Sigma SD1 review

[Ray]

Ray - I have three words of advice on this topic... Remember Occam's Razor.

Michael

Yes, I remember it quite well, Michael. I often listen to the weekly Science program on ABC National Radio. (That's the Australian Broadcasting Corporation, not the American Broadcasting Corporation)

But my understanding is, in order to apply Occam's Razor, one has to be aware of the competing theories before one can choose the simplest.

We've ruled out the Japanese tsunami. Perhaps we've ruled out unforseen and 'expensive-to-rectify' problems at the manufacturing stage. Have we ruled out the imminent arrival of a 'killer' camera from Canon?

My imagination is fertile on this topic, but I have no wish to have the topic closed because of libelous scenarios that I might innocently create.   

[peterzpicts]

A warm thank you to Michael and Nick for doing a great review and Michaels commentary on the SD1's insane pricing. Spot on IMHO
Sigma must be very serious about this camera if they loaned Lawrence and a selection of their best lenses for the review.
I have been shooting with a SD14 for three years and Nikon D70/90 for six years. I do not make my living with my cameras but derive a lot of joy for family and friend with my results.
The bottom line for me is no matter how refined and competent my Nikon gear is I cannot produce the visually stunning shots I get out of the Sigma.  My wife can see the difference, their is something special about the photos I get from the X3 sensor. Their is a realism that we lost with film becoming unpractical for regular folks today many call it Sigmachrome. If it were not for those redeeming qualities my SD14 would have been at the bottom of the Columbia River or sold for pennies on the dollar because as a piece of equipment to shoot with it is a bag of hurt.
Many say the SD15 is much improved functionally, what the SD14 should have been in the first place. But call me spoiled, I am too used to getting more when I upgrade my gadgets, like 2x the computing power.
I would be figuring out the SD1 today if it were not for the D3x size price, I have to work for my money and raise 3 kids.
Sigma's camera's are not at the point where they can be the be all for every body, when shooting sports, theatre, or general junk the Nikon gets the nod, something special like landscape, portraits, trains(I am a railfan) the Sigma gets the nod.
BTW I don't know where folks get the lack of color sensitivity or lack of DR I find that to be the opposite. Shooting on a cloudy day with the SD14 is truly a joy in its own right.
Paul M. Petersen
my gallery
www.peterzpicts.smugmug.com

[Rob C]

If you are confident this is not the business model, then please explain what you think is the business model. One has to credit the business managers of Sigma Corporation having at least a modicum of nous.I understand that National Semiconductors and Dongbu Electronics are involved in the manufacture of the Foveon sensor.

If it is true these companies have 'tooled-up' to produce large quantities of this Foveon sensor, then it makes no sense whatsoever to charge such a high price for the camera.

The only sensible explanation I can think of is that it is now realised that yields, at least for the forseeable future, for whatever reason, are going to be low.

The reason could be that the manufacture of some essential component of the camera or the sensor has been disrupted by the tsunami in Japan, or that this particular design of Foveon sensor, with it's significantly higher resolution than previous designs, really is very expensive to manufacture.

It could be the case that some time ago there was a board meeting at Sigma Corporation where it was suggested that the project would have to be cancelled or at least delayed for a significant time because of the scarcity of certain essential components which would inevitably result in very small production runs.

Perhaps some bright spark in the boardroom suggested that an alternative to cancelling or delaying the project could be to sell whatever few units that could be assembled, at a sufficiently high price that would ensure demand equalled supply.

Got it? 

Edsel, anyone?

Rob C

[fredjeang]

A warm thank you to Michael and Nick for doing a great review and Michaels commentary on the SD1's insane pricing. Spot on IMHO
Sigma must be very serious about this camera if they loaned Lawrence and a selection of their best lenses for the review.
I have been shooting with a SD14 for three years and Nikon D70/90 for six years. I do not make my living with my cameras but derive a lot of joy for family and friend with my results.
The bottom line for me is no matter how refined and competent my Nikon gear is I cannot produce the visually stunning shots I get out of the Sigma.  My wife can see the difference, their is something special about the photos I get from the X3 sensor. Their is a realism that we lost with film becoming unpractical for regular folks today many call it Sigmachrome. If it were not for those redeeming qualities my SD14 would have been at the bottom of the Columbia River or sold for pennies on the dollar because as a piece of equipment to shoot with it is a bag of hurt.
Many say the SD15 is much improved functionally, what the SD14 should have been in the first place. But call me spoiled, I am too used to getting more when I upgrade my gadgets, like 2x the computing power.
I would be figuring out the SD1 today if it were not for the D3x size price, I have to work for my money and raise 3 kids.
Sigma's camera's are not at the point where they can be the be all for every body, when shooting sports, theatre, or general junk the Nikon gets the nod, something special like landscape, portraits, trains(I am a railfan) the Sigma gets the nod.
BTW I don't know where folks get the lack of color sensitivity or lack of DR I find that to be the opposite. Shooting on a cloudy day with the SD14 is truly a joy in its own right.
Paul M. Petersen
my gallery
www.peterzpicts.smugmug.com

Well I think you hit the point Paul,

your post actually shows that you owned a Foveon and been working with it, and print, for years, wich again is very different than the tech analysis based on theory and not real practise.
Yes, I also found a better DR when I had the SD14 (and people jumped on me here claiming the DoX graphs etc...) but this notion of DR is a vast field of muddy waters and a land of endless speculations from people who actually never worked with the cameras they discuss but just reading curves on some websites.

Lots of people take a gear measurement and say this or that has X points of DR. That is completly useless. It's exactly like saying that the average year temperature in Madrid capital is 20 grades.
That does not give you any indication is winters are cold and summer very hot, or if it's 20 the all year etc...

DR is linked to Isos. So yes the Foveon at base isos has a great amount of DR, but it falls apart very fast when you increment the isos. It is a weired behaviour but that's what the eyes and the prints are saying and not a general number taken isolated from the context. The numbers can talk all they want to, the ultimate truth is what our eyes are seeing and the prints are saying and I agree with all your lines.

 

[Dave Millier]

My SD14 produces dreadful results in LR by default - as horrible green cast over everything. Looking around at other SD14 users, it appears all the cameras have this cast in variable amounts. Rumour has it that Sigma profiled each Sd14 in the factory and burned a correction into firmware. SPP can read the correction information, LR can't. Certainly, SPP was much better than LR at producing acceptable default colour.

A couple of months ago I got a colour checker passport and made my own custom DNG profiles. The colour is transformed, the cast is gone.  I tried the same trick with my SD9 which has a tendency to yellowness I don't like -with much less success.  I think Foveon have gradually improved colour with each iteration of the sensor. The rest is software. The SD1 is a new chip design so we may have rebooted to the beginning of the cycle again as they learn how to control the colour. Or it might simply be software that was rushed to market and there is much more neutral colour potential in the camera.

At the risk of getting technical 

Profiles would certainly be a step in the right direction, but there is a limit since the unusual spectral response of the SD1 color channels means pretty high metameric failure which profiles can never help.

I agree.  However, the color processing for a Foveon sensor is far more involved than any Bayer sensor.  It is very possible that it is not practical for Adobe to support this type of sensor with any color fidelity at all without drastically altering their color processing.  It likely wouldn't be worth it for Adobe to do this even if Foveon was cooperative.  That, if true, I suppose could be interpreted as being Adobe's "fault".  Wild speculation of course.  But really, the color processing for this type of sensor is incredibly different from Bayer processing and that may prove to be a significant burden on any RAW processor that considers their product's color performance to be important.

And yeah, like you said, at this price the only discussions worth having are academic ones because it's not like anyone is going to buy the camera...

Ken

[Dave Millier]

According to DPReview's tests, the Foveon sensor has better highlight range but worse shadow detail/noise than is typical. So I guess with bright skies and the like it might look like DR is better because it preserves highlight detail. Until you examine the shadows...

Well I think you hit the point Paul,

your post actually shows that you owned a Foveon and been working with it, and print, for years, wich again is very different than the tech analysis based on theory and not real practise.
Yes, I also found a better DR when I had the SD14 (and people jumped on me here claiming the DoX graphs etc...) but this notion of DR is a vast field of muddy waters and a land of endless speculations from people who actually never worked with the cameras they discuss but just reading curves on some websites.

Lots of people take a gear measurement and say this or that has X points of DR. That is completly useless. It's exactly like saying that the average year temperature in Madrid capital is 20 grades.
That does not give you any indication is winters are cold and summer very hot, or if it's 20 the all year etc...

DR is linked to Isos. So yes the Foveon at base isos has a great amount of DR, but it falls apart very fast when you increment the isos. It is a weired behaviour but that's what the eyes and the prints are saying and not a general number taken isolated from the context. The numbers can talk all they want to, the ultimate truth is what our eyes are seeing and the prints are saying and I agree with all your lines.

 

[fredjeang]

According to DPReview's tests, the Foveon sensor has better highlight range but worse shadow detail/noise than is typical. So I guess with bright skies and the like it might look like DR is better because it preserves highlight detail. Until you examine the shadows...

Yes, the "gain" is indeed in the highlights. At based isos I could not see the degradation in the shadows but yes, when increment the isos, the reduction of DR in the shadows was more pronounced than with the Bayer and fall apart very fast.
 

[ndevlin]

Some great observations above. Laughed till I cried about the SD14 being a "a bag of hurt" as a camera.

And that really captures it. A camera might be somewhat better in some respect, but the quality at all levels is now so damn close that the real question is not is it better, but is it better enough to bother?

There's a reason serious pros generally use Nikon and Canon. It's just easier to shoot with, and whatever tiny percentage of 'improved' IQ one might get from a Sigma, or a Leica or an MF camera is vastly outweighed by the imperative of getting the shot.  A focused, publishable image with ideal composition is far more valuable than a less well-composed image taken a split second after the decisive moment, possessing 30% more resolution or a hint more DR.  

Integrating oneself with one's tools is the most important step in mastery of the art-form.  Not the tools themselves.  That's why the user-experience is SO important on cameras, compared even to IQ, which is now a game of inches.

On this front, Sigma has far to go. I wish them well on the journey, but they are not currently doing anything to help themselves.

- N.

[Aku Ankka]

DR is linked to Isos. So yes the Foveon at base isos has a great amount of DR, but it falls apart very fast when you increment the isos.

The DR can only drop by maximum of one stop per each doubling of the ISO. If shooting at a higher ISO for some reason were to drop the DR more, one should never shoot in that ISO, but just push in post processing - doing pushing is post causes one stop loss for each ISO stop pushed.

(Surprisingly losing a stop per level of ISO is actually better than it dropping less as that would indicate problems in ADC or elsewhere late in the imaging pipeline.)

[Aku Ankka]

BTW I don't know where folks get the lack of color sensitivity or lack of DR I find that to be the opposite. Shooting on a cloudy day with the SD14 is truly a joy in its own right.

Well, the folks get it from science - observations and measurements. Using silicon's wavelength dependent photon absorption will lead to very weak color separation. This is one of the reasons why Foveon images are noisy without plenty of noise reduction. The dynamic range is very limited because of noise - the 3T photodiode used by Foveon sensor is noisy as there is no way for it to eliminate reset noise (through correlated double sampling). No other manufacturer used 3T, but typically some 4T variant. To add to the insult of 3T high noise, all the three layers require their own 3T set.

I am happy you like your SD14 and if you feel it has good colors and high DR for your purpouses, great, but compared to other cameras it is lacking lots on both.

[Aku Ankka]

According to DPReview's tests, the Foveon sensor has better highlight range but worse shadow detail/noise than is typical. So I guess with bright skies and the like it might look like DR is better because it preserves highlight detail. Until you examine the shadows...

DPR doesn't measure dynamic range in any sensible way. They use some arbitary tone curves with arbitary mid grey point (or set by camera), and they eyeball the results they get.

Anyhow, what happens if we expose a camera stop less and then use a different tone curve? We get a stop more headroom, do we not? At DPR they don't realize that their measurements are rather silly when they separate head- and footroom. Dynamic range is the whole range, not one part here and other part there. How much of this dynamic range is used for headroom and footroom is depends totally on the photographer, it is not fixed in some way as DPR presents it.

AFAIK, the Sigma cameras have a tendency of exposing somewhat less than the competition if one applies no exposure compensation. This leads to different headrooom/footroom ratio than what the competition will achieve if one lets the camera to decide the exposure. But I must emphasize that this has nothing to do about the magnitude of the dynamic range itself.

[deejjjaaaa]

Lots of people take a gear measurement and say this or that has X points of DR. That is completly useless. It's exactly like saying that the average year temperature in Madrid capital is 20 grades.
That does not give you any indication is winters are cold and summer very hot, or if it's 20 the all year etc...

DR is linked to Isos. So yes the Foveon at base isos has a great amount of DR, but it falls apart very fast when you increment the isos. It is a weired behaviour but that's what the eyes and the prints are saying and not a general number taken isolated from the context. The numbers can talk all they want to, the ultimate truth is what our eyes are seeing and the prints are saying and I agree with all your lines.

 

please do not mix "average" w/ "maximum"... for 4.6 x 3 mp Foveon sensor the company itself published DR numbers and that will be a little over 10stops, not even 10.5

[deejjjaaaa]

My SD14 produces dreadful results in LR by default

you may be noticed that Adobe does not include their DNG profiles for Sigma cameras at all (no .dcp files - everything is hardcoded and probably was not changed or verified for ages).

[douglasf13]

DPR doesn't measure dynamic range in any sensible way. They use some arbitary tone curves with arbitary mid grey point (or set by camera), and they eyeball the results they get.

Anyhow, what happens if we expose a camera stop less and then use a different tone curve? We get a stop more headroom, do we not? At DPR they don't realize that their measurements are rather silly when they separate head- and footroom. Dynamic range is the whole range, not one part here and other part there. How much of this dynamic range is used for headroom and footroom is depends totally on the photographer, it is not fixed in some way as DPR presents it.

AFAIK, the Sigma cameras have a tendency of exposing somewhat less than the competition if one applies no exposure compensation. This leads to different headrooom/footroom ratio than what the competition will achieve if one lets the camera to decide the exposure. But I must emphasize that this has nothing to do about the magnitude of the dynamic range itself.

Agreed. Whether more range is in the highlights or shadows is just up to how the camera meter is calibrated and how the shooter uses it.

[ErikKaffehr]

Hi,

Well, my view is that the ETTR (Expose To The Right) maximises utilisation of DR. ETTR essentially mean that exposure is maximized still avoiding clipping.

Best regards
Erik

Agreed. Whether more range is in the highlights or shadows is just up to how the camera meter is calibrated and how the shooter uses it.

[Dave Millier]

The default profile for the SD14 is described as "Embedded". I believe the earlier versions of LR must not have supported DNG for Foveon (I read a support forum post to this effect from Thomas Knoll) but if definitely does now. I have four different profiles for my SD14 and they all produce different renderings. Thankfully because I couldn't put up with SPP. It's not a terrible program like Michael suggests and would be perfectly fine if Sigma were your only camera and you were happy to use a basic raw converter then finish up in Photoshop. But not if you want to preserve a single all raw workflow.

you may be noticed that Adobe does not include their DNG profiles for Sigma cameras at all (no .dcp files - everything is hardcoded and probably was not changed or verified for ages).

[peterzpicts]

I can say from my experience with the SD14 Sigma is well aware of the need to expose to the right, so much so I feel the meter is too easily spooked by shadows and will blow the highlights under certain situations and ruining shots.....grrr.
As far as shooting, Dynamic range means to me having details in the highlights and the shadows.  Playing with the x3 fill light I am able to stretch shots a lot further without having them look silly than I can with my D90 files a camera well regarded for its DR in the segment.
As far as work flow, I found getting good color out of SPP very hard prior to version 4. Back then I used LR and ACR and was able to correct colors easier than chasing my tail coming out of SPP.
Since SPP4 I do all my conversion their and buff and polish in Photoshop Elements, I have found the easiest tool to kill the dreaded green cast is to play with the mid tones in the green level channel, .95 usually does the trick for my camera. After that its just a matter of RGB level adjustment and final dose of mild USM.
Happy Snappin,
Pete
www.peterzpicts.smugmug.com

[JLK]

Paul,

That SD14 is one bag of hurt! I've shot with Sigma's dSLRs beginning with the SD10, and in many ways the SD14 was a step backwards. The SD15 is the 'good twin' to the SD14: much better color, great battery life, better ergonomics, and overall a much easier camera to shoot with than it's evil twin. My experience mirrors yours with Nikon and Canon's---the images from my SD15 (and previous versions) are just more satisfying to work with. For me, I'm able to get to a finished print much faster. And like you, I was expecting the SD1 to be a worthy (and flawed) upgrade for my SD14, for a reasonable ($2K) price. Oh well, they missed the mark on that

I find Sigma Photo Pro to be an interesting software package. The Fill Light (a single slider for tone mapping) is brilliant. Most of the controls are reasonable, and I don't miss a lot of other functionality (I take my photos to finish in Lightzone or PS). But in working with SPP5 and SD1 files? Ugh. They've got a bit more coding to do to make it bearable.

Making individual images can be a real delight with the SD15. But if I had to do event work on a deadline with a serious workflow---I wouldn't ever consider the camera (or an SD1 at this point).

Jim

[Ray]

As many of you know, I've been quite voluble in expressing skepticism about the 'qualitative' advantages of MFDB as regards its 3-dimensionality and extra crispness due to the lack of an AA filter.

Such qualities are clearly discernible, and I wouldn't deny they exist. But the price premium one has to pay for an MFDB system is a bit offputting, to put it mildly, not to mention the disadvantages of additional weight and general cumbersomeness of the system.

In a sense, it is therefore surprising that those who are already 'sold' on the 3-dimensionality of the MFDB and the additional 'crispness' of the images, should criticise the pricing of the SD1.

I get the impression, but correct me if I'm wrong, that the Foven design not only has the benefit of no AA filter, but the benefit of 'no need' to interpolate (or invent) image information because of the unequal balance between the number of red, green and blue diodes that the Bayer system has to contend with.

The 'qualitative' benefits of the Foveon system in this respect, disregarding pixel count, should exceed the qualitative benefits of MFDB.

[fredjeang]

As many of you know, I've been quite voluble in expressing skepticism about the 'qualitative' advantages of MFDB as regards its 3-dimensionality and extra crispness due to the lack of an AA filter.

Such qualities are clearly discernible, and I wouldn't deny they exist. But the price premium one has to pay for an MFDB system is a bit offputting, to put it mildly, not to mention the disadvantages of additional weight and general cumbersomeness of the system.

In a sense, it is therefore surprising that those who are already 'sold' on the 3-dimensionality of the MFDB and the additional 'crispness' of the images, should criticise the pricing of the SD1.

I get the impression, but correct me if I'm wrong, that the Foven design not only has the benefit of no AA filter, but the benefit of 'no need' to interpolate (or invent) image information because of the unequal balance between the number of red, green and blue diodes that the Bayer system has to contend with.

The 'qualitative' benefits of the Foveon system in this respect, disregarding pixel count, should exceed the qualitative benefits of MFDB.

Ray,

I don't understand, and it's not aimed to you, why people are seeing the grail to interpolate. They want reso on the cheap. It doesn't work like that. The SD1 is a stunning 15MP camera, and a good 20 ish MP interpolated. If you really want to take all the benefits of the Foveon you need to use the camera at its based resolution.

The moment you interpolate you loose quality, even if in the case of the Foveon you loose less in interpolation, you loose anyway.

You can say that after all it is a good performance that a 15MP camera would deliver a similar output of a 20ish MP bayer, yes, but where is the Foveon advantage then?
Well, the real advantage of the Foveon is that if you take 2 15MP cameras, the Sigma will be truly outstanding, but if you interpolate you'll end in a similar result of the Bayer. If you want to see the "miracle" Sigma is claiming, you need to keep the camera in its resolution wich is 15MP and compared it to other similar cameras on the market: the K5, the 7D etc...

15 packed in a small surface area is not the P65, not even a 20MP digital back. I think it's time to stop dreaming even if it's nice, even if it has 3 layers.

People want to see a sort of MF packed in a dslr... no, the Sigma is NOT the Leica S. It is very far, light years away in all aspects (sensor, built quality, features, image quality, value, image, pro service, lenses).

It's even far away from the Sony body wich is a truth pro body or the Leica M where you can mount on it the best glasses produced on earth.

So, if the Sigma had a large format sensor, it would probably smoke MF gear, but the fact is that it has not.

Also, the price, at that price or you have a brand that smells luxury and quality, or you get pro features. The SD1 does not have any of those.

People would like to feel that they finaly have MF quality for less than 8000euros and that the Robin Wood of the saga is Sigma. They might have a really big disapointment if they beleive so and sign the check.

It's like this sign I saw one day when walking on the beaches district in Toronto: swim at your own risk