4mp compares to 10 seems ridiculous to me.
    • To post a reply Sign In
    • REALLY?

      Sigma SD9 with Leica 19f2.8 Elmarit-R on a Leica Apo 2X Extender-R .

    • To post a reply Sign In
    • I agree there is something unique and special about Foveon images and I find it odd there is such a polarization of opinion. To my eye, it is as close to film as anything I have seen in the digital world. The above photograph says it all and I doubt even a full frame sensor could come close to providing the lush and realistic texture of the Foveon. This quality makes every other digital image from bayer sensors seem flat and lifeless.
    • To post a reply Sign In
    • one of the things i love about the sigma sd series of dslr cameras is the fact that they are really like a reversion back to full manual!!! great quality photos without all the silly and sometimes stupid or confusing options and menu routes, also none of that in camera editing or filters. i like having to do almost everything manually, reminds me of my pentax honeywell.
    • To post a reply Sign In
    • You are absolutely right ,manual operation is actually liberating . That is the very reason for converting your camera to manual lenses.

      This is the camera that took the images displayed above. A Sigma SD9 converted to Leica-R by sigmacumlaude.com

    • To post a reply Sign In
    • I have to say that the 1280x853 rendition of some washed-out, sick-looking foliage above - with a cute vignetting effect - quite conclusively proves the Foveon sensor's merits. As for the SD15, which appears to have been in the works since 2008, I'm not sure which of its specifications appeals to me most. The absolute top ISO of 1600 will discipline me to learn about flash lighting rather than relying on grotty old available light. The SD lens mount frees me from the worry of using inferior lenses by manufacturers other than Sigma. Sigma's lens range is doubly appealing because of its huge diversity; on the surface they appear to only make a few dozen lenses, but in practice each individual lens has a unique performance profile, unlike other manufacturers who try very hard to make each copy of their lenses identical. The fact of having to pay extra to have the entire lens mount removed and replaced with a new lens mount so that I can use other lenses by other manufacturers is a tiny inconvenience and I'm glad that Sigma gave users the freedom to do this. The 150,000 pixel screen is of no consequence to me, because only weak men use the screen for any purpose. People in the film era didn't have a screen and they did fine. The lack of any other modern features - a spirit level, live view, etc - are also of no consequence because they are just crutches for techno-nerds. The top sync speed of 1/180 does not bother me because flash is horrible and fit only for tasteless people who I despise. Oh, I can't keep it up. This sounds like the kind of camera Sigma needed to release in 2005, five years ago. In an ideal world this should have a ten-megapixel, full-frame Foveon sensor. There are people who would kill for a camera with the above specification plus a ten-megapixel, full-frame Foveon sensor. Not many people but people nonetheless, human beings like you and I. Even with an absolute maximum ISO of 1600 and flash sync of 1/180 and so forth. As it stands it comes across as a rehash of the SD14, which was old-fashioned when it was new. A part of me admires Sigma's perverse persistence, but it is not the part of me that controls my wallet, and I cannot see this being a roaring sales success.
    • To post a reply Sign In
    • I know I'm late to this discussion, but Luis, you can't show the superiority of a given camera unless you compare it with that taken with another system. One photo by itself proves nothing.

    • To post a reply Sign In
    •  

      Simply visit Digital photography review, and you can see in the resolution comparisons, which are nothing if not impartial and scientific in their approach, that the DP2, which uses the same sensor as the SD14 & SD15, outperforms the 12mp Olympus EP-1, as well as every other camera they compared it to within its class.

      Here's their verdict: "The Sigma's results are rather over-sharpened by our standard workflow because it includes a sharpening step to compensate for the low-pass/anti-aliasing filter (which the Sigma doesn't have). Even so, the level of detail being rendered is clearly very high and not dissimilar to that of a well-processed image from a camera with a 12mp conventional sensor such as the E-P1. Per-pixel sharpness on the DP2 is way ahead of the E-P1 and even in a large print it seems ulikely that the extra pixels on the Olympus sensor would give you any advantage."

      There IS a lot of polarization around the Foveon technology, because it is different, and it uses a novel approach from the other image sensors out there.  While traditional camera owners don't appreciate Sigma playing fast and loose with such absolutes as "megapixels"  People who own and have grown fond of their foveon digital cameras don't like it when people bash them without fully understanding the technology.

      4 megapixels are devoted to each of 3 RGB layers, just like the old Volare Leaf digital backs took 3 6mp images, with a rotating RGB color wheel.  Did the Leaf Volare produce 6mp images?  No, of course not, it would be absurd to think so.  It took extremely vivid 18mp images (still lifes only, in studio only).  The foveon sensor does the same thing as any of the early high end sensors, only it receives all 3 layers simultaneously, instead of one at a time.  

      I think another polarizing factor is that Sigma isn't as sexy a brand as Canon and Nikon.  Most people who own Sigma lenses do so out of economic necessity, because they cost far less than their competitors.  The very idea that Sigma should come along and dare to herald some new technology galls your typical camera enthusiast, who has spent hours pampering their analogue limited edition Nikons and daydreaming about getting a Topcon 10X macro set up for their Super D.  I have matched up serial numbers on plenty of lenses and cams, and know how flat the name "Sigma" looks on a 28mm Wide Angle lens you thought was SMC Takumar.  But, just as it's hard for a Nikon man to take a Sigma camera seriously, it's equally difficult for somebody who is happy with their Sigma and amazed by it's results to "admit" that it's all a figment of their imagination!  The fact that so many photographers are willing to defend a product from such a mediocre company should be telling you something.

      Where Sigma cameras do fall flat is in the arena of higher ISO values, and most other processing related issues, like multi-shot processing.  This is due to the fact that Sigma is the lone proponent of this technology.  Sigma doesn't have the benefit of years of heated competition to improve their processing technology in the hopes of differentiating their DSLR from the other DSLRs in the market using nearly identical image data coming from similar, ay, often identical image sensors.  Sigma can't reverse engineer Canon's processing engine and stick it in their cameras, so they have to travel the same slow curve that every other camera maker has done over the past 12 years.

      The only reason the DP2s and the SD15 exist is due to improvements in image processing.  Indeed why should Sigma move up to a full frame Foveon sensor when their current sensors produce images competitive with the newest cameras on the market?

      Food for thought, Leica obviously used the Sigma DP1 & 2 as their benchmark, when developing the X1. Leica doesn't spend millions to develop a brand new species of digital camera using the Sigma DP's specifications as a measuring stick because they aren't good cameras. They are relying on camera snobbery to capture your market share, and are happy to not have to come up with something actually new, like, all by themselves.

      I think we all have very nice cameras, and Bayer sensor zealots shouldn't be so fast to deride a new technology.  Every true camera enthusiast knows that it is only by challenging the status quo that we, the end users, get improved technology.  So Sigma got lucky, let them have this one, it won't make their lenses better than... well, almost anyones.

      Once you adjust your eye and workflow to using the X3F raw files, you find there is little difference between the output from a micro 4/3 camera and that of a Foveon X3 sensor.  12MP, 14MP, 10MP, in the end it doesn't really matter that much in terms of over all image quality.  I for one still rather like my Epson R-D1, and it has a 6mp sensor.  If Epson would incorporate an X3 sensor in it's modified Bessa Body, I doubt anybody here would complain.  

      And yes, it would blow the M8/8.2 out of the water.  

       

    • To post a reply Sign In
    • Foveon may well be interesting technology, but it's never going to catch on if Sigma doesn't improve it's snail pace of releasing more camera models with more significant differences/improvements. As things stand, it will be a very rare first-time dslr buyer who would venture into such unusual technological territory. And for those of us who already own one or more dslr's it seems very unlikely that we would be so eager to try out foveon sensors AND yet another lens mount.

       

      Now if Sigma would expand on their model of distributing the same lenses in different dslr mounts and try introducing a model of foveon dslr that came in one or more of the popular mounts, that might be a different story. You might want to take the leap into foveon land because you could use a foveon body in Nikon mount with your existing Nikon lenses, I might be willing try it out as a second body in alpha mount if it didn't mean starting a whole separate lens collection besides my Sony/Minolta lenses, and someone else in eos mount might be curious to try it with their Canon optics, etc etc.

       

      No doubt there would be engineering challenges getting the AF drive of a multi-mount body design to work properly through all the different types of popular digital mounts, but even if those cannot be overcome for emulating some of the big brands, in reality most systems now have a fair number of lenses with their own AF motor, so even that might not be such a big hurdle after all.

    • To post a reply Sign In
    • Paul,

      You nailed the head with your comment.  As a company who has grown up leaching the customers of every other camera company, I can see why they haven't ventured into a more universal camera, mount-wise, to the detriment of the foveon technology.  

      (added: They make their money off lenses, so why create a product that encourages the use of lenses made by their competitors?  Make no mistake, a multi-mount SD16 is the counter-intuitive solution)

      a 42m adapter was the first thing I bought, and now use my pentax lenses all the time.  on some website I saw that pentax autofocus lenses will work with sigma if... if you chop off the aperture pin, which I could never do, would never do, even though I only own 2 pentax af lenses that go to very broken 35mm cameras.  For someone interested in macro, a pentax lens with no aperture control is useless.

      I just bought a 49 dollar minolta to Sigma mount adaptor, because I scored a fabulous macro lens for the 7000 that was 12 dollars.  In the end I paid 61 bucks for a lens and adapter. 

      Like that Sigma-Zealot indicated in his post, there are entire sites dedicated to hacking sigma dslrs for use with Leica R, Nikon, Canon, and even Alpha/minolta lenses.

      My initial comment was just to elucidate the technology of the X3, because I think it's a shame that it is so misrepresented in the press, and in the minds of readers.  I think that Sigma should take a page from Kodak's business plan and make this sensor available to the entire industry, which they have not done, by the way.

       

    • To post a reply Sign In
    • I also wonder why Sigma cameras polarize opinion as well. But heres a secret - I have an SD 14 (well thats not the secret)which I purchased brand new in box for GBP250 recently, with no lens. However due to the fact that Sigma SD fit are quite rare/unpopular I have been able to pick up the following -and this is the real secret:

      Sigma 17-35 90GBP brand new

      Sigma 28-200mm 40GBP in as new condition

      Sigma 18-55mm 30GBP in as new condition

      Sigma EF-500 DG 60GBP

      So for a modest outlay I think I have an excellent but DIFFERENT camera set up to take out with me (I also use Sony and canon DSLRS)and I do not worry about the 4m/14mpix "debate" but am taking pictures with it.

      Its a  harder camera to use - boy do I miss my auto everthing button ! - but to get the best out of it I find I have to think much more about what I'm doing;it does feel like using a film camera in terms of not rattling off lots of shots just because I can (and I know I could do the same with my Sony and canon, but the auto button is too tempting).

      So far I've taken some really poor shots with it and some good shots as well - I am finding it challenging but when its good its very good! Better than Bayer or my Sony and canon - who knows? But I have to say the sigma images have a different "air"about them (I'm struggling to think of an appropriate word)which I like, especially for nature and landscape photography.

      So calm down a bit (both sides of the "debate")and be happy with what you are producing regardless of the "box" your picture comes out of -its the picture thats really interesting isn't it?

       

    • To post a reply Sign In
    • Why Foveon? Some answers here:

      http://www.flickr.com/photos/dougbm/sets/72157624749210093/

       

       

    • To post a reply Sign In
    • Same shot as previous (but on a different slightly duller day) taken with Canon 5D.

    • To post a reply Sign In
    • Crop comparisons

    • To post a reply Sign In
    • Another angle

    • To post a reply Sign In
    • The previous comparison seemed a little soft for the 5D (maybe a touch of camera shake). Remember it was a duller day so not entirely fair but the 5D images always require more working up in Photoshop to give their best.

    • To post a reply Sign In
    • Since Foveon´s first announcement of X3 and publishing of sample photos hit photography fairs around the world, I was intrigued by this seemingly sensible arrangement of color registration in layers, similar to film, with as many sensitive layers as broadcast color channels rgb, but as it turned out Foveon has been ignorant to color problems recognized by film makers even before digital photography was born. Three base colors just aren´t enough apparently, to render nice skin tones in a credible manner. The first images I saw from an experimental camera made by Foveon, struck me as having strange cyan/greenish skin tones. I was told at the stand in Cologne, it was on account of the late time in the day, and the foliage in the setting surroundings. Later when I tried production samples of SD7, SD10, SD14 and also DP1 as well as DP2, this color tint prevailed particularly in skin shots, but also in the corners of particularly dimly shot images. Isn´t it peculiar, that you never seem to see convincing glamour photographs or available light portraits from an SD camera? Vignetting as reported by others is horrible, particularly noticeable in the DP compacts. Initially I thought it was caused by inefficient software, but after seeing the same flaws in later models over and over, I have become convinced that SD15 or SD16 whatever SD must be waste of time and money. I recognize SIGMA´s occasionally great lenses, but their camera bodies are just not attractive, well functioning or even thought through. The analogy to russians slrs which some critics have brought is quite justified in my opinion. If you hold a Nikon D300S, or D700 in your hand for 2 minutes, any SIGMA DSLR to date will feel like junk. The SD14 was able to make sharp pictures in fine weather, but any DSLR can do that these days, and high quality 3´rd party raw converters also abound, so please forget about Foveon X3. The driveling nonsense you hear over and over from enthusiasts coining it " the closest thing to film " is nauseating at best. I even bought it myself, at first, when everyone were still looking for matches to analog media, and I thought the SD7 came close to Kodak EPR, or anyway closer than FUJI S2 Pro and Nikon D100 ccd cameras, but as sensors have improved over the years, it´s now evident to me, that comparing digital products to film only has relevance to people stuck in the past. CMOS sensors from SONY(NIKON) and CANON have surpassed film in every way imaginable. I took my last shots with Foveon X3 using DP2 and SD14, and I will never waste another second on a SIGMA product using the same type of sensor again. I would however love to see SIGMA engage a former NIKON body designer, and build me a camera body with a SONY CMOS, for my Leica R lenses.

    • To post a reply Sign In
    • Thanks to Doug Barry-Martin -

      The only useful test of sensors (or film or cameras or developers etc.) is to take comparison shots and select the ones that create the look you like.  For my eye, the differences in the nuance and detail in the folds of the materials in the last two shots is striking and strongly favors the Sigma.

      As always, much also depends on the subject.  I shoot nature and structure, rarely people, and I am very happy with the results from both the SD and DP1 cameras. 

      I read a review a while back that stated that the only criteria on which the DP-1 beat a high-mp Bayer camera was image quality...and that was intended to a negative comment about the Sigma!  Obviously there are many uses for cameras, but since image quality is one of the aspects that is important to me, I'll just keep pursuing that with whatever comes along.

       

    • To post a reply Sign In
    • To clear up the errors:

      The SD15 does take raw and jpeg at the same time.

      Any claim that this 14.1 MP imager with 4.7 million spatial locations is equivalent to a 4.7 MP Bayer camera needs a comparison to support it. The same applies to the claim that the SD1 with about 45 MP at 15 million spatial locations is the same as a 15 MP Bayer camera. I can guarantee any doubters that a comparison of a camera with a Bayer imager having a similar number of pixels as a Sigma camera with about the same number of spatial locations using Foveon technology will make the Bayer camera look pretty silly. 

       

    • To post a reply Sign In
    • COMPARISON BETWEEN SD14 AND 60D.

      The SD14 has a much nicer viewfinder (better glass). The 60D is much faster (in all ways) with higher resolution sensor. This sometimes shows up the lower res SD14 and sometimes not. These shots are with a 17-70 lens on the SD14 and an 18-200 on the 60D. Funnily it is much quicker to change ISO on the SD14 (something you hardly do) than the 60D (I use auto ISO when possible, but you need to keep an eye on it). Really looking forward to the SD1!

      IMAGES TO FOLLOW

       

    • To post a reply Sign In
    • 60D

    • To post a reply Sign In
    • 100% crop

    • To post a reply Sign In
    • 100% crop

    • To post a reply Sign In
    • Full image

    • To post a reply Sign In
    • Full image

    • To post a reply Sign In
    • To my eye the SD14 wins this round especially as it has a nicer roll off to midtones from the highlights (especially in the fern image). The Foveon sensor doesn't blow highlights the way all Bayer ones seem to. I always shoot a half stop under on Bayer but leave it at 0 on Foveon. But ideally the comparison should be done with the same lens but unless I can borrow a Sigma 17-70 with Canon mount it is not possible. Next I will try using an L series lens on the 60D for comparison.

    • To post a reply Sign In
    • Not wanting to hog these comment pages but I had an interesting recent photo shoot. I had to photograph an artists painting of a palm tree on a white and a black canvas background. I used my SD14 with 17-70 lens and Canon 5D with 24-105L lens.

      The sigma outperformed the Canon and produced images that the client and I preferred. The Canon performed poorly on the palm with a black background producing a washed out background even though under exposed 2 stops. It showed better sharpness due to the lens but otherwise could not match the Sigmas dynamic range and soft roll offs.

      Will post pics if asked.

    • To post a reply Sign In

Sign in or Sign up to post response