The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Sigma SD1

davemillier

Member
Sigma are not Leica though, are they.

They are well known as a cost-effective 3rd party lens manufacturer... and that's about it.

They have made a handful of film cameras no one remembers and a small number of DSLRs that have sold just a few tens of thousands in a decade that have suffered notorious reliability and bugginess issues (I've just taken the bottom plate off my SD14 and and soldered in a bunch of wires in an attempt to solve the very common problem that if you dare to fill the 5 shot buffer, there is a high probability that the camera will freeze up). The cameras are seriously underspecced, years behind in performance, lack any kind pretence of standard features such as dust shakers, live view or video, suffer from poor LCDs, terrible battery life and are stuck with the SA mount (for which lenses are hard to find).

The majoritity of their sales have been to hobbyists, and in particular to a relatively small group of hardcore fans. Most of these people are not in the position of purchasing expensive highend cameras but have supported Sigma/Foveon, buying each iteration despite the faults of the previous model because they are committed to Foveon. Most of those people have waited a long time for a modern, competive camera from Sigma and the SD1 was hyped as the one.

The price is a marketing disaster even if they manage to find some customers and even if they make some money from it. Sigma senior executives publically speculated the price would be competitive with the Canon 7D. They have totally alienated their devout customer base - read the forums, universal condemnation from the only people that ever bought Sigma. Largely, they feel betrayed.

What happens to Sigma, if the fans turn their faces away and then the handful of pros and rich people who buy into the "baby medium format" scam, suddenly come to realize this a Sigma and you are expected to solder it yourself to make it work!
 

tom in mpls

Active member
Hi Tom,

In a Bayer pattern sensor color interpolation HAS to occur as full color info is not available for each pixel. This can lead to color moire if de-mosaic algorithms are not very good (or not present). Foveon sensors have red, green, and blue data for every pixel, so no color interpolation is needed.

Sigma's (and Foveon's) argument vis-a-vis number of pixels goes like this: Bayer sensors claim to have a certain number of pixels, however due to the nature of the Bayer sensor 25% are red, 50% are green, and 25% are blue. RGB info is interpolated based on data from nearby pixels. This degrades sharpness. For example, a 10 MP Bayer sensor would have 2.5MP red, 5 MP green, and 2.5 MP blue pixels. Thus in reality you only have 2.5 MP of actual RGB data before interpolation and there is positional shift information you need to take into account which can to mosaicing.

The Foveon's multi-layer design assures RGB info is present for every single pixel without interpolation. However, this means that in truth only 1/3 of the pixels involved are resolved into image data. Since Bayer sensor makers count every individual pixel in the final count, Sigma/Foveon decided to do the same.
This does not address the issue of moire, which occurs "occurs when a signal is sampled at a less than twice the highest frequency present in the signal". It seems to me that any digital sensor will be subject to this phenomenon.

Since the Foveon has a higher effective resolution, I would suppose it would require a much smaller pattern to produce moire; none the less, moire should occur under the right circumstances. Below is a link to a page showing an example of moire from a Sigma SD-9, which I believe was the first to use the foveon technology.

Link

I suspect that the foveon system may be less prone to moire, but I am not convinced that it is free of moire.
 

Arjuna

Member
Following Tom's link (and then on to the imaging-resource review of the SD9), there seems to be a distinction made between colour aliasing and luminance aliasing: e.g. from the imaging-resource review: "While not subject to the sort of color aliasing seen in striped sensors, the X3 sensor is still a sampled-data device,...", and also: "Thus, while it is pretty completely immune to color aliasing, it is much more subject to luminance aliasing than its competition.".

So, if one uses moiré in the limited sense of colour aliasing, then a foveon sensor is probably free of it. As an example, in Sean Reid's review of the Fuji X100, where there are comparison pictures from the X100, a Leica X1, a Leica M8, and a Sigma DP2, using a WhiBal card which has a pattern on it, the M8 shows much more (colour) moiré than the X100 or X1, and the DP2 shows none. In that case luminance aliasing would probably be much harder to see, and its occurrence would depend on the DP2's sensor pitch compared to the pattern spacing.

Good point Tom.
 

jonoslack

Active member
Sigma are not Leica though, are they.

They are well known as a cost-effective 3rd party lens manufacturer... and that's about it.

They have made a handful of film cameras no one remembers and a small number of DSLRs that have sold just a few tens of thousands in a decade that have suffered notorious reliability and bugginess issues (I've just taken the bottom plate off my SD14 and and soldered in a bunch of wires in an attempt to solve the very common problem that if you dare to fill the 5 shot buffer, there is a high probability that the camera will freeze up). The cameras are seriously underspecced, years behind in performance, lack any kind pretence of standard features such as dust shakers, live view or video, suffer from poor LCDs, terrible battery life and are stuck with the SA mount (for which lenses are hard to find).

The majoritity of their sales have been to hobbyists, and in particular to a relatively small group of hardcore fans. Most of these people are not in the position of purchasing expensive highend cameras but have supported Sigma/Foveon, buying each iteration despite the faults of the previous model because they are committed to Foveon. Most of those people have waited a long time for a modern, competive camera from Sigma and the SD1 was hyped as the one.

The price is a marketing disaster even if they manage to find some customers and even if they make some money from it. Sigma senior executives publically speculated the price would be competitive with the Canon 7D. They have totally alienated their devout customer base - read the forums, universal condemnation from the only people that ever bought Sigma. Largely, they feel betrayed.

What happens to Sigma, if the fans turn their faces away and then the handful of pros and rich people who buy into the "baby medium format" scam, suddenly come to realize this a Sigma and you are expected to solder it yourself to make it work!
HI David
I subscribe to all of this . . . but it does sound terribly similar to what everybody said about the S2 . .

I go back to my original point, which is that the camera really has something to offer with respect to image quality, then there's an argument for the pricing. If it doesn't there really isn't an argument for the camera, whatever the pricing may be.

Sigma's loyal fans are admirable, but they aren't really going to support a new camera.

What's more interesting is how Sigma thought that they'd make money from this.
 

davemillier

Member
I have it on good authority (Laurence Matson who prints the big Foveon prints for trade shows) when we went shooting down in Cornwall after Focus, that the prints from the SD1 are better than any of the full frame cameras and entirely free of noise at the useful ISOs. But he says they're not medium format competitors. Read into that what you want but based on past experience I would estimate resolution will be in the 30-35MP Bayer sensor territory.

That's nice, if it doesn't have lots of other problems - which it may have as it's a brand new chip architecture.

But it's still a basic APS-c camera that based on past experience will need about as many firmware updates as a Kodak to work properly (and about 3 generations of body). It will still take 10 secs to write a file and have a 7 shot buffer and likely freeze uselessly while you wait. At least you can entertain yourself twiddling the second command dial until Sigma realises it could have a use as exposure comp...

You can probably guess I'm a bit cynical about Sigma engineering -

HI David
I subscribe to all of this . . . but it does sound terribly similar to what everybody said about the S2 . .

I go back to my original point, which is that the camera really has something to offer with respect to image quality, then there's an argument for the pricing. If it doesn't there really isn't an argument for the camera, whatever the pricing may be.

Sigma's loyal fans are admirable, but they aren't really going to support a new camera.

What's more interesting is how Sigma thought that they'd make money from this.
 

Tim

Active member
The majoritity of their sales have been to hobbyists, and in particular to a relatively small group of hardcore fans. Most of these people are not in the position of purchasing expensive highend cameras but have supported Sigma/Foveon. Most of those people have waited a long time for a modern, competive camera from Sigma and the SD1 was hyped as the one.

The price is a marketing disaster even if they manage to find some customers and even if they make some money from it. Sigma senior executives publically speculated the price would be competitive with the Canon 7D.
Dave, your points above ring true for me as well.

I had a SD1 on my shopping list and even considering moving other items to get one BUT all that was based on this talk of "7D" pricing. :deadhorse:
One issue I think Sigma have not realized is all the "Sigma lens" sales they will miss out on by not putting a good Sigma mount camera within reach. You won't really have a lot of lens mount options with the SD1 like you do with m43 or Nex.

Right now I'll keep using my DP2 and wait for the "SD1 Saga" to settle out. :watch: If the price is not reachable then I'll play in other camps, perhaps wait for the rumored Sony A77. :sleep006:

Maybe a DP3 will be announced and hopefully I'll be able to afford that?! ;)
 

retow

Member
Well - I'm not so sure that the pricing is a mistake.

It seems to me that there are two possibilities

1. it doesn't offer better IQ than the D3x 5DII A900 etc.
in which case it's dead in the water . . . . and it'd be dead in the water at $3000 as well.

2. it Does offer better IQ than the D3x 5DII A900 etc.
In which case those who want Foveon, and want that quality, and who aren't necessarily requiring all the bells and whistles of the Canikon big guns might see it as a considerable saving over a MF kit. (especially if you start thinking about lenses).

When these pricing arguments come up nobody seems to remember about costs and production runs. If it costs $2500 to make, and you sell it for $7500 you only have to sell 1 camera for every 10 you would have to sell if it's $3000!
I don't quite agree with your costing logic. Your example was true, if all costs were purely variable in nature. As they aren't costs on a per unit basis are lower the greater the volume output as fixed costs and overhead get absorbed by a greater number. Furthermore, if Sigma sold more cameras, demand for lenses would go up as well, thus selling more SD1s would benefit the economics of the lens division and the company as a whole.
 

jonoslack

Active member
I don't quite agree with your costing logic. Your example was true, if all costs were purely variable in nature. As they aren't costs on a per unit basis are lower the greater the volume output as fixed costs and overhead get absorbed by a greater number. Furthermore, if Sigma sold more cameras, demand for lenses would go up as well, thus selling more SD1s would benefit the economics of the lens division and the company as a whole.
Of course I understand about this, I also understand that my example was simplistic - it was meant to be. But it doesn't alter the fact that making up for a big profit margin with increased sales is very difficult to achieve.

And . . Well - it's very unlikely that Sigma either would want to, or even could manage to buy enough components for any obvious economies of scale to be made. Added to which, for a new camera like this, support is obviously going to be an issue, and I doubt if they have the infrastructure to deal with that.

Unfortunately Sigma don't have the same cachet as Leica. Personally I always thought (and argued) that the S2 would be a success, beautifully designed and made and very expensive niche products from luxury companies usually are.

I'd be very surprised if this camera were a success, it depends really on how much of a niche it represents, and if serious photographers can see a real IQ advantage.

But this doesn't alter my point that making it cheaper would be very unlikely to make it more of a success for Sigma!

all the best
 

Jorgen Udvang

Subscriber Member
Sigma are not Leica though, are they.
No they aren't, but since Leica survived the M8, there's a chance Sigma will survive the SD1, some way or another. But that price is not very likely to survive, and neither is the $7,000 price that has appeared later. The SD1 seems to offer a lot in the area of image quality, probably more than the likes of the D3X. But most of Sigma's potential customers have lots and lots of Nikkor, Sony or Canon lenses that won't fit this new camera.

If they want people to change, or even buy the SD1 as an additional camera which is more likely for most, at least to start with, making the change easy would be the obvious thing to do. Sigma has done the opposite. Strange ways...
 

Corlan F.

Subscriber Member
Jorgen,
i wanted to stay out of this debate for various reasons, but seriously and with all due respect, i hardly understand why you keep saying that about IQ, especially from the official samples available.

Got all of them DLed, along with many 5DII and D3X files on my computer. The samples look a lot like the (good) 5D ones, but no way anyone with a D3X would ever be tempted to jump ship analysing these samples.

As a second body for 3-4 K$ maybe.

I don't have time now but will post a couple of similar shots in comparison. Or alternatively send a couple to you for what i can't post.



.
 
Last edited:

Jorgen Udvang

Subscriber Member
Jorgen,
i wanted to stay out of this debate for various reasons, but seriously and with all due respect, i hardly understand why you keep saying that about IQ, especially from the official samples available.

Got all of them DLed, along with many 5DII and D3X files on my computer. The samples look a lot like the (good) 5D ones, but no way anyone with a D3X would ever be tempted to jump ship analysing these samples.

As a second body for 3-4 K$ maybe.

I don't have time now but will post a couple of similar shots in comparison. Or alternatively send a couple to you for what i can't post.



.
Corlan,
It's a problem to find relevant, full resolution samples from D3X for comparison, so that would be interesting. At the moment, both cameras are well beyond my reach anyway, but I could hope for a full frame D7000 with the D3X sensor, somewhere south of $2,000, couldn't I. Or even better: a pro spec m4/3 camera with a smaller version of the SD1 sensor. Or is that called daydreaming :LOL:

The problem with me is that I would like to have a pocketable camera with MF resolution and pin sharp images with a series of f/0.9 lenses, preferably with a build quality that makes it usable as a chair and/or a hammer as well :ROTFL:
 

tom in mpls

Active member
The problem with me is that I would like to have a pocketable camera with MF resolution and pin sharp images with a series of f/0.9 lenses, preferably with a build quality that makes it usable as a chair and/or a hammer as well :ROTFL:
Isn't Leica currently working on a medium format pocket hammer? Don't even ask about the expected price.
 

woodyspedden

New member
Of course I understand about this, I also understand that my example was simplistic - it was meant to be. But it doesn't alter the fact that making up for a big profit margin with increased sales is very difficult to achieve.

And . . Well - it's very unlikely that Sigma either would want to, or even could manage to buy enough components for any obvious economies of scale to be made. Added to which, for a new camera like this, support is obviously going to be an issue, and I doubt if they have the infrastructure to deal with that.

Unfortunately Sigma don't have the same cachet as Leica. Personally I always thought (and argued) that the S2 would be a success, beautifully designed and made and very expensive niche products from luxury companies usually are.

I'd be very surprised if this camera were a success, it depends really on how much of a niche it represents, and if serious photographers can see a real IQ advantage.

But this doesn't alter my point that making it cheaper would be very unlikely to make it more of a success for Sigma!

all the best
Reminds me a little of the old saw......."Lose a little on each sale but make up for it with volume!"

We learn to relive our pasts every day

Woody
 

Tim

Active member
I would like to have a pocketable camera with MF resolution and pin sharp images with a series of f/0.9 lenses, preferably with a build quality
Laugh as we may, in reality this is what we all want. Likely One day it will exist IMO. What I am Not sure is, if there will be an improvement in Photography.
 

Quentin_Bargate

Well-known member
Good, but not the equivalent of a Bayer 45mp sensor, which it claims to be. Probably around 30mp equivalent, which is in line with my expectation of roughly 2x Bayer performance (subject dependent).

Given we should soon see 35mp cameras from Canon, Nikon and maybe even Sony, with all the professional bells and whistles, in my opinion it is grossly overpriced and under-specced to compete against them. $2,000 and it is a different story.

Quentin


I won't go as extreme as Jorgen to 160MP, but here's one of the samples taken by Carl Rytterfalk which I upres'd to 45.4MP (8251px x 5501px).

Whole image, resized:



100% crops from the upsized, 45MP file:





Full-res (45MP) file here: http://farm4.static.flickr.com/3463/5745152282_5cc4cd0953_o.jpg
 

Corlan F.

Subscriber Member
Jorgen -my apologies for taking several days to post some samples... long days, short nights...

Disclaimer: as said above, i didn't really want to get into the argument for various reasons. First because there's no point in (or being perceived as) placing any kind of pride in a piece of equipment, nor getting into a "mine's better" useless fight. I'm not. They are tools. Second, i might -extremely reluctantly- put up my D3X for sale soon, only for financial reasons needing to raise some extra cash for an upcoming, non photography related project. That said and clearly stated, let's get into the actual "details" :)


All the info below is just intended as casual documentation, sustaining why repeatedly reading that "the SD1 samples show better fine details resolution than the top end DSLRs, namely the D3X" is in my eyes a bold statement. And why, as a D3X user, i do not feel that these available samples would be inciting to consider a platform switch. Especially not a relatively comparable price point (yes, that counts).
Therefore, all the info below is absolutely not intended as a comparison or a demonstration, just a visual support for my previous dissenting remark.


Quickly browsing the hard disks, i tried to find some elements of "comparison", at least a bit relevant with some of the SD1 samples.


Technical notes:

In order to stay fair, all crops were made from jpegs (not RAW or TIF).
All of them are shot hanheld, and unless noted otherwise wide open or close, generally f2.8 or f4.
All are 8bit sRGB, and for most no sharpening at all (USM or else) applied for the purpose of this posting.

IOW what is posted below is clearly on the low side of the possible resolution on the D3X.

Using tripod / lens stopped down / MUP / Adobe RGB or ProPhoto RGB profiles / 16 bit / sharpening etc., one can expect to at least double the fine resolution showed here, and probably much more.

No processing except when noted, all 200% enlargements are single-click bicubic in Photoshop (no fractals etc.).

Most of them are definitely not demo material
, but... let's start...


(examples quoted from page 1)

One more. I've marked the area of the crop here. This one is taken with the 50mm at f/9.0.




No canyons here, but some water. Here's a file shot from across the river, in fact from across yet another elevated street on my side.
Unfortunately this one was handheld at f/2.8, so it's really on the softer side of things.


The final used crop was about 20MP, you can see it with the full original frame (dimmed areas).




100% crops:






200% crops:







(one can easily make the the hidden car logo)

Again, that's hanheld @f2.8.

Please also take note that i don't fancy myself as a master when it comes to optimizing vibration or stability. Though i did shot a lot of natural light, dim lit portaits at 400 ot 800 ISO in the analog age, it's not paramount for me and there's a lot of great photogs out there with better MO in this regard. Among our friends here, Steen (Bondo) pops up in mind as a good point in case, on many shots the sheer details he gets from his D300 is amazing (e.g. the recent crushed (or is it crashed? :D) cans) and i'd bet on par with or close enough to what many would achieve with a 5DII, SD1, D3X, and MF gear handheld. What you (Jorgen) do with panning is another example. Sure, in both instances crop sensor small mirror kinda helps, but still.

Just saying, again, the examples here represent at best an average minus level of what can actually be achieved with a D3X.
Certainly not the best resolution possible, and by far.


Some more stuff below...


-----------------------------------------------------------------------------------------------------------------------------------
 

Corlan F.

Subscriber Member
(cont'd)

Same day, a few minutes later. Strictly no processing, one-click standard camera profile RAW to JPEG then direct 200% crop in PS.
As you see not only it's a candid with the boat moving. Not a speedboat mind you, but still advancing at a good pace. I was using AF-S as usual (not AF-C).


General view:




200% crops:

lips and face expression:


detail in the shadows:




This next f2.8 handheld shot from another batch has strictly zero artistic value all the same, i stumbled upon this scene and wondering what these guys were doing so i just pointed and shot.
No processing, just like the previous example.

General view:




100% crop with bright / shadowed area:




200% crop same area:




Two more 200% crops:






The three files in the posts above in particular are certainly not demo material, but that's what i had off hand sharing similar traits with some of the SD1 samples, so why not... :)



-----------------------------------------------------------------------------------------------------------------------------------
 

Corlan F.

Subscriber Member
(cont'd)

The photo below one is from another SD1 sample:

General view:




100% crops:

in-focus area, lower right:


random area:




Here's a similar subject shot from the D3X (one that you (Jorgen) know well, but different since this file here followed the minimal process described above. Full image no crop, no processing, no NR, no USM etc.
Adobe profile RAW to JPEG in ACR)

This shot was then taken handheld at f3. there was some roadworks and i stepped out of the car to take it, traffic was resuming and the yellow jackets guys starting to yell at me, so it' was a one-shot deal at this distance.


General view:




100% crops:

In and out of focus areas:








Lower left corner (strictly uncorrected):




400% crop lower left corner (strictly uncorrected):




(a complete at 400% would result in a 24000 pix. long side and a 500MB+ file)

-----------------------------------------------------------------------------------------------------------------------------------
 
Last edited:
Top